miiminm10111:420=0111m:101m110111m electronics...tf 2604 voltmeter £395 tf 2702 inductor analyser...

MIIMINM10111:420=0111M:101M110111M

ELECTRONICS& WIRELESS WORLD

AUGUST 1987 1.95

Short -hauloptical fibres

Heaviside -champion ofinductancetwommommimaraCurls, divs,vectors andscalars411111161111111ffil

Clearer,simplerflowcharts

%AMP"

Variable -

speed, C -

format v.t.rsrprrobwromme

Satellitebroadcasting

Denmark DKr. 63.00 Germany DM 12.00 Greece Dra. 680.00 Holland DEL 12.50 Italy L 6500 Spain Plas. 700.00

Singapore SS 11.25 Switzerland SFr. 9.50 USA $ 6.50.

ELECTRONIC BROKERSThis is just a sample of our huge inventory - contact us with your requirements.OSCILLOSCOPES

Hewlett Packard180 TR 100MHz Scope Mainframe

(Mint)PhilipsPM3266 I 00MHz Storage ScopePM3305P 35MHz Digital Storage. I onlyPM3540 Logic Anal/ScopeTektronix465B/DM44100MHzScope/DMM4656100MHz Scope475A/DM44 200MHz Scope/DMM485 350MHz Scope608 Monitor634 opt 1,20 Display

Monitor5223 Scope WF (Mint)71041GHz ScopeM/F7603100MHz WF7704A 200MHzMainframe £25007834 Fast Storage ScopeM/F £6500R7603 100 MHz RackMINT WF £17507904 500 MHz ScopeWF £5500A large selection of7000 series plug -insavailable at up to 60%saving on list. Pleasecall for quotations.

ANALYSERS

£950

£3950£1925£950

£1550£1450£1950

£3000£995

£500£3600

£16000£1500

Hewlett Packard141T DisplayMainframe £20003580A SpectrumAnalyser £30003582A SpectrumAnalyser £75004193A ImpedanceAnalyser £52508552B IF Section £18508553B RF Section,1 I OMHz £2650MarconiTF 2300A Mod Meter £495TF 2370 110MHz SpectrumAnalyser £7250TF 2371 200MHz Spectrum Analyser £6500Tektronix7 Ll 4 Spectrum Analyser P. In308 Data Anal ser

£9250£600

SIGNAL SOURCESHewlett Packard4204A Digital Oscillator80076 Pulse Generator8011A-001 Pulse Generator

20MHz8601A Sweep Generator86408/001/002 Signal Generator

1024MHz86260A Sweep Gen. Plug-inI 2.4-18GHzMarconi2015-1 AM/FM Signal

GeneratorTF 2002BANt/FM Signal Generator

£200£750

£400£1000

£4250

£1500

£650£695

PhilipsPM5519ITVPattem GeneratorWavetek184MHz Sweep Generator10801 GHz Sweeper

£595

£425£1500

HEWLETT PACKARDCOMPUTERS

85A Microcomputer86A Microcomputer110 (45710BU) The Portable150B Touchscreen2673A Thermal Printer69408 Multiprogrammer7470A opt001/002 A4 2 -pen plotter7475A opt 002 A3/A46 -pen plotter7910H opt 015 Disk Drive82913A 12 inch Monitor82938A HPIL Interface82939ASerial Interface82940A GPIO Interface9111A Graphics Tablet9121D Single -sided Disc Drive9122D Dual Double -sided Disc Drive9133XV15MB & 270KB Disk Drive

£995£600£950£350£425£950£575

£1050£200

£75£150£100£250£825£450£775£950

9845A Desk Top Computer9816A Personal Computer9825A opt 001-002 Desk Top

Computer98256A 256K RAM board,

series 2009915AComputer

£1650£1800

£700

£250£750

GENERAL PURPOSE T & MFarnellSSG 520 + US 520 Transmission

Test Set

Fluke7220A Comms. Freq. Counter £350

Hewlett Packard4282A-001 Digital High

Capacitance Meter £750436A Power Meter £1050

467AAmplifier £6003403C True RMS

Voltmeter £18503406A Sampling

0 Voltmeter £12503465A 41/2 DigitD.M.M. £10053008 + 5305B1300MHz Counter £3505381A 80 MHzCounter £2258447A Amplifier £400

1611111118447D Amplifier £600MarconiTF 1 245A Q meter

£750TF 1246Oscillator £500TF 1313A LCR Bridge0.1% £350

TF 2173Synchroniser

for2016 £250TF 2603 RF

Millivoltmeter £300TF 2604 Voltmeter £395

TF 2702 InductorAnalyser £550

TF 2905/8 TV PulseGenerator £650

TF 2915 Data Monitor £1200TF 2950/5 Mobile Radio T.S. £995PhilipsPM5580 IF Modulator £3000PM558I RF Convertor £1000PM5582 RF Convertor £1000Tektronix520A Vectorscope (NTSC)521A PAL VectorscopeSi Sampling HeadS.3A Sampling Head1411C opt 03 TVSignal GeneratorA6901 Isolation Monitor4041 System ControllerP6451 DA ProbePM 102 Personality ModulePM 108 Personality Module

£3500

£3750£3950

£850£1250£5500£275

£4650£250£500£650

SEND NOW FOR OURLATEST BROCHURE

All prices are exclusive of VAT and correct at time of going to press Carriage and packing charges extra A copy of our trading conditions is available on request

0E1311,7_10ff,g,__YL:g El. (g

Electronic Brokers II (1),

r° Electronic Brokers140-146 Camden Street, London NW1 9PB

Ll Fax: 01-2677363. Telex: 298694. Tel: 01-2677070ENTER I ON RENA ( .NRI)

AUGUST 1987 VOLUME 93 NUMBER 1618

COVER

False -colour electron micrograph of inte-grated circuit leadouts bonded to chip padsis courtesy of Standard Elektrik Lorenz of

Stuttgart.

ISOLAN: A FIBRE OPTICETHERNET

774For a fibre optic network the most suitable

layout turns out to be a 'snowflake'topology

composed of multiport repeatersand active hubs.

by D.A. Rowarth and N. Howe

VARIABLE -SPEED VIDEOTHE ULTIMATE

782Report on new developments for C -formatprofessional video tape recorders up -dates

our seven -part reviewby John Watkinson

FLOW DIAGRAMS -A NEWAPPROACH

787Novel approach produces clear and easy -to -

maintain diagrams more suited to high-level and assembly language than convent-

ional flow chartsby David J. Sweeney

PIONEERS: OLIVER HEAVISIDE

789In his last years he bought his supplies via

the local policeman who would attractHeaviside's attention by blowing his whistle

through the letterbox.by WA. Atherton

COUNTDOWN TO SMALL -DISHSATELLITE TV

791A recent satellite broadcasting conferencerevealed details of Britain's direct broadcastsatellite tv service due to start late next year

NEW OSCILLATORS794

Three new oscillator proposals -a tunableactive -C for circuit integration, single-

amplifier active -R oscillator, and a currentconveyor design with single -capacitor

tuning.by M.T. Abuelma'atti and colleagues

Optical -fibre links will shortly be accom-modated in the IEEE802 'Ethernet' stand-ard for local area networks, as our featureon page 774 explains. (Photo courtesy ofPilkington Communication Systems ofRhyl.)

DATA CAPTURE FORFOURIER ANALYSIS

797How to get data into memory with a fast

capture system that digitizes in tenmicroseconds.

by Peter Finch and David Tayler

STEREOPHONIC IMAGELOCALIZATION

799Wavefront reconstruction technique

predicts image location for stereophonicsystems that use interchannel phase

differences.

TOP BRAINS EXODUS802

The brain drain is really about the loss ofskills rather than numbers; top researchers

going abroad are not being replaced.

SIMPLE VOICE -CONTROLFOR IBM's PC

807Take advantage of an undocumented butuseful one -bit i/o line to switch keyboard

mode with simple 'stop' and 'go' voicecommands

by B.J. Sokol

ALL ABOUT CURLS AND DNS

809A gentle introduction to vector field theory

for those who missed out in theirundergraduate days

by 'Joules Watt'

25 YEARS OF SPACECOMMUNICATIONS

813The world's first communications satellite,Telstar, was launched into low earth orbit

on 10th July. 1962.

TONE GENERATOR SYSTEM817

Microcomputer -controlled design scanseproms to obtain high quality waveforms

by S.J. Kearley

STREAM ENCRYPTION821

A stream crypto system generates keysequences containing equal numbers ofones and zeros so as to appear random

by Brian P. McArdle

NEW LOOK AT GAIN -BANDWIDTH PRODUCT

834Does the bandwidth of a feedback amplifier

necessarily have to shrink as thegain is increased?

by B. Wilson

Q AND OSCILLATOR STABILITY

843Recent interest in the origins and use of Qis widened by involving pendulums and

timekeepingby D.A. Bateman

COMMENT 769FEEDBACK 786,840

RESEARCH NOTES 797CIRCUIT IDEAS 803

TELECOMMS TOPICS 827APPLICATION NOTES 829

BOOKS 836PRODUCTS 845

RADIO BROADCAST 855TV BROADCAST 858RADIO COMMS 859

UPDATE 862APPOINTMENTS 865

ADVERTISERS INDEX 870

ELECTRONICS & WIRELESS WORLD 767

The Board LevelI.

rsales OffICCIKAW-yoifir ticketta success.

r: E ri 55 ON REPLY CARD

IC OMCommunications

ICOM introduces the IC -R7000. advance-, ,-chnology. continuous coveragecommunications receiver With 99 programmable memories the IC -R7000covers aircraft marine. FM broadcast. Amateur radio. television and weathersatellite bands For simplified operation and quick tuning the IC -R7000 featuresdirect keyboard entry Precise frequencies can be selected by pushing thedigit keys in sequence of the frequency or by turning the main tuning knobFM wide. FM narrow AM upper and lower SSB modes with 6 tuning speeds0 I I 0. 5 10 12 5 and 25kHz. A sophisticated scanning system provides instantaccess to the most used frequencies. By depressing the Auto -M switch theIC -R7000 automatically memorises frequencies that are in use whilst it is in thescan mode. this allows you to recall frequencies that were in use Readout isclearly shown on a dual -colour fluorescent display Options include the RC -12r.fm r d r-rri-.P syr.thPsi:.(1- ,;:(1 HP I hPaciph,-,nPs

release rush me details of the IC -R7000 and my nearest ICOM dealer.

Name

AddressTel

Post to: ICOM (UK) LIMITED , Dept WW, FREEPOST,Berne Bay, Kent CT6 8BR. (no stamp). Tel 0227 363859.- J

ENTElt '29 ON It E:1'1.1 CARD

SMALL SELECTION ONLY LISTEDRING US FOR YOUR REQUIREMENTS WHICH

MAY BE IN STOCK4 4

Latest bulk Government release - Coss°, Osc.Ilos, our CPU, iS II I.

Racal Solid Stale Communscatron Receivers -RA1217 "Racal RA171. Conontrn.c.O.on Rece.vers

Frac,. Synthes.sersMA250

MAI 3S0 MA2S9G..RA '37 and RA);

RA98RAI21.

PR' 5rGanstelMatrA Printers '

Army. 1.eld telephone sets

Don 10 Telephone CableN.qIn Venvinq .ntra - rod

Staticperiscopes

.nveriors

recorders,,us TF 893AIrepoenCy

AY Riotlers and penSignal Generators vac

RacalTeldrono plod .its

EXPORT TRADE AND QUANTITY DISCOUNTSJOHNS RADIO, WHITEHALL WORKS,

84 WHITEHALL ROAD EAST. BIRKIENSHAW.BRADFORD. BD11 2ER TEL NO.10274 684007.

WANTED: REDUNDANT TEST EQUIPMENT -VLVES - PLUGS - SOCKETS, SYNCHROS

ETC. RECEIVING AND TRANSMITTINGEQUIPMENT

173MHz FM TELEMETRY + TELECOMMANDRADIO LINKS Versions for Remote Switching,

Voltage Monitoring + Serial DataTransmission.

Line of Sight Range typically over300 metres with 1mW ERP + over900 metres with 10mW ERP

Approved to MPT1309

Each Module 86 x 176 x 45mm+ requires only low current 7.2V(or optional 12V supply)

ADENMORE LTD(Manufacturers of Radio Equipment since 975)

27 Longshot Estate, Bracknell, Berks. RG12 1RL Tel: 0344 52023

ENTER S ON REPLY CARD

76$ ELECTRONICS & WIRELESS WORLD

COMMENT

ENT( )RPhilip Darrington

DEPUTY EDITORGeoffrey Shorter, B.Sc.

01-661 8639

TECHNICAL EDITORMartin Eccles01-661 8638

PROJECTS EDITORRichard Lambley

01-661 3039NEWS EDITORDavid Scobie01-661 8632

DESIGN & ILLUSTRATIONAlan Kerr

01-661 8676

Roger Goodman01-661 8690

ADVERTISEMENT MANAGERMartin Perry01-661 3130

Michael Downing01-661 8640

CLASSIFIED EXECUTIVESusan Platts01-661 3033

ADVERTISING PRODUCTIONBrian Bannister

(Make-up and copy)01-661 8648Jackie Perry01-661 8649

PUBLISHERShobhan Gajjar

01-661 8452

auREE

SID

BUNFCSPUBLISHING

ForethoughtIn the beginning, around the time someone discovered how to walk upright, wasscience or knowledge. During the succeeding several million years, the pace ofscientific advance was scarcely dramatic, and its effect on life barely discernible.

There was the occasional flurry of excitement over fire and iron and the wheel andthe husbandry of grain crops, but in general the seeker after knowledge and itsapplication to progress did not have to spend a great deal of time worrying about thepossible consequences of his actions in the years to come.

Passing lightly over the ensuing period with the merest of nods at the inventionof gunpowder and the discovery that coal will burn, and arriving in the late 19th and20th centuries, one is made very aware that invention and discovery quickly beginto carry with them an imposing array of unlooked-for repercussions. It is nowbecoming ever more difficult to foresee and, therefore, avoid quite disastrous sideeffects of progress in almost every field of acitivity.

The list already accumulated is formidable: everyone can reel off a long statementof afflictions that have come about through well meant and apparently civilizedinnovations. As examples, the use of earth resources cannot have seemed likely tocause excessive depletion in the foreseeable future to early exploiters: and it musthave appeared improbable that the pushing of fumes out of a few factory chimneysand horseless-carriage exhausts would lead to acid rain and brain damage. Butthese effects have happened and they are the responsibility of scientists andengineers.

By definition, engineers are problem -solvers. Since, by their ingenuity andknowledge they have created these problems, albeit unknowingly (for the mostpart. at any rate), it is now within their area of responsibility to alleviate theireffects. They are presented with the greatest problems to solve that anyone can everhave seen. Pollution, the ransacking of the earth's store of materials, the recklesshaste to harness atomic power, the threat of military hardware, destruction offorests, over -exploitation of animals, the use of electronic "entertainment" toconvert whole populations into zombies: all these and many more are problemswaiting for a solution - quite the reverse of much modern engineering effort.

Perhaps it is not now possible to call a halt, or even to arrange a breathing space,in the insane rush to develop more technologies to satisfy immediate requirementswith not a thought, or at least only a quickly suppressed one, for the future, thefuture being far too remote to worry about. But if it is, the future begins to appearless than inviting-and relatively short.

It is, of course, little use addressing oneself to "science" and "engineering". Theattitudes of scientists and engineers must be radically changed, as must those whoemploy them and who direct their efforts. And since the problems to be solved areglobal ones, any effort to orchestrate the problem solving must also be worldwide,undertaken by a body comparable in might and authority to the United NationsOrganization.

Whoever takes the job on, assuming anyone thinks it necessary, had better startsoon, because the day the last of the ozone layer disappears is not a good time tostart thinking about the advisability of rethinking spray cans.

Electronics & Wireless World is published monthlyUSPS687540. Current issue price 11.95. back issues£2.10 at Retail and Trade Counter. Units :&2. BanksideIndustrial Centre. Hopton Street. London SE I Telephone:01-928 3567. By post, current issue £2.25. hack issues iifavailable' (2.50 Order and payments to 301 Electronicsand Wireless World. Quadrant House, The Quadrant.Sutton. Surrey SM2 5AS. Cheques should be payable toReed Business Publishing Ltd. Editorial & Advertisingoffices: Ettit' Quadrant House. The Quadrant. Sutton,Surrey SM2 5AS. Telephones: Ediloriat 1)1.661 :1614.Advertising 01-661 :HMI 01-661 8469 Telex: 892084B1SPRS 1EEPi Facsimile:01.6W :1948 Groups II & 1111Beeline:01-661 8978 or 01-661 8986 :Milhaud. 7 data bits.even parity. one stop -bit. Send ctrl -Q. then EVVW to start:NNNN to sign off. Subscription rates: I year normal

rate, 1:23.40 UK and £28.50 outside UK. Distribution:Quadrant House. The Quadrant. Sutton. Surrey SM2 5ASTelephone 01-661 3248. Subscriptions: Oakrield House.Perrymount Road. liar wards Heath. Sussex RH16 3DHTelephone 04444 59188. Please notify a change of address.USA: $116.011 airmail. Reed BusinessPublishing LISA Subscript IMIS Office. 205 E. 42nd Street.NY 10117. Overseas advertising agents: France andBelgium: Pierre Mussard. 18-20 Place de la Madeleine.Paris 75005 United States of America: Jay Feinman.Reed Llusiness Publishing Ltd. 205 East 42nd Street.New York. NY 10017 Telephone 12121 867-2080 Telex23827. USA mailing agents: Mercury Airfreight Inter-national Ltd. Inc., b Englehard Ave. Avenel N.J. 07001.2nd eta,s postage paid at Rahway NJ. Postmaster - sendaddress to the above.©Reed Business Publishing Ltd 1987. ISSN 0266-3244

INSTRUMENTSCOMPONENTSEQUIPMENTACCESSORIESOSCILLOSCOPESHAMEG . '203 6 Dal 20 MN/ I rod, probes £314.00204 Vilith sweep delar £418.00205 atonal storage dual 20M11/ £527.00605 Dual CO M111 plus sweep delay £583.00CROTECH wall Component tester3031 Single 20 tritil 95cm CRT

HITACHI V212 Oual Nace 20 1411/8 r 10cm £320.00GENERAL 05T5m smote trace S MK3 CR £139.09HUNG CHANG Dud' t' I A

620 Dual 20M H/ componenl test £295.00635 Dual 30M111 Sweep delay £399.00615 Dual 15MHZ bat Mains £399.00650 Dual 5081111 Sweep delay £579.00

£199.00

Official orderswelcomed.

Export supplied.Please add 15 I VAT(UK orders only)

UK FREE DELIVERY

111",

Add storage to your scope MAsampling £195.00

COUNTERS AND TIMERS

270/40V AC mains 8 digit TED

Counters 0.INZ resolution 5mV sensitivity

I MANZ 2 ranges

600M112 3 ranges

1000M NZ 3 ranges15110IANZ 3 ranges 199.00Timers 0.001N2 Resolution 100M1115mY sensifiyily

OPI 0 Standard model £219.00OP100 Inc) display hold etc stopwatch RPM etc

£285.00

243 0 24 welts 0/3 amps OC245 0 24 volts 0/5 amps DC154 5/15 volts 0/4 amps OC

£99.00£126.00£175.00

DC POWERSUPPLIES2400 AC input Stabilisedvariable output voltagesDirect meter readingvolts and amps

£43.26£56.26£33.26

Also 138V BC 3 amps max Freed Ins mere, I £13.000 30* 15A DC with volt meter £29.950 30* 0/2 amps OC LC0 display £150.00

GENERATORS 220/2400 AC operation

2c:1;r:F=41::rtnztt

00 001

FUNCTION sine sg triangle III 0 PJ500 01141 to 5001011 £110.00T6101 0 0211/ to NOW £110.00T6102 0 2111 to ?FAH/ £160.00past Square Complement 111 0 PT6105 SHIM 51A111 £105.00TV -VIDEO PAL VHF 'UHT 55/6/65MHZsound carriers RGB and SYNC 0/P s AM FM sound mod

Colour etc £199.00

GENERAL EQUIPMENT81/110 Probe kits 1250M1111 £10.00200 IODANI 110MHZ R.0 OSC £65.17221) 20012 213131012 OSC £78.22C512130 Bench Dipltal Cap Neter £89.00114301 LCD Olpital Temperature £59.502o16 Audio Power Meter £220.0014501 Elidnanic insulation tester £65.00PF512130A 200M112 Pocket TED Counter £75.50OM 350 8CH Scope Mumpleser £179.00141)516250NHI TR DIP Meter £43.042016 1H2 !MHZ Generator 06 £180.00

DIGITAL MULTIMETERSk3 digit LCO displays

506 15 range 200mA DC 2MEG £19.96 Alab 14 Range 2A DC £21.50szer 13 range plus Hle Mil £26.04sex 28 range 113A AC DC £35.65am 28 range 10A AC DC 20 MEG £33.50703001°° version of above £39.50576 20 range auto - memo, r gold 10A AC, OC £34.50615 19 range plus rite test 20 MEG IBA DC £30.395010 31 raixe I OA AL DC tun NIA AC /0C £36.50T51357 30 range 10A ACS DC bun £39.005010 EC as 5010 cap Hie cowl £52.13

ANALOGUEMULTIMETERS1015 15 range omits! I nig £7.501035 19 range MK V 10A oct10.39103)1 ig rams Km ac is rovv£12.61NC2111 19 Ramp 2111l/V 12A AC/DCE15.61

20213 20 range i0A OC Hie 201( /V £19.0930101Z 24 range 10A DC boner 300/11 £20.83155m 21 range 10A AC/OC 2011/V £22.575050E 41 Range FFT 20M1 P £28.65133011 26 Range 10A AL: OC 301( V £25.17

TT

LCR METERTc200 18 range digital handheld teller6 ranges resistance

7 ranges capacitance

7 ranges inductance £85.00

SWITCH MODE PSU'sOpen Frame Models I/P 100/240V AC$ X 510 to 5v ladmetablel (Damp £7.78Enclosed Model85/97W 5* 6A 1211 4 75A 121105A £19.95

PROBE/PULSER6106 20MHZ logic probe & Weeper £16.95625 50M112 Probe A pulser osolsOoot £22.9507 208111 logic probe E 9.52

BENCH DMM'S . ge=3 Digit 1.0 10A

AC/OC 20 meg Buzzer etc

322 025°. £ 89.0032100.1%TruellMS £119.00

STOCKISTS FOR:Logic Analysers Insulation Testers temperature meters Security Systems Public Address Communications Tools Computer Equipment TV/VideoAccessories (See Catalogue)

AC CLAMPMETERS With gaveling leadsMO 300 tops 15 ranges)6009 ranges) 1 A OHM £32.83

e-

4300 Autorange Digitai 300A

S0011 cone test £85.00

DIGITAL CAPACITANCEMETER 8 range LCD 200pt to ?DOOM F 0Pushbutton selection 160131 £55.00

ORDER BY POST OR PHONE. OPEN 6 DAYS A WEEK FOR CALLERS.

HEfIRY'S ralE)143onicsatiaj 301 Edgware Road. London W2 1 B N. 01-724 3564

-7-7- 404 Edgware Road. London W2 I ED 01-724 0323TELEX 298102 TRANS G

FREECATALOGUE

EDUCATION AND ACCOUNT ENQUIRIES 01-258 1831ENTER 17 ON 1a:14.1( CARD

FORTH INSTRUMENTATIONCOMPUTER

................11,3",731: .........................

...........31111,10

A powerful control computer based on the new Hitachi6303Y and high level language Forth. 100mm x 72mm.30K bytes RAM, 16K dictionary RAM/PROM, 256 bytesEEPROM, 16K Forth. You can attach 64 key keyboard,LCD and I=C bus peripherals. Built in are interrupts,multitasking, time of day clock, watchdog timer, fullscreen editor and symbolic assembler. 32 parallel and twoserial ports. Single power supply and low power 3mAoperational mode.1 off £194.95 including manual and non-volatile RAM.

TDS9090

Triangle Digital Services Ltd100a Wood Street, London E17 3HXTelephone 01-520-0442 Telex 262284 (quote M0775)

ENTER 28 ON REPLY CARD

STEREO STABILIZER 5

Rack mounting frequency shifter for howl reductionin public address and sound reinforcement.

Mono version, box types and 5Hz fixed shift boardsalso available.

SURREY ELECTRONICS LTD.,The Forge, Lucks Green, Cranleigh.

Surrey GU6 7BGTelephone: 0483 275997

I NEW, AFFORDABLEPCB CAD

FROMCHANNEL MICRO PRODUCTS LTD Maximum circuit board size greater than 2' x 2'. Powerful zoom facility and fast screen re -draw. Full on -screen editing. Mouse & Trackball compatible. Hard copy at 1:1, 2:1 & 4:1 from Epson FX or RX printer.

. Plotter drivers for most popular plotters available. Low cost BBC Micro hardware keeps system cost below £1500.

Software only £240.

Please write or telephone for full details.

227 Canterbury Street, Gillingham, Kent ME7 5XB.Tel: 0634 570256 / 63228


7711ELECTRONICS & WIRELESS WORLD

IEEE PROGRAMMABLE INSTRUMENTS9818 PROGRAMMABLE CURRENTVOLTAGE CALIBRATOR

* 5 Current ranges from 100uA to 1* 4 Voltage Ranges from 10mV to 1* 0.01 % Accuracy.* 5 ppm Resolution.* IEEE/GPIB/HPIB/IEC Bus Comp* 30% Overrange on all ranges.

9814 PROGRAMMABLE VOLTAGECALIBRATOR

A. * 6 Ranges from 10mV to 10VOV. (100V and 1kV ranges optional).

* 0.01 % Accuracy.* 5 ppm Resolution.

atible. * 30% Overrange on all ranges.* IEEE/GPIB/HPIB/IEC Bus Compatible.

Other programmable units include -24 Way Switch, 900 Way Scanner, Programmable Resistance, Power Supplies and Multi Function Calibrator

Full catalogue of calibration equipment available, which incljdes:Inductance, Resistance, Capacitance Standards, Decade Boxes, Voltage References, DVM and Multifunction

Caliorato-s plus IEEE Programmable Instruments.

MOD STD 05-24, 05-26 NATO APPROVED

TIME ELECTRONICS LIMITEDBotany Ind. Est., Tonbricge, Kent, England.DIRECT SALES (0732) 355993. Telex 95481


Thermal Linescan Recorder Dual channel high definitiondry paper grey scale recorderusing a revolutionary thermalprocess

time st free islet operalienRugged corstruction Low cost of owneHigh reliability

ANALOGUE OR DIGITAL INFUTSIEEE 488 CONTRO_ INTERFACE

Multi -purpose system for applicatior s in: Spectral analysis Sidescan sonar ima Lofargram displais Sate lite imaging

Medical imaging Facsim le copy Video hard -copy Comp.r.er graphics

Seismic profiling Infra red linescanFor further, cnforrnati cot,-

11114DINFTV Dowty Maritime Systems _imited

WAVERLE ELECTRONICS

WAVENLEY DIVISIONWw.erltv Road. Weyelo dh. Cassel. Englani C T3 5HLTel Vir-yrrou'r (0303('34738 -eex 4147' Fax (0305) 77791$

A Dm"' .9ectron. Sysien, .Tvovo, Conoany



Not only in height is the range of Clark Masts wide but also in thefield of application. Every model, mechanical or air -operated,has been created in response to customer requirementand proved in service; for over 25 years.Major users in the Communications Industry,Broadcasting, Civil Authorities and MilitaryCommands worldwide, have all contributed and benefited from Clark Masts'reliability and ceaseless engineeringimprovement. V

Write for yourcopy of ClarkMasts' "FastGuide to MastSelection.. andsee what we mean.

CLARK MASTS CLARK MASTS TEKSAM LTD. Binstead. Isle of WightEngland Tel 109831 63691 Telex 86686

CLARK MASTS TEKSAM NV . Woudstraat 213600 Genk fielworn Tel 011 38 08 31 Telex 39354

I :NTER I I ON REPLY ('ARD

M & B RADIO (LEEDS)THE NORTHS LEADING USED TEST -EQUIPMENT DEALER

SPECTRUM ANALYSERSHP1417 8554L 8552A 1250MHzHP8551B 10MHz 40GHzHP8755A 10MHz 18GHzTektronix 491 1 5-40GHz

C5500C1500

£750£1350

OSCILLOSCOPESRecent quantity purchase of Cossor CD150

Compact Solid Stale Oscilloscopes. Dualtrace 35MHz with X1 -X10 probes andprotective cover Tested and in goodcondition C175

HP183A 1830A 1841A 500 500MHz 0950Tektronix 465 100MHz DL Sweep £750HP1741A 100MHz Storage £1500HP180A 50MHz DL Sweep plug in £550HP183A 4 Trace 50MHz 0350HP182C 75MHz 1825A Time base £1250Telequipment D83 50MHz DL Sweep £325Philips PM3233 10MHz C195

GENERAL TEST EQUIPMENTHP8620A 1-2GHz 5 9.9GHz

8-12 4GHz £5000HP333A Distortion Meter £550HP654A Test Oscillator £750Radford DMS 3 0350Radford LDO2 DMS2 Pair £375Marconi TF2120 Wavform Gen £250Wavetek Programmable Waveform

Gen £450Marconi TF2331 Distortion Factor

Meter £425HP400EL AC Voltmeter £200HP3400A RMS Voltmeter £450Racal 9903 Counter Timer £200Racal 9917A Counter £385HP3311A Function Generator £295HP5381A 80MHz Counter £175Wavetek 1001A 200MHz Sweeper £275Marconi 2300 Modulation Meter £375Marconi 2330A Wave Analyser C425Gen Rad 1362 UHF Osc 220 920MHz C295Bradley CT471 Electronic Multimeters RF

measurements to 1200MHz C35Tektronix 308 Data Analyser as new £750Gould 920-13 Logic Analyser C750Marconi PCM Test Set TF2801/4 TF2823 POA

Bradley Scope Calibrator 156 C425Marconi TF2012 400-520MHz AM FM 0375Marconi TF144 10Hz 72MHz C75Marconi TF1060 450-950MHz AM FM 0195Marconi TF1066 10-470MHz £250

PM5324 100KHz 110MHz Sweep 0350Thorn TV Sync Generator 0550Tektronix 529 Waveform Monitor £250Bradley 171B Multimeter Cal £850Marconi TF1026 Freq Meter 5- IGHz £35HP415D SWR Meter £250Lyons PG73N Pulse Gen £250Marconi TF2212 X -Y Displays £65RS Poly Scopes 2 5-1 2GHz £185Marconi TF1313A Bridge £450Marconi TF2700 Bridge £195Marconi TF2701 Insitu Bridge £185Marconi TF1245A TF1246 1247 £575HP419A DC Null Voltmeter £150Fluke 6160A Synthesizer 30MHz £275MI Sanders No 19 Attenuators C95Eddystone EC958 101(Hz 30MHz C450Eddystone 830,7 Receivers C350Racal RA17L 117 etc from £175Nagra III Tape Decks £450Uher 40001 Tape Decks £85Racal 9056 Selective Analyser £250Philips PM8110 Mono Recorder £250Hitachi HV62K CCTV Cameras C85Brandenburg 807R 30KV PSU £150Philips 1649 PSU 150V 7A £250HP6453 PSU 0-15V 0-200A £500Marconi TF2019B Noise Gen £235Marconi TF2092C Receiver C450Racal MA79H Transmitters £300Marconi TF2333 Measuring Set £475Various Wendel & Golterman Line TestInstruments in stock eg Levell GeneratorsSPM6 Levell Meters SPM12 SignalGenerators PS3

MISCELLANEOUSHarris 1KW FM Broadcast

Transmitters C55005KW VHF Dummy Loads £200Ruston Diesel Alternator sets designed forremote and unattended operation. 62KVA 3phase. 100Orpm. turbo charged as new £3750

ALL PRICES PLUS VAT AND CARRIAGE

86 Bishopgate Street, Leeds LS1 4BB.Tel: 0532 435649

IKESTREL ELECTRONICCOMPONENTS

All items teed to

LTD.manufacturers spec.9

Many other items available.'Exclusive of V.A.T. and Post and Package'

1. 50i 1 + 50 1 50 74LS00 0.13 0.10 6821 1.40 100 2764-25 2.60 2.4074LSO4 0.13 0.10 6502 3.10 2.50 27128-25 3.80 3.6074LS08 0.13 0.10 6522 3.10 2.50 27256-25 3.95 3.7074LS21 0.15 0.12 6809 4.00 3.50 6116LP-3 1.50 14074LS37 0.13 0.10 6810 0.70 060 6264LP-15 2.40 2.0074LS74 0.13 0.10 8085A-5 1.30 1.20 62256LP-15 12.00 10.0074LS86 013 0.10 8255-5 1.90 1.60 4164-15 1.00 0.8074LS138 0.25 0.20 4 meg Crystal 0.30 0.25 41256-15 2.60 2.3574LS139 0.25 0.20 LM324 0.25 0.20 Z130ACPU 1.40 1.0074LS244 0.40 0.28 LM339 0.25 0.20 Z8OACTC IA0 1.0074LS245 040 0.28 LM386 050 0.40 Z80AP10 140 1.1074LS373 0.35 0.28 LM556 0.35 0.25 1488 0.25 0.1874LS374 0.40 010 ICM7555 0.65 0.55 1489 0.25 0.1874LS399 0.30 020 ICM7556 0.85 0.75 27C128 300 2.60

All memory prices are fluctuating daily, please phone to confirm prices.

178 Brighton Road,Purley, Surrey CR2 4HATel: 01-668 7522

ENTER 53 ON REPLY ('ARD

I.*NIL so

BOXING

PROBLEMS??

LOOK

NO

FURTHER!!Choose from our standard range, or if you are looking for a specific size. then wecan produce YOUR OWN CUSTOM BOX WITHOUT ANY TOOLING COSTS, withall holes, slots, PCB grooves, etc, already machined in, ready to assemble.For sizes and details of Custom Service please contact us.

BAFBOX LTDUnit A, Park End Works, Croughton, BRACKLEY. NN13 5LX

Telephone: 0869 810830ENTER 57 ON ItEl'IN CARD

772


ELECTRONICS & WIRELESS WORLD

ca t Tra 01-943 4477Carstonsell Cardontsuy

Carston2nd USER

TEST EQUIPMENT& COMPUTER PRODUCTS

WHATEVER YOUR EQUIPMENT NEEDS ARE,

DO PLEASE CALL US!

FOR LATEST STOCK

INFORMATION RING 01 943 4477

CARSTON ELECTRONICS LTD.2-6 QUEENS ROAD, TEDDINGTON, MIDDLESEX, TW11 OLR

TELEPHONE: 01-943-4477. TELEX: 938120 CARLEC-G

EN FE It 37 UN REPLY CARD


Isolan: a fibre -opticPthernet

the most suitable layout is antmultiport repeaters and active

774 ELECTRONICS & WIRELESS WORLD

The use of fibre optics in local areanetworks is a relatively simple applica-tion of the technology; point-to-point

optical links with a maximum distance of2km and 10Mb/s Manchester encoded datarate do not demand a high degree of techni-cal sophistication. Nevertheless, the applica-tion to local area networks as in Isolandemonstrates the ability of fibre optics toovercome with ease the difficulties withsafety and performance which may be ex-perienced with metallic data networks span-ning large sites and several buildings.

International standards for local area net-works providing open systems interconnec-tion have made substantial strides in recentyears, but there are still significant areaswhere even at the physical layer thestandards -making process is not complete.One such area is the use of fibre optics in lansconforming to the IEEE 802.3 standard'which up until now has covered only thecoaxial medium. The appearance of thisstandard, which has effectively supersededthe earlier joint proprietary Xerox Ethqrnetspecification2, and more recently of its inter-national counterpart3 has greatly encour-aged the adoption of Ethernet lans, sinceusers know that they can source complyingproducts from a number of vendors and beassured of compatibility.

The ability to mix media will furtherenhance the attraction of Ethernet as aneffective simple -to -use Ian solution for avariety of applications, and work is in handto extend the standard to specify fibre -opticelements. These Ian components may beused as part of a coax network or to provide afully fibre -based installation. This furtherevolution of the standard is not proceedingon an abstract basis; the work is undertakenby engineers who are directly involved inadvancing the technology. The productsdescribed have been developed in associationwith this ongoing work in the standardsarena.

Ethernet is a contention bus system oper-ating at 10 Mb/s. with up to 1024 nodes in asingle network. Access by a node to transmitonto the network is determined by a carrier -sense multiple -access protocol with colli-sion detection. If a node needs to send apacket of data it waits until the bus isinactive and then starts to transmit. If a nodeelsewhere on the network has also started totransmit, a collision will occur which isdetected by each node; both abort transmis-sion before trying again after a randomdelay. Asa result of the Manchester encodingscheme used for packet transmission, thereis a fixed mean d.c. content of each packetwhich makes identification of collisions re-latively simple in a copper -based network,though this simplicity may not apply in thecase of a fibre optic implementation.

FIBRE OPTICINTER -REPEATER LINK

As published, the IEEE 802.3 standard de-fines a multidrop bus network consisting ofcopper coaxial segments which may be inter -linked by repeaters to form a network withup to three copper segments in any serial

This article is based on information first published bySPIE at Fibre Optics 86.

Isolan fibre -optic hubs, available in 6 and 12 -port versions, form the basis of a simplestar network and provide the means of connecting devices up to 1/2 km apart on anall -fibre network.

Fibre -opt c multipart repeater allows the interconnection of up to seven fibre opticsegments and one ethernet segment Used with a fibre -optic hub, an all -fibre networkcan be implemented with a diameter of up to 4km.

Co-cx segment

MAU

TransceiverAUI

DTE

MAU

AU! cable

OTE

_LLMAU

MAUAUI

_fa_MAU

Fig.1. Typical ethernet configuration based on theIEEE802.3 standard illustrates basic networkingcomponents. Included are repeaters that allow upto three copper segments in any serial pathbetween corrected equipments. (AUI: attachmentunit inte-face, MAU: multiple access unit DTEdata terminal equipment).

her ne

Repeater

AUI

MAU

DIE

"SI

MAU

AUI

MAU

path between connected equipments, Fig. 1,and permits the network to have a maximumdiameter of 1.8 km. It allows expansionbeyond this by the inclusion of repeater linksto connect segments together giving a maxi-mum diameter of 3km (Fig.2).

Studies on network characteristics sincethe standard was published have shown thatup to five repeaters may be used to connectactive segments in series within the overalldiameter constraint of about 3km, provided

that the repeaters meet specific performancerequirements. An appropriate amendment isplanned for a future issue which will detailpossible topologies in a more flexiblemanner.

The 802.3 Working Group established atask force to draft a vendor -independentfibre -optic i.r.l. standard so that differentsuppliers may implement end -equipmentwhich will interwork over an optical link.This fibre -optic inter -repeater link enables


isolated copper segments to be incorporatedinto a single Ian without introducing metal-lic connections between them -a key featurefor safety. security, reliability and resistanceto e.m.i. when the segments may be inbuildings separated by as much as 2km. Thetask involves the definition of a fibre -optictransceiver in terms of its interfaces, func-tionality. and compatibility. The configura-tion is shown in Fig 3: a duplex fibre -opticcable interconnects fibre -optic transceiverswhich take the place of standard coaxialtransceivers and are plug -compatible withthem at the repeater's attachment unitinterface connection.

At the time of writing, the IEEE 802.3working group has approved the draft of anadditional section defining the fibre -opticlink, and allowing the use of all popular fibresizes. but optimum for 62.5/125 and 85/125microns. Following balloting by the IEEEstandards board this is expected to be fullyapproved by the third quarter of this year.

FIBRE OPTIC ETHERNET LAN

Since coaxial carrier -sense multiple -accesslans are configured as passive buses a sim-plistic approach would be to attempt toduplicate this structure using optical fibre.However, the undirectional nature of lightdoes not readily lend itself to the require-ments of bidirectional data transfer neces-sary in a bus structure, nor is it feasible tohave many optical taps in series.

The approaches considered for im-plementing a fully fibre Ethernet mainly use

duplex fibretransmission. In effect the bus has shrunk toa point at the hub of the star, and the duplexfibre links extend this out to the nodes Fig.4.Collisions now occur at the hub, and argu-ably the most crucial aspect of designing andspecifying the network lies in determiningan effective method of detecting them. Thesimple electrical method of collision detec-tion by examining the mean d.c. content ofthe signal level as employed in the coaxialcase is not applicable for superimposedoptical signals which may differ substantial-ly in amplitude. A further requirement isthat it should allow the specified maximumof 1024 nodes to be achieved - and boththese needs should preferably be met by thesimple assembly of off -the -shelf networkcomponents from a mix of vendors withoutalignment or adjustment, with easy additionof nodes to cover later needs for expansion.

Four basic approaches have been ex-amined for an optical Ethernet to be plug -compatible with the coaxial system at thelevel of the attachment unit interface. Thesediffer in their characteristics. as reviewednext.

Passive starwith collision detection by signal analysisThis approach relies on an entirely passivehub consisting of a transmissive star couplerwhich distributes to all nodes a packettransmitted from any node. Fig.5. The re-ceived power level at a node can varysubstantially depending on which node istransmitting, and care must be taken insetting up the network to minimize therange of variation so that strong transmit-

Fig.2. Connection of three ethernet seg-ments achieved using two repeaters withfour transceivers expands area coverageto a maximum diameter of 3km. (MAU:multiple access unit)

AU1

Duplex fibre -optic cable

Fig.3. Fibre optic inter -repeater link canbe used to connect buildings and main-tain electrical isolation, and as a byproduct increase the network diameter.Fibre -optic transceivers are plug -

compatible with coaxial transceivers atthe repeater's AUI connectioin. (AUI:attachment unit interface, MAU: multipleaccess unit)

more difficult as the number of ports at thestar coupler rises, so this type of approachtends to find more application for the small-er networks.

Passive starwith detection by time domain analysisTo overcome the requirements for carefulmatching of received optical powers, a sys-tem relying on signal tuning has beenproposed4. The network configuration is

identical to the passive star with signalanalysis, but collision detection is performedin the transmitting transceiver by timingfrom start of transmission to start of receiptof the looped -back packet. If a transmissionarrives early it must be a colliding packet.The disadvantage is that each node mustnow be tuned for its particular path length tothe coupler. Unfortunately, the collisiondetection can never be 100% effective be-cause a tolerance band must be allowed onthe correct loop -back delay.

Active starTo overcome the difficulties with collisiondetection posed by the use of a passive star.and active hub approach may be used, asshown in Fig.6. The fibre optics consist ofindividual duplex point-to-point links: thereceivers can accept a wide dynamic range,so the received optical power can vary widelyamong the ports without impact on thesystem performance.

Collision detection is achieved within thefibre optic transceivers by the simultaneousdetection of transmission and reception of apacket at the optical ports. The hub facili-tates this process by performing the follow -

Fibre -optic cable Multipart fibre -optic repeater DTE

Fig.4. Fibre -optic ethernet in which the bus is effectively shrunk to a point at the hub,duplex links extending this out to the nodes. Collisions now occur at the hub and thedesign of collision detection becomes crucial. (FOT: fibre -optic transceiver, AH: activehub, MPR: multiport fibre optic repeater.)

ters do not swamp weaker ones.Various schemes have been proposed by

which a transmitting transceiver can ex-amine the signal to detect collisions, includ-ing sensing of average power level, monitor-ing of pulse widths to detect code violation.bit -wise comparison between transmittedand looped -back signals, as well as patentedmethods of coded high -amplitude pulses inthe preamble. At best. collision detectionwith this type of system is only modestlyeffective, and only if the received signals arewithin a few decibels amplitude of oneanother. 'Tuning' the network becomes

ing switching algorithm on the incomingpackets.

- If only a single packet is being received by aport, that packet is output from all portsapart from the port on which it is received.

- If more than one packet is received by thehub, a collision exists and the hub trans-mits valid carrier with null data to all itsports as output. In some variants of theproposal a special distinguishable signal isused by the hub rather than the null data.


This method give 100% effective collisiondetection within the full dynamic range ofthe receiver.

Hybrid starThe remaining technique, shown in Fig.7, isoptically the most complex. It also addressescollision detection by an active approach atthe hub, but only collision detection isactive, with the data path still using atransmissive star coupler. The collision in-dication signal is superimposed on the trans-mitted data, rather than replacing it as in theactive star approach; the transceivers have todetect this superimposed signal to becomeaware of the collision. This hybrid systemcan provide collision detecting approachingthe effectiveness of the active hub system;however, as with the passive cases, additionof extra nodes is not straightforward and thedesign is best suited to smaller networks.

Examining the four techniques againstthe essential criteria, the only approach thatfulfils all the requirements is the active star,provided that adequate redundancy techni-ques are used in the design of the hub.

The standard for fibre optic Ethernet isstill in development at the time of writing.The IEEE 802.3 working group has re-quested the fibre optics technical advisorygroup to advise on the suitability of thevarious design approaches, with particularreference to collision detection. In view ofthe technical advantages of the active huboutlined it seems likely that the approachwill finally be adopted if the specificationcalls for the same interoperability at networkcomponent level as in the coaxial case.

NETWORK COMPONENTS

To provide a flexible implementation of fibreoptic or mixed -media Ethernet complyingwith the existing and developing require-ments of IEEE 802.3, a group of networkcomponent products is required.

Fibre -based products (fibre optic trans-ceivers, active hubs, and repeaters) de-veloped by BICC Data Networks in associa-tion with the fibre optics advances in the Ianstandards arena are discussed next.

Fibre -optic transceiverThe transceiver is a fundamental compo-nent, since it establishes the duplex fibre -optic link upon which the other productsdepend, and is incorporated in hubs andrepeaters. The transceiver interfaces be-tween a standard IEEE 802.3 a.u.i. electricalconnection and twin 9mm SMA opticalconnectors, and is implemented using acustom thick -film hybrid receiver. The linkcarries a continuous signal at all times.contributing to the excellent 25 dB dynamicrange and allowing detection of fibre breakor low light level conditions.

The wide dynamic range permits usewithout adjustment on links of any lengthup to 2km with any of the graded -indexfibres likely to be selected as standard forlans (50/125, 62.5/125.85/125, 100/140 mic-ron. The stand-alone transceiver is pack-aged in a robust and compact die-cast case.

Fibre/optic two -port repeaterThe two -port repeater contains one internal

The transceiver has been designed to use all the popular fibre -optic cable diameters forease of installation, and features jabber protecticn, idle signal modulation, loss of lightindication, and switch -selectable 'heart-beat' (SQE) test.

DTE

DTE

DTE

DTE

L

E

RxControl

CD

jom

Tx

DTE

DTE

Fig.5 Passive fibre -optic hub acts as a junction box directing and redirecting the lightpaths, but is sensitive to all errors on the network Collision detection is only effective ifthe received signals are within a few dB of each other. Fig.6. In the active hub system,which isolates segments as well as regenerating signals, collision detection is achievedby simultaneous detection of transmission and reception of a packet at the opticalports. Switching algorithm (see text) gives 100% effective detection within the fulldynamic range of the receiver.

DTE

CD

Jam

DTE

DTE

Fig.7. In this hybrid star the data path is passive whilst collision detection is active. Anindication signal is superimposed on the data rather than replacing it as in the activestar approach. Transceivers detect this signal to become aware of a collision. Schemeis very effective but addition of other nodes is not straightforward and it is best suited tosmaller networks.


fibre -optic transceiver, the other port beinga standard a.u.i. connection to an externaltransceiver. Two units can thus be used toform an inter -repeater link between coaxialsegments, as shown in Fig.3. The unitperforms the normal functions associatedwith an 802.3 repeater (preamble recon-stitution. fragment extension, data retim-ing, and auto -partitioning of faulty seg-ments) and also incorporates the fibre -optictransceiver functions of loopback of data.jabber protection, 100% collision detection.idle signal modulation, and loss of lightdetection and indication.

Multiport fibre -optic repeaterThe repeater section of IEEE 802.3 is writtenin such a way as to allow a repeater withmultiple ports, able to form a repeatingradial hub within a network. The fibre -opticmultiport repeater which has been de-veloped has seven fibre ports and one a.u.i.port, and allows economical interconnec-tion of multiple fibre links as well asoptionally connecting into a copper Iansegment. As with the two -port repeater, itcarries out preamble reconstruction, frag-ment extension, and signal regeneration.Automatic partitioning and reconnection isprovided to isolate a faulty segment from therest of the network, implemented with theminimum number of components to ensurehigh reliability, the same thick -film hybridoptical receiver circuits as the transceiver.

The optical links for a multi -port repeatermay be connected to another multi -portrepeater, a two -port repeater, an active hub,or to a fibre -optic transceiver, giving greatflexibility in network construction.

Fibre -optic hubThe active fibre hub detects collisions andsignals them to all ports, as well as distribut-ing received packets to all ports except thesource port. It differs from a repeater in thatit is not required to perform the functions ofpreamble reconstruction and timing regen-eration. The design is constructed in twoconfigurations: Six fibre ports plus an a.u.i.port or 12 fibre ports plus two a.u.i. ports,allowing the simple establishment of anoptical star network with 100% effectivecollision detection, and with optional cen-tral connection to a copper portion of thenetwork or to facilities such as gateways toother services. As with the other products.each port has detection for fire break or lowlight level.

OPTICAL NETWORKCONFIGURATION

A single hub will provide a star with up to 12arms, but by interlinking active hubs andmultiport optical repeaters to form a snow-flake topology, as shown in Fig.4, extensivebut cost-effective networks can be built. Thisprovides a simple means of achieving

maximum -size fibre -optic lans with up to1024 nodes and up to 4km in diameter usingoff -the -shelf units and without the need foradjustment or optical alignment.

Fully effective collision detection isguaranteed throughout the network, andthe functions of the repeaters ensure trans-mission performance to IEEE 802.3 require-ments as well as providing automatic parti-tioning of the network in the event of localfaults. Connections to copper segments canbe established by using the attachment unitinterfaces on repeaters and hubs, permittingmixed -media Ethernets in which the mostappropriate medium is used for each portionof the network.

References1. IEEE 802.3: CSMA/CD Access Method andPhysical Layer Specifications. July 1983.2. The Ethernet, a Local Area Network. Data LinkLayer, and Physical Layer Specification. Version1.U. DEC. Intel. Xerox Sept 1980.3. ISO DIS8802/3: CSMA/CD Access Method andPhysical Layer Specifications. July 1983.4. S Moustakos and H Witte: Passive Optical StarBus with Collision Detection for Local AreaNetworks. 100C 83. June 1983.

Don Rowarth is director of advanced tech-nology at MCC Data Networks. Nigel Howeis in the market support group.

Software interfaces for IsolanThe first networking product thatallows interworking of the popularPC networking systems as well as

any mini or mainframe computer usingindustry -standard protocols has beenannounced recently by BICC Data Net-works. Together with a plug-in controllercard, the new software enables PCs toshare 10mbit/s networks for rapid com-munication by allowing different pro-tocols to work together in sharing filesand printers on the same network, and todirectly access mini and mainframe hostsat the same time.- The new software, called Multi -Protocol Support, allows communica-tion in a network between, for example.an IBM-compatible PC, a mini such as aDEC VAX, and IBM mainframes, andusing MS -Net, Novell NetWare and DR -Net, all of which can co -exist on the samenetwork.

Ease of connection is simplifiedthrough additional packages for simplenetworking of small PC clusters. Multi -Protocol Support ensures optimum net-work efficiency' by stopping any oneapplication from monopolising the con-troller.

The handler is provided with and with-out ISO transport; both are on the dis-tribution disc as 'terminate and stayresident' applications. The controller it-self is an intelligent IEEE802.3 Ethernetinterface with the extensive input buffer-ing necessary for multi -tasking. Supportfor Unix environments -in two formats,

Ferranti AT

4ASNet WkstnPC

P

6-01IBM

Sperry IT (AT)

Gateway Ser4

Gateway

TangogateGateway

I CL

IBM XT

'MSNet WkstnAnirryst -2 screens

Tangogate -sawn

&PrInnt

UI

Torch 's

CAW :V1.%%%%%%%%WiAS%%%%%

1.1

chTT. AOC (AT)

MSNetServer

J./

/SOLANCoal/pc/ration

for Networks 87

Olivetti SP M24

NetWareWorkstation

Zenith AT

NetWareServer

the MIT implementation and a file trans-fer protocol - is avaiable separately, as isthe PC Serial Connection for terminalemulation.

The company says its commitment tothe international Open Systems Inter-connection standards underpins its Iso-Ian range of products. which now includeMulti -Protocol Support. Both OSI andnon-OSI protocols can now run simul-taneously on the same network. "It'sabout time users were given the oppor-tunity to make full use of the benefits that

ISOLANMonitor

ISOLANTest Unit

ICL Ouattro

COOS:RNe

,C ornmsstonsw Vowdata

ITT Extra

coosI Dunk

DRNHConnissionaof

)

an open system can give," said Ian Laur-ence, BDN's marketing director "and notbe forced into buying products withdubious lifetimes".

Formed in 1984, BICC Data Networkhas already established itself as Europe'sleader in the design. manufacture andmarketing of local area network productswith a turnover approaching 12 millionper month. Its Isolan products are thefirst to fully comply with IEEE802.3standard. and won a Queen's Award forTechnology earlier this year.


IMPROVEYOUR LANGUAGE

Don't end up cursing the limitations of relay logic.Essex have a range of systems, tailored to your

specific requirements and offering a choice of highlevel languages, from 'C' to BASIC.

One Essex module can provide up to 48 intelli-gent i/o lines at a much lower cost than thetypical programmable controller, so you can builda control system that is cheaper, has moreprogramming power and much greaterflexibility. Take the FORTH Microcard for in-stance, it provides the powerful FORTH pro-gramming language, 40 i/o lines and serial

communications - all for around £100.We have a range of industrial grade interfaces

including analogue, opto-isolated digital, andmains power modules. And, for a completepackage, the Essex Development System pro-vides you with everything necessary for ef-ficient software preparation.The Essex modules are building blocks

for system design, provid-. ing a rapid solution to your

industrial control- problems.

So throw out your swearbox, and give us a call.

EssexElectronics Centre

WIVenhut; Pale, Culur,ustur, Essex LO4 s3U lei ,020b) abb0b9distributed by

RC.S.IIRICROSYSTEMS LIMITED141 Uxbridge Road, Hampton Hili. Middlesex TW12 1 BL Tel 01-9792204


Used equipment - with 30 days guarantee. Manuals supplied if possible. This is a verysmall sample of stock. SAE or telephone for LISTS. Please check availability before

ordering. Carriage all units f 16. VAT to be added to total on Goods and Carriage.

OSCILLOSCOPESTEKT ItONix 577 Cory.. I car ei L2500TEKTRONIX 475 Dual is. e 200tH. Dee, Sweep LI500TEKTRONIX 46513 Oitai Trace 100tH, Dewy Sweet, 1I.200H P 17(50 poor 'tell e 27,051.17 Delay Sweep 1I2007 E KTRONix 455 Dos lin e ; 005677 Delay sweet, L909TEKTRONIX 454 Dual tint e 1 5.751H,i Dela, sweet, L5507ELEQUIPMEN7 075 Dual hate 5051Hr Delay Sweet L325H P 180A 501.107 Delay Sweep 1325ONLY

If QUIPM1 NT Cad Dow Ise e 29407 Dinar sweet 1250Si L ABS SE I I : Dow Trai e I iiMil7 AC ,,, 1 urinal :isOpeiaboo L165TEL EQUIPMENT 554 i.,,,, 4,,RAH., L100TEL EQUIPMENT 554A Single Ira, e I 01.07 . Soot ',IA, £110PHILIPS PM32 0 51 DRA .7 (hi. Henn 17 7.4 , 1500IEL EQUIPMENT 01.164 510847a Lhiai ivi e 101.0yHCornpkteL50

SPECIAL OFFERS, . ;WHO C=lay Sweep.

,, NOW ONLY L1110 ea -rut Font Protect. Corer C.ontaining 2 Probes and Vievemyho

1

ARTRON OSCILLOSCOPE CD1400 Dual Beam 15MHz withC.NL V L85 ea

- ,VALVE TESTER CT 160 Suitcase styleMos (Kb L4

es Precwon CRT Restore, 'Analyser Model 467 Supplwd witses and manual. p&p17 ONLYL125, P'ow Cola., Bat General°, KG) 8 Test Patterns

..; '54 ONt 0 140 ea i

MULTIMETERS

Ca we LIO. 8 MAY Complete wan batteries. Hart and case 190

TEST SET No 1 (Military version ol 800 81Comptele withines leads 8 cairnica. L65' LEADS suitable or VOMEIERS Red 8 &act. with ?Ctoc C14,0

'sOPfriT'O'IP LS, M:ii 7iMETERS Tint, Yoder Aso et7

L 10

MARCONI MOBILE RADIO TEST SET 182950 ON, Y 1950 earn

GENERATORSMARCONI TF2006 TM 10- ILX)0MH., 1I.400MARCONI TF2006 AM 4- SOOMH.; 1950MARCONi 712008 AM Fla 10.0i, 5101407 11200MARCONI 7F2015 AM FM 101416 5201.1Hz with 102171Synchroniter 11.000MARCONI 1820026 south Synchronise, 712170B I 0717 88M.1..1M 06 1500V ARCONI IF 106613 Ala 0 IA 10MHz-470MH: L350V ARCONI 189954,5 1 5-2201414, Narrow D.A. L250ADVANCE Type SG63E AN FM 4-2305116 1100ADVANCE Type 50628 AM 1501(Fla 22014 Hz £45

BRUEL & KJOER EQUIPMENTL3.50011.000L1000

' 1I.750Pe . i 7 11.400

1,e0 Ckcinat L400

COMMUNICATION RECEIVERSy ALAI 8417 500106-30MH7 troth mania, . ONLY 1I50 ea.'; DOYSTONE 73074 4BOKFle- 30MH7 *nth manual ONL Y L110 earn

k DRIVE PSU 2400 in 5S . ie s .. . ;A Jul Suemin /10 ea

71' HIV KEYBOARD as in LYNX MICROf Push to mateed ONLY L5 ea

7. 'CHED MODE PSU 120 0 256 `78, 15A 240 14A

,'4on 3000 PROM COPIES (Up., uP to 101120

,.,, i 7 1150 eachMARCONI.1

MARCONIMARCONI.

MARCONIMARCONIMARCONIIKlITHIFY

()Meter T8I245 with111246 a TF1247 JtS00P POWER METER 431C with Thermistot Mourn 10MH7-10011, te" 4-18GH2 . moo

IN MIL L NOL THEY ER 182603 508147-1.5001151, LaoVALVE VOLTMETER TF2600 10/17-109/1/ Imo-3000

,0 fopinsau UNNERSAL BRIDGE T F 27C IUNIVERSAL BRIDGE 1E2700 Battery Operated L150TF1313 Universal Bodge L100

P RF MIL LIVOL TMETER 41IA 5001(16-50051111 L60ELECTROMETER Type 6108 wan Adaptor 61013 1500

11 ART RON TRUE RMS VOLTMETER VM1464 L50 .,[,! RM5 VOLTMETER 3400A 1014,1CMH i L400

NEEQUIPMENT iAMEG OSCILLOSCOPE 605 D,oR Trace 60MH7 Delay Sweep

roponent Tester L583"MFG OSCILLOSCOPE 203 6 Dual Trace 20Melt Component

L114

. . L219Jo ...KN.-, 7/1.. 07,1,0 0 7, r oie ' L285

7...K STAR FREQUENCY COUNTERS 1P80 t 7

I 001.1117 L99.' sye WO -600106 1126.' eut 1000- IGO, 0175

17 1( STAR JUNIOR 500 FUNCDON GENERATOR ....1110,.. 7.1 :40

ANA lb013 SPECT RUM ANALYSE R 0 sal L3 000HNITAR LOGIC ANAL YSER 1.52160 le Clviii .

1,17 / 1 750isVDAR LOGIC ANAL YSER 702080 101,s,/1 - -

1050

ISOLATING TRANSFORMERS - 240V Input,

115 en h Non IS 10006 L6 ea. h p&p 1

10' 200VA L4 each pip I.:

L39 50133 50

III

STEWART OF READINGREADING,

Telephone: 073468041_BERKS RG6 1PI. AE

IIIMI"IMITIMMIt 110 WYKEHAM ROAD,

Callers welcome 9am to 5.30pm. MON-FRI. (UNTIL 8pm. TliURS)


06°(F-c'e\e6.1\P''''

4:"7"-s''tV *10 \

00 0 0 c o

ere

_00cjos.ec'6\ \e ck vsg

6 0 c,c3 441.,0,"?,(ae2P e,< 4

,<Za 9\-{ 444-,,t.ee

\1Ae

0\>\

\CeNv"

4 6,`°\Y es -

6 ce

6eN- c,'25 o).

44''40 c,e 6e 0c -N\e

eec07-

Nox sc,00,

cr 9,0

N.1*O

R\-

e0.1 W"` p.0



The Archer Z80 6bCThe SDS ARCHER - The Z80 based single boardcomputer chosen by professionals and OEM users.* Top quality board with 4 parallel and 2 serial ports,

counter -timers, power -fail interrupt, watchdog timer,EPROM & battery backed RAM.

* OPTIONS: on board power supply, smart case,ROMable BASIC, Debug Monitor, wide range of I/O &memory extension cards.

from £185 + VAT.ENTER 21 ON REPLY CARD

Thc5owman 68000 MCThe SDS BOWMAN - The 68000 based single boardcomputer for advanced high speed applications.* Extended double Eurocard with 2 parallel & 2 serial

ports, battery backed CMOS RAM, EPROM, 2 counter -timers, watchdog timer, powerfail interrupt, & anoptional zero wait state half megabyte D -RAM.

* Extended width versions with on board powet supplyand case.

from £295 + VAT.ENTER 22 ON REPLY CARD

act -wood Data 6y8tem LtdSherwood House, The Avenue, Farnham Common, Slough SI,2 Tel. 02814-5067

EMSPower Systems

4

3E

EMS manufactures DC Power Supplies andBattery Chargers both linear and switch modein a range from 5 VA to 3.2 KVA.Also a complete range of Standby, UPS andMains Stabilizer Systems 35 VA to 1 KVA. EMSspecialises in the manufacture of customisedproducts and has a full design anddevelopment facility.

EMS (Manufacturing) Limited,Chairborough Road,High Wycombe,Bucks HP12 3HH.Tel: (0494) 448484


TARGET ELECTRONICS16 Cherry Lane, Bristol BS1 3NG

Telephone: 0272 421196

Subminiature 0 type plugconnectors. Gold platedcontacts. Solder bucketterminals

9 -way male

9 -way female

9 -way hood

I5 -way male

15 -way female

I5 -way hood

25 -way male

', way female

y hood

1 10 10038p 28p 25p40p 30p 26p

34p 25p 22p44p 33p 29p48p 36p 32p

38p 28p 25p60p 48p 38p65p 55p 45p

42p 32p 27p

RS232 WALL PLATESwith 25 pin D socket.,

25 WAY D GENDERCHANGERS

Three typesMale to MaleMale to F e r, £3.15Female ,

Link 7500: 19013485Telex: 94012950 AVON G

DATA SWITCHBOXES

N.. 'Met required H gP quaia, longlile heavy duty operationParallel36 way centromcsI to 2 way changeover i28 95Parallel36 way centronicsI to 0 way changeover 13795Serial25 way RS232

I to 2 way changeover !2650Serial

way RS232I r.4 changeover /3795AMSTRAD PC 1512Prised Lead ParallelFlexible moulded cable 25 wayD male. 36 way centronics.I 8rn £450Parallel Printer Lead36 way 36 way centronics. 2m

1600

05\ENTE R 6 ON REPTICARD

IEEE 488 CONTROLLERSfrom M A Instruments Ltd

IBM PC and Compatibles including Amstrad PC1512.Half length interface card. software including Turbo Pascal or Quick BASIC interfacecable and comprehensive operating manual. C295

Acorn '8' and MasterInterface in separate :ase. software in PROM. IEEE 488 cable and comprehensive250operating manual

New software for existing owners of Acorn GPIB controllers which offers compatibilitywith the new IEEE 488.2 standard. E28

All controllers offer a command structure which follows the IEEE 488.2 standard. Thissimplifies the control of existing GPIB devices. reducing the number of odrnmands

needed and is compatible with new devices using IEEE 488.2 format commands.

Desisned. manufactured and supported in Great Britain

For more information contact.

M A INSTRUMENTS LIMITEDMoordene, Axtown Lane, Yelverton, Devon PL20 6BU.

Tel: (0822) 853585Prices exclude, VAT and delivery



VALVES L,PECIALQUALITY Please orlon, for tarn quolatun V A T included

A 1, ,0, 1,4041293 7.00A2900 12.75AR8 1.15ARP3 1.15ARP35 0.70A TP4 0.90812H 6.90CY31 1.40DAF 70 1.75DAF96 0.75DE 722 32.80DF 92 0.65DF 96 0.70DH76 0.750L92 1.100086/87 0.65DY802 0.70E92CC 2.80E18000 11.50E1148 0.58EA76 1.60EB34 0.70EB91 0.60EBC33 1.85EBC90 0.90EBC91 0.90E BF 80 0.95EBF89 0.80EC52 0.65EC91 440EC92 1.85E CC81 0.95ECC82 0.95ECC83 025ECC84 0.60ECC85 0.75ECC88 0.80ECC189 0.95ECC804 0.65ECF80 0.95ECF82 0.95ECF801 0.95ECH42 1.20ECH81 0.70ECH84 0.80ECL80 0.65ECL82 0.75ECL85 0.75ECL86 0.90EF9 3.50EF22 3.90EF37A 2.15EF39 1.10

i ,i, 0.65L' 83 3.90EF 85 0.60E F86 1.25E F 89 1.60EF91 1.60EF92 2.15EF95 0.95EF96 0.60E F 183 0.75EF 184 0.75EF812 0.75EFL200 1.85EH90 0.85EL32 0.85EL34 2.10EL34' 4.55EL82 0.70EL84 0.95E L86 0.95EL90 1.75EL91 6.50EL95 1.25EL504 2.70E L 509 5.85EL519 7.70EL821 8.45EL822 9.95E LL8OSE 4.50EM80 0.80EM87 2.50EY51 0.90EY81 0.75EY86/87 0.60EY88 0.65EZ80 0.70EZ81 0.70GM4 8.90GY501 1.30GZ32 1.40GZ33 4.20GZ34 1.400134" 3.800237 3.956166* 15.50KT77- 14.00KT88 17.006188" 25.00ML4 3.20ML6 3.20MX12001 29.50N78 9.900A2 0.700B2 0.80

PCL82 0.95PCL84 0.85

PCL86 0.80PCL80585 0.95P0500510 4.30PF L200 1.10PFL200' 2.80PL36 1.10PL81 0.85PL82 0.70PL83 0.60PL84 0.90PL504 1.15PL508 2.00PL509 5.65PL519 5.85PL 802SE 3.45PY80 0.70P081800 0.85PY82 0.75PY88 0.60PY500A 2.1000V0310 5.9500V0310' 7.5000003 204 27.5000V06 40A 28.5002006,404 49.50OV03,12 5.75SP61 1.801121 37.50T122 37.50UABC80 0.75

0.70U8F89 0.70UCC84 0.85UCC85 0.70UCH42 2.50UCH81 0.75UCL82 1.60UF 41 1.35UF80 1.60UF85 1.20UL84 0.95UM80 0.90UM84 0.70U082 0.70U085 0.85VR10530 1.45VR150,30 1.80X61M 1.70X66 1.802749 0.752759 19.00Z8000 3.45

28010 3.7528030 18.5(Z900T 2.451A3 2.751L4 0.65IRS 0.801S4 0.65I S5 0.75114 0.751U4 0.802X2A 3.803A4 0.703Al2 3.403B28 12.003828' 19.50306 0.603E29 21.853S4 0.704832 18.25SR4GY 3.355U4G 1.855V4G 0.75SY3GT 0.95513 2.805Z4G 1.255Z4GT 1.156 30LS2 0.806AB7 0.706AC7 1.156AG5 0.606AK5 0.956AK6 2.856AL5 0.606AL5W 0.856AM5 6.506AM6 1.606AN8A 2.506A05 1.756A05W 2.306AS6 1.156AS7G 4.956AU6 0.908A 840 i 1.306AX5GT 110684G 1206BA6 0456BA6' 1.806BE6 0.65613E6' 1.8568060 1.6068J6 L.,6007A 0.856BR7 4.806BW6 6.10

6BW76C46CH66CL66CW46CX86C Y56D66F6G6F6GB6F76F 8G6F126F146F156F176F236F246F336FH86GA86GH8A6H66J46J4WA6356.1501636636W6JE6C6JS6C6JU666761066L66L6GC6L6GT6L186L0206L06607G6SA7'6SG76SJ76S676SL7GT6SN7GT69076SR76V6G60601604605016Y6G6Z4

1.651.107.902.757.404.601.152.501.951.102.800.851.50.

;91.303.100.651.15

10.5018.80

1.950.901.601.951102.300.900.652.806.606.906.351.457.404.606.251.950.700206.601.301001801.801.350.851.500.954.60ISO1.401.500.752.801.30

7Z491)611E212A612AT612AT712AU7I 2AX7138A6120E6133H7taEt12J5GT12K7GT12K8G11207GT12SC71 2SH7125.171256712S07G112Y413D313D619A051963190619'15200120E120P1252601

85A2

85A2'807807812A813813'82988203'E2,66Aooc.,931A931A'954,"'9565763

,060,1,96-136614688068900190)290)3

1.902.15

19.501.000.750.950.950.751.401.90285

19.950.551.151.250.750.801.251401.452.200.702.800.901.35

11.5010.353000

0.801.300.601.6025240

0.751A02.552203A0

315028.5044.0016.0024.00

7.704.95

13.9519.80

1.101.101.205.75195

2.807.30

10.3512.500.950.950.95

VALVES AND TRANSISTORS NEW PYE EQUIPMENTTel' ','s etc & SPARESRehr . . , . . , 0899

HARNESS A' 8 B CONTROL UNITSFIELD TELEPHONE. CABLE TYPE 010 A R Jr J2FIELD TELEPHONES TYPE V'. Tropical in metal cases10 -line MAGNETO SWITCH -BOARD. Can work with U .. t' :

,, 0

every type of magneto Telephones Larnet sets eit,

POSTAGE AND PACKING CHARGES: El -E3 50p 0-05 60p. 65-610 80p 110-615 (I 00 615-12C 11 50Over C2012 00 but below 2kg Parcels over 2kg at Cost

COLOMOR (ELECTRONICS LTD.) 170 Goldhawk Rd, London W12Tel: 01-743 0899 or 01.749 3934. Open Monday to Friday 9 a.m. -5.30 p.m.


IEEE488Technology that hits the mark

For IBM -PC/XT/AT/IC/RT6150and all other compatiblecomputers

For PHILIPS PC :YES HP commands (enter, clear etc.)

implemented SRQ/ASYST compatible 64 kByte memory capacity DMA and INTERRUPT can be

activated by simple commands

DEALER + OEM WELCOME

HELP functions, SYNTAXmonitoring in clear text

BASIC, BASIC(compiled(.ITURBO-(PASCAL MODULO -2,FORTRAN, C, ASSEMBLER

DOCUMENTATION -ANALYSIS -software for IEEE -488 sys- emse.g.transient-recorder,PHILIPS -memory oscillosccpe

Ines GmbHNotiennefor Maw455000561n 41West Germane

Mon. 49 221 4391 59T«a...2827.22r 42 37 gHonSAX 49-221 491901

ENTI.:ft 510N 1{1.:19.1

11,11Mrii111:1:01=:1:=2:Z114.1:1121.1=erell:144:i1MIE

ELECTRONICS& WIRELESS WORLD

OCTOBER ISSUEThe October issue of Electronics and Wireless World,published 16 September, reviewsR.F. SIGNAL GENERATORS.The modern signal generator is a far cry from those in useeven ten years ago. This in-depth feature examines newdesign techniques, facilities available and the characteristicsof a wide range of instruments from a few hundred kilohertz tothe gigahertz end of the r.f. spectrum.TO ADVERTISE WITHIN THIS FEATURE CONTACT:MARTIN PERRY on 01-661 3130 NOW'

TEST METERS

III I's 111 NI \ 1 l IR Range 0.2Hz to2M11/. Stile. trimgli.. rectangular, pulse, ramp, saw -tooth, and slewed sine wave can he generated. Variable dutycycle, push button range and function selection switches.

111.111/ t!s:( )1'N FER I ionzto 00.1111/ with 8 -digit I El) display, frequency counter, periodcounter with a range dow i to 400ns, and a totalise eventcounter mode. 1CMIIi.c-vstal oven time base.

sr ,444M014,444.1.

all,f130.39

VA 5Y

ll,Ili\Il 1.1 'ITN. 1 )N q N Li( 1 AILR 'autowith 8 -digit LED display, freq. counter. period counter

with a range down to 400r s, and a totalise e, ent counter mode.10MI I/ and 3.90625MIL crystal oven time base.

1111111111

to.

1)10 /MAIM' )R 1111,2 tar 'ookH1. Low distortionsine & square wave outpu s. Synchronisation input. Output freq-uency displayed on 4-digi LED display. Output 2.5V r.m.s. into600It with continuous ark fixed -step output attenuator.

IZZ0000. Ya.00,

N

e e

A -1)11.1 I DI \ II I I IME 1 Reads IX:voltsfrom 10i.tV to 100(IV. . I OµV to 750V, DC and ACcurrent 10nA to I u:\ and r2sistance 10mi l to 20M11. Basic DCaccuracy ±0.05".. 2 dig is ). Continuity and diode test ranges.Similar specificatit 'II I 00aA min. current hand-held type alsoavailable YJ8 IC 0 1 .1. 1:60.82.

MP S MA.PLIN PROFESSIONAL SUPPLIESP.C. Box 777, Rayleigh, Essex. SS6 XIX.

Telephone 0702 552961. Telex 995695. Fax 0702 553935.

tr In meters . . we're miles ahead.GO LDAll p se' art net trade 111,15 HI 11 .5 I titter Niken ere aCallable Tor 404111111es T1

T 1-1.41. per Item All tn..ter 240% 111.1111S Hixadlcsi m an raL IT, c grey re 1th tillden. T....LT:pi 1'1611.



781

Variable -speed video-the ultimate

This report on new developments in variable -speedplayback for C -format professional video tape recorders

up -dates our seven -part review.

In the C -format, one diagonal tape trackcontains one video field, and the tape iswrapped almost completely around the

drum so that a single head can be in almostcontinous contact with the tape. The inter-ruption, known as the format dropout, istimed to coincide with the vertical intervalwhere there is no picture information. Theadvantage of field -per -scan operation is thatvariable -speed operation is eased when com-pared with formats which segment the fieldinto several tracks. In variable speed, thedrum speed remains the same. locked tofield rate, but the linear tape speed ischanged by driving the capstan asynchro-nously. A track -following head in the rotat-ing drum deflects to follow tape tracks, andjumps during the format dropout to omittracks at higher than normal speed, or torepeat them at less than normal speed. As aresult, the field rate remains the samewhatever the tape speed over the speed rangeof the track -following system, usually -1 to+3x normal speed because of geometricrestrictions.

The physical length of the track is deter-mined by the field period, the signal band-width and by the minimum wavelengthwhich can be recorded reliably. In theC -format, this results in tracks which areabout 15 inches long, and a rather largedrum. Timebase stability of helicalscan re-corders can never be as good as transversescan machines because the long length offlexible tape around the drum is excitedalong its length by head rotation and tensionvariation.

The timebase stability needed in a profes-sional video signal is a few degrees ofsubcarrier phase, so that composite signalsfrom various sources can be mixed. Sinceone cycle of subcarrier represents 225 ns.the time stability is of the order of ananosecond. Since the field period in PAL is20ms, this represents a proportional accura-cy along the tape track of one part in 20million, which is asking rather a lot of aflexible medium.

It is not an exaggeration to assert thatwithout the digital timebase corrector, theC -format would not be a viable medium. The

JOHN WATKINSON

Sampling clock phasePhase error at line end

-7 'velocity error'

One tv line

Fig.1. Conventional timebase corrector updates the phase of the sampling clock ateach burst, but in the presence of velocityerrors this results in a phase error at theend of the line.

timebase errors introduced by the transportfall into various caterories, all of which thet.b.c. has to correct. The drum constructionis such that the upper half is free to rotatewhereas the lower half is fixed. Tape entersthe drum at the lower elevation, and risesaround it. As a result. the tension does notchange uniformly around the scan, as theforce needed to pull the tape over the fixedlower drum is greater than that needed topull it round the air film boundary on therotating upper drum. Fortunately. the sameeffect takes place on recording and playback.and largely cancels out. The effect of tapetension is critical, as a minute percentageelongation can result in an enormouschange in field period when it is measured incycles of subcarrier. Dimensional changesdue to humidity variation have similareffects to tension changes. These effectschange relatively slowly. and timebase cor-rection can be assumed to be the same overthe duration of a video line.

There are also dynamic phenomena whichare more difficult to deal with. Head impactand the release of tension as a dragging headlifts off the tape, known as expact, causeshock waves which propagate down the tapecausing the instantaneous head to tapespeed to rise and fall slightly. The shockwaves will be at the mechanical resonantfrequency of the tape's mass with its ownelasticity, typically a few kilohertz. Thedynamically varying subcarrier phasechanges within the line period, and is much

harder to correct. Unfortunately the term'velocity error' has been given to this kind ofinstability, possibly to distract attentionfrom the fact that its origins are in accelara-(ions. When the machine is in variable -speedmode, tape acceleration as the speed is

changed will also cause significant subcar-rier phase changes throughout a line. Suchinstabilities will also be found in recordingsmade on portable machines that have beencarried while in operation.

Fig.1 shows the subcarrier phase chang-ing as the result of a head impact. There is nosubcarrier in the chroma signal carried inthe composite video because it is suppressed,and the only information that can be used todetermine subcarrier phase, and hence themeaning of the chroma. is in the burst. Asthe burst only occurs once per line, thesubcarrier phase transmission is a sampledsystem, and according to Nyquist, the fre-quency with which the phase changes can-not then exceed half the line rate.

Fig.2 illustrates the digital timebase cor-rectors used with C -format machines. wherethe incoming video waveform is sampled atthree or four times off tape subcarrierfrequency. obtained from the burst, andquantized into eight or nine bits, the lastallowing for more processing in the digitaldomain. Samples are stored in memory andare read out according to a stable samplingclock derived from reference timing. Drop-out compensation and colour processing arealso necessary.

Returning to Fig.1 you will seen that ifthere are velocity errors, this apparentlysimple process becomes difficult. The onlyphase information which can be used forsampling comes from the burst. and thesampling phase then remains constant untilit can be up -dated by the next burst. Inpractice, the subcarrier phase changes con-tinously, and there will be a chroma phaseerror which increases towards the end of theline. The conventional solution to this prob-lem is to measure the phase change betweensuccessive bursts, and thus establish whatsubcarrier phase was at both ends of the line.In this way the phase of the memory readclock can be dynamically shifted as the line


proceeds in order to oppose the phase errorbetween the sampling clock and the chroma.

Fig.3 shows how this conventional form ofvelocity compensation is performed. butilustrates that it is quite unsuitable wheredigital processing of the stored samples iscontemplated, because the phase of thesamples with respect to subcarrier is con-stantly changing.

In composite video signals such as PAL.there is an endless structure where therelationship of the chroma to the line timingonly repeats at eight field intervals, tointerleave the sidebands of the chorma intospaces between the luminance sidebands.When a tape is played at variable speed, theresulting head jumps cause the eight fieldsequence to be broken. The sequence mustbe restored on the video output of thetimebase corrector so that it can be mixedwith other signals. Traditionally this hasbeen done by analogue circuitry after the d -aconvertor, or before the a -d convertor.which decoded the composite video to base -band according to the burst phase offtape.and encoded according to reference burstphase. This circuitry is generally referred toas the colour processor.

An improvement in the performance andstability of the colour processor could beobtained by using digital techniques, butthis would require some means of aligningthe phase of samples with offtape subcarrierin the presence of velocity errors. It would bepossible to construct a digital interpolatorwhich could compute the values of thewaveform at arbitrary points between sam-ples. to produce a sample stream with aconstant phase relationship to subcarrier.but this approach is very difficut to imple-ment because of the phase sensitivity of thesignal. When working to a few degrees ofsubcarrier, the number of interpolaterphases necessary would be extremely large.and the coefficient store would be enor-mous. The wordlength of the samples wouldalso need to be increased to hold downchroma noise.

In the Ampex Zeus video processor, digitalcolour processing is used, requiring a stablephase relationship between sample pointsand offtape subcarrier. This is ensured by

,controlling the phase of ;he sampling clock.If the sampling clock phase is to be made

to track the offtape subcarrier phase instabi-lities. it is necessary to use information fromthe burst before and after the line of interest.As the burst at the end of the current line willnot be avaiabele until after the line haspassed, a one -line delay is necessary. In theZeus Processor, a c.c.d., delay is used.running at 27MHz. The bandwidth and noiseperformance of this delay are engineered tobe in excess of the constraints set by thesampling and quantising parameters of themachine. so it is essentially transparent.

As Fig.4 shows, the delay is actuallyslightly longer than one line. and an ana-logue switch allows the delay to be bypassedbriefly, so that part of the sync. pulse isreplaced by the burst from the end of the linewhich has been advanced by omitting thedelay. This signal is fed to the a -d convertor.which makes it available to the memory andto the sampling clock control circuits.

Off tape

video

A -to -dconvertor Memory

Off tape timinggeneric.

0-to-a Referenceconvertor I ocked video

Reference timing I

genercor

Fig.2. Simple timebase corrector temporarily stores samples which are produced fromunstable video and reads then out with a stable reference -- derived clock, Dropoutcompensation and variable speed circuitry are not shown.

Acceleratingwaveform

Steady sample clock

Deceleratingread clock

Stable waveform

Half cyclesof reference

Fig.3. In conventional velocity compensation. the real clock is phase swung to correct thephase error on sampling. This works well for an a ialogue output but the phase betweensamples and chroma becomes arbitary.

The effect of sampling the burst with afour -times subcarrier clock is to produce arepeating set whichreveal the phase relationship between theclock and the burst in rectangular co-ordinates. In practice there will be noise onthe burst. but by adding together the sets ofsample values from several cycles, and divid-ing by the number of cyles acquired. anoise -free determination of the co-ordinatescan be made. The actual phase relationshipcan be expressed in angular terms by sometrigonometrical calculations. By successive-ly processing the two bursts coming fromthe analogue switch, it is possible to com-pute the current phase error in the samplingclock - i.e. the amount by which the clockphase needs to be changed to correctlysample the beginning of the line. and tocompute what the phase error will be whenthe end of the line is reached. To get intolock, it is necessary to know the approximateclock frequency needed. This is obtained bytiming the period between h -sync. pulses.The sampling clock is obtained from afrequency synthesizer, which has no drift ortolerance rnechansisms and thus needs noinitial or periodic adjustments. The equiva-lent of the v.c.o. control voltage is the digitalburst phase errors.

There is very little time between the end ofthe second burst and the start of the activeline. and so the digital burst phase filters andphase measuring circuitry must work veryfast indeed.

When the speed of the tape is varied. theline period also varies. and the length of thedelay must be adjusted to suit. This isachieved automatically by clocking the c.c.d.delay from the digital synthesizer. If the tape

Input cMore than 1 lineTo a - to -d

conv

Switch

IANext line.

Current line

Line end burst Line begin burst( , -

DelaySvil* No delay

Fig.4. Using a delay of a little over one line itis pcssible to switch the delay out momen-tarily so that the current and next burstsappear side by side. Both are digitized tocompute the velocity error over the line.

speed were to increase. the first sign of thiswould be that the advanced lundelayed)burg would have a phase lead relative to thesampling clock, whereas the delayed burstwould not. This would cause the synthesizerto speed up the clock to cancel the phase leadover the line. The raised clock frequencywould then make the delay period correct forthe next line and so on. In this way thesampling clock tracks offtape subcarrier


over the speed range where colour is pro-vided. As with all phase -locked loops, thedamping has to be critically set to the bestcompromise between jitter and speed ofresponse. In a sampled system, the responsecannot exceed half the line rate, and so thesynthesizer is damped so that it cannotproduce frequency changes more rapidlythan that during sampling of the active line.

An advantage of this method of clockgeneration is that the a -d convertor iscontained within the loop because it is thedigital samples which are used to control theclock. Phase errors due to drift in theconvertor or component tolerances are thuseliminated, along with correspondingadjustments.

When memory chips were of relativelysmall capacity, it was convenient to assign agiven set of chips to the storage of one tvline. Now that chip capacities have in-creased, it is no longer practicable to do that.In modern t.h.cs there is just a large ram andthe individual lines can only be discrimin-ated by their different addresses. This meansthat the memory must be able to read andwrite at the same time. There is a furtherproblem that the high rate of digital video isbeyond the access rate of economic ramchips. The solution to both problems lies ininterleaving the memory.

A number of samples is assembled into asuperword by a serial to parallel register. Thesuperword rate will now be the sampling ratedivided by the number of samples in thesuperword. The superword period is nowdivided into a period when the memory canread, and a period when it can write. Theswitching between read and write functionsinvolves both address and data multiplexers,and arbitration. In general, the memorytiming will be locked to stable referenceclocks, and when the memory is being read,the write process is locked out. A smallbuffer silo is installed before the memorywhich allows the write process to be inter-rupted by reading, and accommodates thedifference between stable read clocks andunstable write clocks.

For variable operation, the write clockrate will vary. At lower than normal speedsthe data rate falls, so that not every writeperiod will be used. Clearly at higher thannormal speeds this approach cannot be used.The solution is then to subdivide the super -word period into three time slots, one forreading and two for writing if necessary.Thus at low rates, the silo may not have asuperword ready when a write period ocurs,whereas at high rates it may have twosuperwords ready. Fig.5 shows the generalarrangement of an arbitrating ram andshows the input silo. In some machines,such as the Ampex TBC-6, the silo performstimebase correction to the nearest super -word, whereas in the Zeus, the silo has a fourline capacity and corrects to the nearest lineleaving the main memory to correct timingin integer line steps.

The digital colour processor of Zeus willnow be described, beginning with the func-tions it has to perform, which result fromthe structure of the PAL signal. The eightfield sequence comes about as a result of thequarter cycle offset of subcarrier against line

Reference clocks

Memoryarbitrator

Asynchronoussamples in

Serialto

parallel

Writerequest

Raj

Read/writecontrol

Silo output ready

Inputsilo

Superwords

Fig.5. In a ram -based timebase corrector ram is reference synchronous. and an arbitratordecides when it will read and when it will write. During reading, asynchronous input databacks up in the input silo, asserting a write request to the arbitrator. Arbitrator will thencause a write cycle between read cycles.

Read/wrsuperword

-`erence lockedsamples out

rate. There are 2833/4 cycles of subcarrier inone line so that four lines must pass beforethe same relationship occurs between sub -carrier and h. The odd number of lines in aframe necessary for interlace means thatfour frames must elapse before the four -linesequence assumes the same relationshipwith vertical sync. It follows that the task ofthe colour processor is to take one of the fourline types from the input field and make itinto the necessary line type for fieldbeing output. The presence of the V switchon alternate lines which gave PAL its namecomplicates this process.

Fig.6(a) shows that from line to line, thephase of U advances 90 degrees due to thequarter cycle offset, and the phase of alter-nately leads and lags U. Fig. 6(b) shows thatif chroma is sampled on the U and V axes theresult of these two effects is that from oneline to the next, the chroma samples repre-sent different sequences of U and V. The shiftto the right of the U sample due to thequarter cycle offset can be seen. It is notimmediately obvious how one line of chromacould be made into another with thisapproach.

The solution adopted in the Zeus is tosample at 45° to the U and V axes, i.e. inphase with the burst. Fig.6c shows thatwhen this is done, the samples representU cos45 + Vcos45 etc. If everything is

normalized by dividing by cos45, the sam-ples become U+V, - U+V etc.

If it is required to convert a line in to thenext line type, having the opposite sense of Vswitch, this can be done by inverting thesamples where U and V have different polar-ities, but not where they have the samepolarity.

If the difference is two lines, the onlyrequirment is to invert the whole chromawaveform, as V switch is, the same after twolines, and there have been two quarter -cycleoffsets of subcarrier making 180°. If thedifference is three lines, an inversion com-bined with V -switch reversal is needed, andthis can simply be done by inverting samples

where U and V have the same polarity andnot where the polarity is different.

The line type is entered into the memoryalong with the samples for the line, and isavailable when the line is read out. Therequired line type is available by decodingthe ScH phase of the reference sync. input,and by comparing the two-line types, theconversion to be performed by the colourprocessor can be determined. The chromamust be separated from the luminancesignal before the colour processor can oper-ate, and this is done by a digital filter. Acombination of chroma band pass and combfiltering is used. The comb filter gives goodseparation of sidebands, but fails when thereare vertical chroma phase changes in thepicture. The chroma phases at the beginningcentre and the end of the delays are moni-tored by a corellator, and when the combfilter fails to mesh, the bandpass signal aloneis used. The filtered chroma signal is thensubtracted from the composite signal to givedigital luminance. Once chroma and lumi-nance are separated, a digital chroma gaincontrol can be incorporated simply by pas-sing the chroma samples through a multi-plier.

Owing to the use of interlace, the headjumps may result in an odd field being playedwhen an even field is needed, and vice versa.The vertical picture movement which re-sults from outputting the wrong field can becompletely removed by vertical interpola-tion.

The output raster and the input raster areshifted slightly so that neither input fieldaligns with either output field. By usingmemory to produce three delays of one lineeach, samples representing four -points in avertical line on the input field are simul-taneously available to a four -point f.i.r.filter. If the output field is to be the same asthe input field, the phase of the interpolatorwill be such that the sample value for a line1/4 of the way down from an input line will becomputed. If the opposite type of field isrequired, the order of the coefficients to the


56n sec

Line n

Line n. 1

Line n2

Linen. 3 "N.V

1 cycle of subcarrier

-v

v\U

Fig.7. In vertical interpolation a f.i.r. filterwill take an odd or even input field andinterpolate an odd or even output fieldfrom it by choosing coefficients to obtainthe correct impulse response.

f.i.r. filter will be reversed such that a line 3/4of the way down from an input line will becomputed. Fig.7 illustrates how this processresults in the image staying at the samevertical position on the screen irrespective ofthe relation of input and output fields.

The luminance conveys the subjectiveresolution, and so the vertical chroma inter-polation can be done with a simpler filterhaving only one line of delay to give twopoints.

Dropout compensation in v.t.rs is bysubstituting samples from nearby in thepicture. The v.t.r. sends a logic level signal tothe t.b.c. when the replay r.f. level falls dueto a dropout. When samples are stored in thememory, one of the 512 codes is used tospecify the presence of a dropout so this canbe detected when the memory is read. Theimproved type of drop -out compensationused in Zeus uses delay lines to give simul-taneous access to the previous line and thenext line. Where a dropout code is detectedin sample values, samples from the previousand next lines are added and divided by twoto give an interpolated pixel value. A correla-tor watches for the condition where dropoutoccurs in the same place on two of the lines,since in this case interpolation is impossible,and samples from the remaining line areused instead. As the chroma phase andV -switch sense are different on successivelines, processing is necessary to obtain thecorrect type of chroma to substitute for thedropout.

V.1) -VU -V-U V -U

Line n

Line n.1

Line n.1fromline n

Invert Invert

Fig. 6(a) above left. The combined effects of V -switch andquarter -cycle offset in PAL cause the above relative phases of Uand V from line to line, in a four -line sequence which only repeatsafter four frames.Fig.6(b) left. If a sampling system tapes samples on the U and Vaxes it -esults in the samples having the values shown here. It isthen not possible to conver from one line type to the next because.for example. V cannot be processed to become U and so on.

Fig.6(c) above. If sampling is performed with sample phasebetween U and V axes (thin arrow) the samples represent sums anddiffererces of the two signals (cosine term neglected). Line n canbe be made into line n 1 by inverting the samples when U and Vhave the same sign. When the sample is made, phase information islost anc only amplik de is known. For this reason the non -invertedsamples phase mirror (broken arrow) because the phase isdetermlned in the context cf adjacent samples.

IEven in, even outx. coefficientsOdd in, even outo.coefficients

Even in, odd outx=coefficlentsOdd in, odd ouro.coetficients

E,en input

field tines

Odd inputfield lines

Even outputfield line

Odd output fieldline

Once the dropout compensation, colourprocessing and vertical interpolation arecompleted, the digital chroma and lumi-nance can be recombined in an adder, andthe functions of the processing amplifier canbe undertaken. These include black clipping.to prevent luminance going blacker thanblack level, black level and pedestal adjust-ment. the re-insertion of syncs and burst atthe desired ScH phase. and a chroma phasecontrol. Traditionally, all of these processeshave been carried out in the analog domainafter the d -a convertor, but in Zeus. these are

all carried out in the digital domain prior tothe d -a, with resultant stability and freedomfrom periodic adjustment. A further advan-tage of the all -digital approach is that allparameters are set in output ports of acontrol microprocessor. and this makesremote control very easy. The t.b.c. iscontrolled from the control panel of thev.t.r. by a serial link to the microprocessor.

John Watkinson. M.Sc. is senior technicalsupport engineer and educational consul-tant with Ampex Great Britain Ltd.


The etherI believe the concept of an etheris a necessary construct to enablereasoning about reality with aphysical mechanism. It iserroneous philosophizing to saythat the slick mathematics ofasymptotic equilibrium phe-nomena does not need it. The'iterative' causal process of anether mechanism is the in-termediate 'scaffolding' that de-fines and converges to theasymptotic limit - but seeminglyvanishes in importance (like allconstructional scaffolding) inthe final oversimplified formula.But it is an essential explanatoryintermediate.

The ether model I envisage, ofa virtual electron -positron sea, isnot the same as Maxwell's ether.In modern understanding, mat-ter and energy are interconverti-ble. Extrapolating that one stagefurther, I would say that matterand energy are one. Matter isenergy trapped in the ether, andnot Maxwell's ether somehowpassing through the intersticesof matter.

So, in ether -drift experiments,it is not the ether that passesthrough matter, but the energyof matter that has to propagatethrough the ether the same asany non -localized light rays inthe apparatus. Hence the nullresults of all the misconceivedexperiments.P. J. RatcliffeStevenageHertfordshire

M. G. Wellard's letter in the Mayedition of EWW rationally andconvincingly argues the case forusing a model based on 'etherwind' to help explain and under-stand magnetic phenomena. Allphysics relies on using modelswhich reflect the 'perceivedphysical reality' to a degree de-pendent on the current 'range ofperception' or in other words,the state of measurement tech-nology. It may or may not bepossible in the future to in someway perceive an ether, and untilsuch time it could be that theether does not 'exist'. It couldequally well be said - leavingaside current controversy - thatrelativity effects did not exist inNewton's time; his laws of mo-tion described the perceivedphysical reality to an accuracy farbeyond the range of perception

FEEDBACKat that time. The 'existence' ornot of an ether, while perhaps ofphilsophical interest, is hardlyrelevant to its scientific applica-tion (show me a complex num-ber) and only its usefulness inunderstanding and predictingphysical phenomena will (orshould) determine its eventualacceptance or rejection by thescientific and technical world.

As an afterthought, it wouldseem reasonable to suppose that,if an ether model proves success-ful in explaining 'instantaneousaction at a distance' for magnets,then possibly a second ether tomodel electrostatic force may beappropriate. Naturally, gravita-tional force must have its ownether and of course time is mere-ly another type of 'ether wind'.Electromagnetic radiation isthen an interaction between themagnetic, electrostatic and tem-poral ethers and mass, (which isof course equivalent to energy aslong as you're not buying apound of butter) is obviously tiedup with temporal and gravita-tional ethers.

Not quite fire, earth, air andwater but maybe Eddie is in thespace-time continuum after all.Keith WoottenReadingBuckinghamshire

My letter at page 44 of theNovember 1986 issue of EWWdrew attention to the initial re-port in Nature of the experimentperformed by E. W. Silvertoothwith US Air Force sponsorship,which had detected our motionrelative to the aether. The experi-ment measured variation ofstanding -wave node displace-ment with orientation of a laserlight beam comprising light raystravelling in opposite directionsand supplied from a commonlaser source. A speed of some 378km/s relative to the aether hadbeen detected.

Arising from my letter, severalof your readers have writtenasking me for more information.Accordingly, I offer the followingcomments.

An account of the details of theSilvertooth experiment hasalready been reported by the edi-tor of a book, just published,entitled: "Progress in Space-timePhysics 1987". This is an editedcollection of 28 papers on cur-rent research in the anti -relativist field. The book isobtainable from the publisher,

Benjamin Wesley, Weiherdamm-strasse 24, 7712 Blumberg, WestGermany.

Of special interest, however, isthe fact that the Silvertooth ex-periment has already been repe-ated in Austria by S. Marinov,using less sophisticated detectorequipment and Silvertooth'sfindings are fully confirmed. TheAustrian tests gave the speedthrough the aether on 4 January1987 as 386 ±38 km/s in thedirection of right ascension12.5 ± 0.5 h and declination-22°±6°. Marinov's paper on thisis one of the 28 in the book justmentioned.

Also, though Silvertooth hasnow built and demonstrated animproved test apparatus, whichconfirms his earlier findings, areport of his first experiment hasjust been published in his paper"Experimental Detection of theEther" (Speculations in Scienceand Technology, volume 1Q,number 3, 1987).

The sad situation, however,which echoes the views of DrEssen (Wireless World, page 44,October 1978), is that these ex-periments are being performedby 'anti -relativists' and are,therefore, being ignored by thepro -relativity establishment.

It may be, therefore, that wemust await a practical applica-tion of the Silvertooth discovery,perhaps as an aid to space naviga-tion, before the situation reallyclarifies. Meanwhile, your read-ers may also like to hear that oneto those writing to me, Dr J. A.Briscoe, tells me that he prop-osed a standing -wave detectorfor measuring our motionthrough space some 29 yearsago. He even obtained a BritishPatent for it in 1958 (No.884,830) and, in his paper in theWesley book, he writes "my ideahas now been substantiated bySilvertooth . . ." Hopefully,there will soon be more scope forinventions as we come to realizethe potential of exploiting whatthe aether has to offer as relativ-ity slips aside. Remember the lastwords of Essen's Wireless Worldarticle:

"space contains an unlimitedamount of high frequency energywhich could possibly be extractedand used with safety and effi-ciency".

H. Aspden,Department of ElectricalEngineeringUniversity of Southampton

Michelson-MorleyIn your December 1986 issue youhave an article and two letters onthe Special Theory of Relativity.Einstein deduced in 1905 fromthe results of the Michelson-Morley experiment that therewas present a Lorentz -invariantproperty. At that time little wasknown about quantum mecha-nics, and he concluded that theLorentz -invariant propertyappeared to be in the coordinatesystem, which made it incom-patible with the Newtonian coor-dinate system.

The Lorentz -invariant proper-ty follows from Heisenberg'sprinciple and is a property pos-sessed by some of the Heisenbergeigenstates. Such eigenstatescan exist within the Newtoniancoordinate system.

The reflection of light involvesthe absorption and re -radiationof a photon by an electron, andHeisenberg's principle statesthat in the interaction between aphoton and an electron, if thereis a change in the values of twoparameters the dimensions ofwhose product are those of ac-tion, then the changes in theparameters can only be deducedto within a quantum of action.Two such parameters are dis-tance and translational momen-tum, the last of which can beregarded as proportional to thetranslational velocity. Thus if thepositions of two eigenstates aredetermined, then nothing can bedetermined about translationalvelocities associated with the twoeigenstates, and this is theLorentz -invariant property.

In the Michelson-Morley ex-periment, by positioning half -silvered mirrors and the plates ofan interferometer it is attemptedto establish the positions ofHeisenberg eigenstates, so thatcomparisons can be made of thetimes that photons take totraverse a particular path. It fol-lows that nothing can be deter-mined about translational velo-cities associated with theseeigenstates, and this is theLorentz -invariant property.

Einstein deduced that therewas present a Lorentz -invariantproperty, and the form of hisformula is correct, but it repre-sents a relation obeyed by someof the Heinsenberg eigenstateswithin the Newtonian coordinatesystem.R. Fricker, Surbiton


Flow diagrams -anewapproach

Novel approach produces clear and easy to maintain flow diagramswell suited to both high-level and assembly languages.

DAVID J. SWEENEY

hen I was first introduced to thesubject of computers and comput-ing, the accepted means of planning

and designing a program was to draw a flowdiagram. As anyone who has ever performedsuch a task knows, it is far from simple.

By analogy, if you were to read a book onthe subject of car maintenance, inevitablyyou would come across a description of howto remove or disassemble some complexitem (a carburettor say). In the book theproblem was disposed of in a sentence. Inreality, the job took an hour of valuable time.

The difficulties of using conventional flowdiagrams show themselves most vividlywhen you try to alter them. The boxes arenever big enough to add those few extrawords; you have to rub out and redraw boxesand text just to insert a box with i= i+ 1 in it:and you have to draw lines intersectingothers and make the diagram unreadable.

So who in their right mind would use aflow diagram? Well. I am about to suggestthat you should. A picture is worth athousand words, and a flow diagram is - orshould be -a clear picture of the programyou intend to write.

By applying the kind of reasoning used inthe evolution of the circuit diagram to theflow diagram. I can now claim to have a flowdiagram which is easy to maintain. clear inits presentation of the facts and flexibleenough to cater for high-level languages andassembly -language code.

BASIC PRINCIPLES

Using my method. flow of control passesalong lines and proceeds down the page.Conventional start -of -program and end -of -program boxes are used. Fig.1. Actions aresignified by a small black circle on a line. thedescription of the action being given to theleft of the flow line.

Actions only take place on vertical lines.When a decision has to be introduced. theflow line branches to the right. The branchmay be caused by either a true or false

Fig.2. Three loops. Outermost is a 'for'loop in which counter I is indexed. Next isa 'repeat -until' loop where the exit condi-tion appears at its end. The innermostloop fordo -while' continues for as long ascondition X>50 is true.

A

B

X (

Y

PLOTVAL

A2.9 (BREAK)

P=0 (CONTINUE)

Fig.3. Subroutines are represented thus.Two parameters, A and B, are passed intothe subroutine and one value. Q. is re-turned from it.

PROGRAM

1=0

B,27?

C=1.8/5

X,./11417

X

Calculate Z -FACTOR

I< 50

Calculate Y-VAL

condition, this being signified by a small T orF in the box as shown. Where secondary flowrejoins the main flow an arrowhead is used.Note that it would be wrong to draw thesecond decision rejoining the main flow inFig.] . The correct method is as shown.

Subroutine calls are represented by arectangle. The flow line is drawn throughthe rectangle, and examination of Figs 1-4illustrates one of the major features of thismethod. All of the symbols are chosen sothat if you need to insert a symbol, no linesneed to be erased!

Figure 2 shows how loops are drawn. andthe clarity with which the nesting of loopscan be seen. I view this as a valuable featuresince it allows immediate identification ofsuch transgressions as jumping out of. orinto, a loop. Two symbols not shown on thediagram are the symbol for an early exit fromthe loop. called BREAK in the C programminglanguage, and the symbol for returning tothe top of the loop for the next iteration.called cosriNi E in C. These symbols areshown in Fig.3.

Parameters for subroutines are shown asan equals sign when they are being passedinto a subroutine: inside the subroutine areshown as a greater -than symbol. Bracketsaround the parameter symbols signify thatthe parameter is being passed by reference

Fig.1. Twodecisions areshown; the firstbranch will betaken if B> 27 istrue and thesecond if 1< 50 isfalse, as indicatedby letters T and Fwithin therectangles.


A=B4112

Z=20

A=A/9

Z> 50

C=2/7

Fig.4. Control flow passes to the statementA A/9 if Z is not equal to 20. After thiscontrol passes down the escape route.Fig.5. Medes' line and circle drawingalgorithms (bottom) redrawn to illustratethe new approach to flow diagrams(below).

CIRCGEN

X.=0Y tRA -2

B =1-2*R

X =X1A =A45 = BA

13,.0?

LINEGEN

For a straight tine from 10,01 to lx 1st, y1s ) ty

with 0<y1st Gy 1st

X..- 0

A ...2*y1st8 - A - xistC 8 - xist

PUT PIXELIX.T;

True

X 4.

Y + 1+ C

B

rather than by value.In Fig.3, value P has been returned by the

routine 1.1.01-vm.: value Q is being returned bythe subroutine which comprises the wholeof Fig.3.

ESCAPE ROUTES

Figure 1 shows how decisions are nested.With conventional flow diagrams, a difficul-ty arises when this nesting structure has tobe translated into a high or low-level lan-guage. Actions taking place on the 'arm' of abranch do not align themselves with thesource code which has to be written. Thefollowing Pascal example should clarify mymeaning. Compare it with Fig.1.

IF B>27 THENBEGINIF 1<50 THEN CALCWAL (Q)ELSEBEGINB:=9:

LINEGEN

X tOY=0

A .241T1STB =A-X1STC =B-X1ST

XY

PLOT

X =X1

B<O7

Y

B -BC B tBA

X.X1ST?

CIRCGEN

Fora circular arc centre (0,01. talus 'with 0<.< y

y

0

Y

A -

B+1- 2* -

X X + 1

A +

B + A

Y- Y - 1B -B - 4*T

I: =5ENDELSEBEGINB:=B+ 1:C:l + B/5:X:= B*7:CALCZFACT (X)ENDEND:

Notice how the layout becomes linear whencompared with the diagram. It may not beeasy to convince yourself that this is indeedan equivalent program to that in Fig.1. Theescape route is the bent arrow appearing onthe left of Fig. 4. A description of the actionsin Fig.4 should illustrate its function.

The calculation is performed and a testmade for Z=20. Assuming that Z does notequal 20. A is divided by 9. This done thesmall triangle diverts control flow to theescape route leading to the bottom of thediagram.

As with loops, escape routes can be nested.and a feature not allowed in decisions (well.frowned on by me). that of a second decisionjoining up with an outer one, is positivelyencouraged with escape routes.

AND FINALLY...

I have been developing and using my flow -diagram technique for the past ten years.However, I feel sure that having startedalong this road some of you will offerimprovements and enhancements.

In the February issue of E&WW. Medesput the case for using flow diagrams, givingtwo examples. Figure 5 shows these sameexamples redrawn using my technique.

David Sweeney B.Sc. is a real-time softwareengineer currently working with GEC soft-ware. Since graduating in applied physics atLiverpool Polytechnic in 1975 David hasdone various jobs in data processing, includ-ing work on Tornado. One of his mainambitions is to win the Micromouse com-petition.

TRANSCONDUCTANCEAMPLIFIERS

CORRECTIONS

Several errors appeared .n the abovearticle in the June issue, for which weapologize to readers and to the author.

The order of the first three figures, onp.580, was incorrect leaving the captionswhere they are and transposing the fi-gures in the order 3:2:1 corrects thelayout

v ROE 2

VAF R1

In equation 4 on p.581, VFE should readVAE, and a line is missing in the sixthparagraph from the end of the article.beginning "The differentially configuredoutput stage...". The third sentence shouldread "An isolated signal source may beconnected directly across A and B andsupply rejection is achieved even if thesignal source has finite impedance."


Sarcastic. Eccentric genius. A man withless than a competent grip on theordinary affairs of life. A legend in his

lifetime. Irascible. A gentleman of the oldschool. A dear old man. A hermit. All of thesecomments have been used to describe OliverHeaviside.

Heaviside was a self-taught mathematicalphysicist whose discoveries made long-distance telephony possible. Traditionalmathematicians and physicists struggled tounderstand his pioneering publicationswhich initially he had to fight to get pub-lished. To practical engineers they weredouble -dutch.

He waged war on mathematicians (whomhe said were as plentiful as mushrooms). thescientific establishment (the 'brays of theBritish Asses' was a phrase he used ofmembers of the British Association for theAdvancement of Science), and the infesta-tion of the then current system of units withthe 'disease of 4n..

He was probably the first to use theconcept of negative resistance, was one ofthe earliest engineers to use vectormathematics and he used his own calculussymbols. He suggested the existence of aradio -wave reflecting layer in the upperatmosphere and was the first to elaborate theskin effect. It was he who simplified Max-well's electromagnetic theory having clearedaway the debris of the battle fought byMaxwell. according to one biographer. And itwas he who showed - insisted - that induct-ance is a necessary part of a telecommunica-tions line. This insight led to the inductiveloading of telephone cables and the subse-quent leap in their performance. In short.Heaviside rewrote the theory of telecom-munications.

Because of his unusual, near -incomprehensible approach, recognition ofhis genius was slow in coming. When hon-ours did come his reaction was typical of theman: "I think honours have been very muchoverdone; the more honours the less value.It is depreciating the currency". One or two,he felt, were enough: and yet graciously hefelt it unpleasant to refuse.

Perhaps the Faraday Medal of the Institu-tion of Electrical Engineers (!EEL of whichhe was the first recipient. was the honour

PioneersW. A. ATHERTON

8. Oliver Heaviside (1850 - 1925):champion of inductance.

hutitutionof Kitt:Inc:II Engineers

that gave him most pleasure. This annualaward was. in part, specifically created tohonour Heaviside.

He was a Fellow of the Royal Society andonce wrote that maybe he should havestopped with the tail FRS, "with perhaps agood honorary degree to balance the tail, bygiving one the title of Doctor". Even so, heobserved. there are doctors in nearly everystreet (times must have changed) and to thepublic FRS means "nothing at all, being lessthan FRHS, to which our respected garden-ing townsman belongs by paying a guinea".

When the Faraday Medal was taken to himat his home in Torquay in 1922. he was 72years old and living alone in a pleasant housedecaying from long neglect.

J. S. Highfield, then President of the IEE.took the medal on behalf of the Institution.He recorded that he was met by Heaviside ina weed -covered drive and taken into a

furniture -laden hall all covered with dust. InHeaviside's room the walls were paperedwith prints and reproductions of publica-tions. They presented a pictorial record ofhis life. He retained his impish criticism andcaustic wit and he was genuinely pleased atreceiving the award. The fact that the medalwas made of bronze and not gold compen-sated. in his eyes, for the "wasteful expense"of the leather -covered vellum document.

"His way of life made a pathetic back-ground to his mental activity," wrote High -field*, "but I am sure he did not regard it aspathetic."

Highfield's account is full of pathos. Onreading it one is left with a sense of wonderand the wish that these unique moments inthe life of a unique man could have beenfilmed for posterity. That would not havebeen to Heaviside's liking, however. Even apublished photograph, he thought. "makesthe public characters think they really arevery important people". On one occasion hegave away a framed and glazed photograph ofa group of LEE members to a Newton Abbotfurniture dealer along with an old kitchentable. He described the photo with thewords, "Giants at the back. Pigmies at thefront".

BEGINNINGS

Little is known of Heaviside's early life. Hewas born at Camden Town. London, on May18. 1850. the fourth son of Thomas andRachel Elizabeth Heaviside. Charles Wheat-stone was his uncle. having married hismother's sister. Both of his parents livedlong enough to see him achieve some fame.

The family came from Stockton-on-Tees:his father. brothers, and two uncles werewood engravers. His mother came fromTaunton. Evidently Oliver inherited someartistic talent for two of his drawings sur-vive. one at least from the age of eleven.

Progress in photography must have madelife harder for engravers and it may explainthe family's move to London about 1849.

Oliver was educated at Camden HouseSchool and examined at 15. coming fifth andwinning the Natural Science prize. He re-ceived no university education, though.What he subsequently learned of mathema-tics, physics and engineering was self-taught.

In his mid -twenties. supported by hisfamily. he studied and worked through thenight in his room at his parents' home. withwindows and door shut, and an oil lampburning. One has a picture of a serious manat work, oblivious to the stuffy atmosphere

'In "Oliver Heaviside.' hy Sir George Lee. British Council.1947.


14,

'1 1..Svar

around him. It is said that when he wasworking during the daytime. food would beleft outside his door for him to take when hewas ready.

From around 1866 to about 1874 itappears that he was employed by the DanishCable and Great Northern Telegraph Com-pany. He became the principal operator atthe Newcastle-upon-Tyne end of theNewcastle -Jutland submarine cable, usuallyoperating on the day duty. His brotherArthur was a divisional engineer with thePost Office at Newcastle. There the twoexperimented with duplex telegraphy.

Heaviside's introverted and at times suspi-cious nature was perhaps encouraged byphysical disabilities: he was deaf and hadlittle if any sense of smell. For short periodshis deafness vanished and he could hearperfectly. He enjoyed walking, a pipe ofstrong tobacco. music. 'and especially cycl-ing. One of the few photographs of himshows him with his bike. He hated alcohol.

In 1889 his parents moved to Paignton inDevon and Oliver went with them. Aftertheir death he moved to Newton Abbot(where he gave away the Giants and Pigmies)and lived there alone from 1897 to 1909. IIIhealth struck and he was offered accom-modation in Torquay by a friend. Subse-quently he bought that house and saw outhis days there as a hermit in solitary theore-tical research. In his last years he bought hissupplies via the local policeman who wouldattract Heaviside's attention by blowing hiswhistle through the letterbox.

In this house, with its shortage of homecomforts, he became a scientific guru - afont of all wisdom. Researchers wrote to himfrom many countries and he developedlengthy and friendly correspondences withseveral, including Heinrich Hertz. Oxfordphysicists dubbed his hermitage the Inex-haustible Cavity. One story has it that thePost Office even recognized the title, deliver-ing a letter addressed to him at "Inexh. Cavy.Torquay."

It was a far cry from the welcome given tohis earliest papers.

-Oliver Heaviside. the Man by the late Dr G. F. C.Searle. mentioned in our June books column. isavailable at £12 post free from C.A.M. Publishing.P.O. Box 99, St Albans AO 4HQ.

This drawing, made when he was about 11years old, is inscribed "2nd work by OliverHeaviside (no others preserved)".

THE FOOLOMETER

Heaviside's first published article appearedwhen he was 22 years old. in The EnglishMechanic. A year later he had progressed tothat respected scientific journal the Philo-sophical Magazine. with a paper on the bestarrangement of the Wheatstone Bridge foruse with specific circuit configurations. Heeven discussed the question of a name for thebridge, finding none of the terms then usedwholly satisfactory: these included bridge,balance, lozenge. parallelogram, quadrangleand quadrilateral.

His careful attention to the choice ofwords led him to adopt many names forelectromagnetic phenomena at a time whenthe terminology was in a state of transition.

Much of his terminology has remained,including resistivity, conductance. induct-ance, permittivity. impedance. admittance.and distortion. There were others. The vec-tor mathematics term 'curl' was suggestedby Maxwell, used by Heaviside and universal-ly adopted. Kelvin grieved over that - he hadpreferred rotation or spin. Heaviside alsoused 'div' and may have originated it. And headopted from Rayleigh the term attenuationas being "the very thing I wanted".

As Heaviside's writings grew more com-plex the troubles began. One editor stoppedpublishing his articles because hardly any-one read them and another observed thatonly a few professors read them: "Professors.you see, do not advertise".

A commercial magazine bravely took himon: and from 1882 to 1887 The Electriciandevoted about 500 pages to his articles.Whilst these established his reputation,understanding of his work was slow to dawn.

Eventually his articles on electrical com-munications and electrical theory were col-lected and republished in book form. Theybecame classics: 'Electrical Papers', twovolumes, 1892: and 'ElectromagneticTheory', three volumes, 1893-1912.

Editors pleaded with him to make hiswritings more understandable. Apart fromhis unusual (vector) mathematics, he often

omitted steps in the argument. "Our re-ferees . . . complain of the exceeding stiff-ness of your paper. One says it is the mostdifficult he has ever tried to read", wrote oneeditor. To another Heaviside explained thathe was not writing for posterity but for thepresent "stiff-necked" generation.

Even Hertz on one occasion said he couldunderstand Heaviside's letters better thanhis book. Another engineer -scientist sug-gested that Heaviside write a simple bookand offered his services as a "foolometer".

TRIUMPH

In 1889 came what one writer has calledHeaviside's complete triumph. At the formaloccasion when the Society of TelegraphEngineers changed its name to the Institu-tion of Electrical Engineers. Lord Kelvingave the address. In doing so he reflectedthat his own famous and well -used telegraphequation had ignored inductance completelybut that now Heaviside had worked it out infull. From then on. Heaviside's work re-ceived more of its proper due.

Over long distances on a telephone linethe different frequencies tended to arrive atdifferent times, rendering long-distancetelephony impossible. Heaviside showedthat a 'distortionless circuit' could beachieved if inductance were used to compen-sate for the effects of capacitance. Theofficial view in Britain and abroad disagreed.Only when M. I. Pupin in America proved thepoint in practice did views change. Eventual-ly. induction loading coils and continuously -loaded cables made long-distance telephonya reality. Heaviside meanwhile viewed offi-cialdom with scorn.

Eccentric genius is hardly adequate as adescription of him. To some opponents hissarcasm hardly wavered, and yet he held nopompous view of himself. From about 1918he called himself The Worm and signedhimself Oliver Heaviside. W.O.R.M. Heshowed an impish delight when others tookthat as a distinction.

Much earlier he said of himself. "I am notfit for a cook: I forget. Then it all goes tocinder". G. F. C. Searle recalled having teawith Heaviside at Newton Abbot in 1905."The tea party was a bit odd." he wroteaffectionately: "the teapot spout was blockedwith tea leaves so Oliver Heaviside simplypoured from the lid."

Searle was perhaps the last to die of thosemen of science who really knew Heaviside.He described him as a strange but veryattractive person, with an unquenchablespirit of fun.

Early in January 1925. Heaviside wasfound unconscious in his house by hisfaithful policeman friend and he was re-moved to a nursing home. There he im-proved, was full of fun and enjoyed theattention of the nurses and the good food.

To physicists and telecommunications en-gineers he was a legend. To those nurses hewas simply a dear old man. The dear old mandied on February 3, 1925.

Tony Atherton works at the IndependentBroadcasting Authority's engineering train-ing college in Devon.


Countdown tosmall -dish satellite tv

A recent satellite broadcasting conference revealed details ofBritain's direct broadcast satellite tv service,

due to start late next year.

Barring accidents duringthe launch. British view-ers will have a further

three television channels tochoose from by Christmas nextyear. British Satellite Broadcast-ing, the contractor appointed bythe IBA. has now selected theHughes Aircraft Company toprovide the two satellites neededfor its service. A draft contractarrived from Hughes the nightbefore the conference began.

Making a break with the con-ventions of the commercialsatellite world, Hughes will hand the satelli-tes to BSB only after they have been tested inorbit. This supply -and -launch arrangementis designed to help restore confidence in thespace industry after its spate of recent fai-lures, and it will greatly reduce the financialrisk to BSB. Other companies, includingBritish Aerospace, were on BSB's short-list,but they failed to match the Hughes package.

Each satellite will be equipped with three110W transponders capable of an e.i.r.p. of59dBW in the 12GHz band, giving high -power coverage of whole United Kingdom.Reception will be possible on 30cm dishesand BSB say the receiving terminals willsoon become a mass -market item costingonly £200.

Plans for the new services were outlined tothe conference by BSB's Graham Grist. Thethree channels will be used for four pro-gramme streams. Now would be a news andevents channel with a major input fromIndependent Television News; Galaxy wouldinclude light entertainment. quizzes andshows; and the remaining channel would beshared between Zig-zag, with daytime prog-rammes for children and mothers at home,and Screen. which would offer in the even-ing a mixture of feature films ranging fromnew releases to cinema classics. The lastservice would be available on subscriptiononly, at about £10 per month; the others

Hughes Aircraft's HS 376 satellite, to beused for BSB's d.b.s. services to the UK,weighs about 630kg and has a design lifeof 10 years. The antenna and telescopingsolar panel extend to about 7m overall inorbit. Broadcasts will be receivable on adish 30cm in diameter. Thirty versions ofthe spacecraft have been purchased bycustomers in six countries. BSB's twospacecraft will cost it £200M.

would be supported by advertising. But allwere to be encrypted. This was because BSBwould buy only the UK distribution rights toprogrammes and would have to preventreception abroad. However, decoders wouldbe freely available on the home market.

The IBA, said Mr Grist, recognized theneed for commercial success of the newservice, and had exempted BSB from some ofthe regulations which govern its other con-tractors - for example, the duty to supplyprogrammes for minority tastes and toguarantee a wide showing for programmesof merit. However, BSB would aim for a highgeneral standard of programmes. with im-partiality and without offence to decency. Itexpected to spend over £100M on program-mes in its first year and it would accompanythe launch by a promotional campaignwhich would put it among the top sixadvertising spenders.

Transmissions will be on theD -MAC standard, for which Brit-ish Government approval is stillawaited. No chip set is yet avail-able for D -MAC. but Mr Grist toldthe conference that the Nordiccountries now had on the draw-ing board a set which wouldcover all three MAC standards, C.D and D2. RCA were prototypingthe video chip and would havesample quantities ready by theend of the year; Plessey and ITTwere also taking an interest. Heexpected adequate supplies in

time for the launch of the BSB service.A special feature of the d.b.s. service will

be the high data -carrying capacity of the D -MAC transmission. This BSB will use inthree ways: for programme -related teletext.for subscriber services such as financialinformation. and for toll services such asbusiness messaging. The company intendsto maximize its use of this capability, tounderpin the service whilst advertising re-venue and a subscriber base are building up.

Mr Grist was confident that his servicewould begin on time. The launch vehicle. aMcDonnell -Douglas Thor Delta I, had showna 98% success record over the past 10 years.The standby satellite would be placed in orbitby a second launch, using the more powerfulDelta II; this was scheduled for six monthsafter the first launch.

DATA SERVICES

More information about the possibilities ofbusiness messaging was provided by AdrianNorman of Direct Business Satellite Sys-tems, who hoped to reach agreement withBSB shortly to start such an operation.Through the vast capacity of the Britishd.b.s. satellite, his company could offer astore -and -forward electronic mail servicewhich would significantly undercut its ter-restrial rivals.

Four data channels on each of the satel-


Astra will bring 16 television channels to a European audience which could grow to 40 million by 1996. The satellite. to be operatedby the Luxembourg company Societe Europeenne des Satellites, is due for launch on Ariane's Flight 27 in June or September 1988.

lite's transponders would give a capacity of13.4Mbit/s averaged over the day. and out ofprogramme hours the entire picture fieldwould he available for data. The cost ofsending a megabit over the space segment ofpath worked out at about 11 p - enough tocarry four pages of faxed typescript, a minuteof compressed 64Kbit/s voice, or 128Kbyte ofcomputer data.

Messages would he received via set -topdemodulators, which would add relativelylittle (£10043001 to the cost of a d.h.s.receiver. Demodulators would he individual-ly addressable: only those terminals author-ized to receive a particular message would he

able to display or print it.To ensure low administrative costs me.

sages would he pre -paid. Each one wouldcost a few pounds to send. However, mostitems sent through the present-day mailwere copies or near -copies of other mess-ages. And since the satellite system would heable to address the same basic message tomany recipients. the cost of an additionalcopy could he as little as 1p. Mr Norman saidthat he was re -introducing the penny post.

For senders. the service would mean theycould fire and forget: next -morning deliverywould he guaranteed. Recipients would in-cur no costs in receiving messages. other

DMAC TRANSMISSION STANDARD

Listed here are features of the EBU D -MAC packet system to be adopted by the IBA and BSB.A full specification will be published later: some details remain to be settled. The IBA willput out test transmissions to help manufacturers check compliance and tolerances.

Vision: f.m. with pre -emphasis and d.c. restoration. Time -compressed luminance andchrominance information is multiplexed on to each line: alternate lines carry RN or B -Ysignals. Basic standard for Europe is 625 lines. 25 pictures per second.

Data: duobinary coded packet system at 20.25MbitIs, f.m.

Aspect ratio: normally 4:3: 16:9 possible. Line 625 is a status line giving details of this andother transmission parameters, enabling receivers to be reconfigured automatically.

Sound: stereo or mono, 15kHz bandwidth. Commentary channels, 7kHz.

Teletext carried in the data multiplex.

Vertical blanking interval: reserved for future high -definition enhancements.

Encryption: by double -cut component rotation and the shared key over -air addressingsystem known as EBU System B.

Full details of the MAC systems are given in the European Broadcasting Union's technicaldocument number 3258.

than hardware costs. With present e-mailsystems, they have to pay for computer timeeven if they find no message waiting.

Demonstrations of the system would takeplace on the Olympus satellite and DBSS waslooking for early customers willing to takepart in the trials.

Voices from the floor commented thatd.h.s. transmission was a one-way processand asked how senders might be given theconfidence that their messages had arrived.A small number of non -received items mightcause havoc in the system. The speakerreplied that some redundancy would beprovided: a list of messages to come wouldprecede the main transmission by somehours, and the list would he repeated sometime afterwards. If a recipient missed amessage, he could call the centre: largerorganizations could use viewdata to call forrepeats or to resolve problems. The expectedhit error rate was 10 ' or 10 and so themost likely cause of failure was accidentalremoval of the receiver's mains plug.

D -MAC AND KITCHEN FOIL

Brian Salkeld of the IBA spoke in support oftwo IBA papers in which reasons for thechoice of D -MAC were set out.

In choosing standards for the new service.he said. it was necessary to take a long-termview. Chips now being made for the forth-coming French and West German d.b.s.services did not allow scrambling and therigid arrangements for sound lacked theflexibility of the D -MAC packet system.

The paper by Chris Dauhney of the IBAargued that the principal alternatives. C-MAC/packets and D2 -MAC/ -packets were


Inding a message. Theid subscription costs at

ner recipient's end (Direct'0 v Id)

co.sts.0 $''' '

/2,0 ; 'oc-d ode44,0 e se.0,,,00 v'ssSz' 13 18 100"-s) 10 25 25

10 20 40..40s) 5 10 36

...--.1 iIts) 1 25 200IP 0 words) 1 2 200le

min. fYP- max.

.uropean satellite tv systems (Philips -

France).

mice Vision Sound.CIR Pal/Secam 1 f.m.

France, FRG Pal/Secam 2 f.m.France Pal/Secam 4.6 digital ASweden Pal/Secam 8 digital AUK BBC Extended Pal Digital

-462 UK IBA MAC Digital6/1983 EBU MAC C/packets6/1984 Independent MAC D2/packets12/1984 EBU MAC C/D/D2 packets7/1985 Ind..Fr..FRG MAC D2 packets

Applications of the various MAC systems (GwynMorgan. Logica). Opinion is divided as to whetherD2 -MAC can evolve to cope with high -definition tv.

HDTV DBS Cable SMATV Home video

D2

C0

both less satisfactory: the former because itcalled for a more expensive receiver(although it gave slightly better resultsunder low signal strength conditions) andthe latter because it had only half the datacapacity and so failed to meet the criterion ofmaking full use of the channel.

At the end of his presentation. Mr Salkeldproduced with a flourish a small 12GHz dishantenna which looked as if it might havecome from BBC-tv's 'Blue Peter': it wasmade, he said, of kitchen foil, a plasticslemonade bottle and its cap. Plasticene. ashort length of copper water pipe, and papiermache composed of old copies of the Finan-cial Times (the newspaper owned by thePearson group, one of BSB's founder share-holders). Setting the aerial on the top of a tvset, he was rewarded almost immediately bya healthy set of colour bars - which showed.he said, how simple d.b.s. reception wasgoing to be. Discussing the antenna after-wards, he revealed that the demonstrationhad been genuine: the microwave transmit-ter at the other end of the conference hallwas set to give the actual field strengthexpected from d.b.s.

TAKING THE RISKS

All this, of course, is contingent upon asuccessful launch of the BSB satellite. Aspace insurance consultant. Rodney Buck -land, examined some of the technical risks.With civilian communication and earthobservation satellites, losses up to theapogee kick motor stage had risen from 8%to 18% over the last decade (taking afive-year moving average). Nothing seemedto have gone right for the space industry.Challenger. Titan. Delta and Ariane had allfailed within months of one another lastyear. Atlas. Centaur, two Protons and anIndian satellite had also failed. Mr Bucklandhad just re-examined the figures for Deltaand had found the true success rate to be

well below the 98% optimistically quoted byBSB. That figure. he said, related only to thefirst stage of the launch. Of 39 basicallysimilar Delta flights over the past ten years.three had resulted in a total loss of thesatellite due to the launcher. Further fail-ures had occurred on transferring satellitesinto orbit or during commissioning, makinga cumulative loss rate of about 13%: betterthan the industry average of 16-18%, butworse than Mr Grist had suggested. A speak-er afterwards observed, to laughter, that thisone -in -six rate was the same as the chance ofbeing killed in Russian roulette.

Last year. underwriters had lost $400M onsatellite business on a turnover of $900M (afew claims were still unsettled) and so had agood deal of ground to make up.

Traditionally, satellites had been builtlaunched on a 'best effort basis' by thecompanies concerned, with no guarantee ofsuccess. Mr Buckland congratulated MrGrist on securing his 'buy in the sky'package deal with Hughes. which offeredthe industry a way forward where insurershad failed it.

WHO WILL PAY?

The main motive power for d.b.s.-tv comes.of course, from the advertising industry.Robert Dodds of BBDO. the second largestadvertising agency in the world. outlinedsome of its potential advantages over themore restrictive climate of terrestrial tv.Currently, only two countries in Europeoffered iv advertisers the freedom to buytime as and when they wanted it: one wasItaly (with an astonishing 750 minutes ofadvertising time per day) and the UK (withabout 144 minutes). In other countries,commercials were shown in blocks of up to12 minutes; airtime might have to he appliedfor up to 15 months before transmission.This system denied advertising buyers theability to book specific time -slots or to targeta particular group of viewers; and it gaveconsumers the opportunity to miss thecommercials altogether. In addition, it de-prived advertisers of the leverage a sym-pathetic editorial environment could bringto their commercials.

In this context, the audience was divertedby another speaker from the marketingworld who sought to demolish the viewexpressed by one sage that satellite tv wouldprove to be ler. excuse us) a cornucopia ofcrap. The hubbub which followed his show-ing of a montage of programme highlightsseemed to hint that there were still a fewsceptics around.

BBDO estimate that around $1400M of tvadvertising business (at 1986 prices) is cur-rently unable to be placed because of restric-tions. Some proportion of this, said MrDodds. would be available to finance satellitetv through multinational companieswishing to promote pan-European super -brands.

ASTRA'S 16 CHANNELS

Europe's other high -profile d.b.s. vehicle isAstra, a privately -operated medium -powersatellite due to be placed in orbit by therevived Ariane programme in mid -1988.

Astra will essentially be a carrier of program-mes provided by others rather than anoriginator. The spacecraft will offer sixteen26MHz-wide downlink channels (with sixspares) in the range 11.2 to 11.45GHz. all ofthem battery -backed and available 24 hoursper day. Power output from the t.w.ts will be45W, giving a nominal e.i.r.p. of 50dBW(48dBW in the north of Scotland). Viewerswill be able to receive Astra on an 85cmdiameter dish; though if they want BSB too,they are likely to need a second microwavefront end - current models have insufficientbandwidth. Astra now has a tie-up withBritish Telecom. which is to build a fourthantenna at the London Teleport in theDocklands to uplink two of the channels. BTwill also market Astra services in the UK.

Marcus Bicknell of SES, the Luxembourgcompany behind the Astra project. discussedsome of its aims. UK viewers, he said, wouldbe offered two or three English -languagegeneral entertainment channels, a pay-tvfilm channel, plus five or six pan-Europeansingle -subject thematic channels devoted tomusic, arts or sport. These thematic chan-nels would be shared by other countries inEurope. where a similar pattern would berepeated. By this means Astra would giveFrench viewers double the number of chan-nels offered by their TDF satellite. One-thirdof leisure spending power in Europe was inGerman-speaking countries: these too couldexpect to be well -served.

Technical standards for the systems hadnot been decided. but would be announced

mid -July and early August.A likely contender is the D2 -MAC packet

system chosen for the French TDF1 andGerman TVSAT satellites. This was describedby Jacques Geninet. d.b.s. project managerfor Phil ips-Portenseigne. D2 -MAC is a

narrow -band MAC system with reducedsound and data capacity, tailored to suitcable tv networks in continental Europe.After the presentation, Paul Ratliff of theBBC Research Department commentedfrom the floor that if we were to adopt D2, wewere being constrained by yesterday's cablesystems. D2 had no capacity for expansion:high -definition tv could be accommodatedonly by doing without the sound. Geninetresponded that C -MAC was the best possiblesystem for satellite transmission: but thequestion was whether we should optimizefor satellites or go instead for a multi-purpose system such as D2 -MAC.

Another speaker at the conference, DavidCutts, of Direct Broadcasting Ltd. a Britishcompany specialising in subscriber author-ization for tv, observed that it was odd thatthe EBU - having got fairly near to astandard for MAC transmissions - actuallyencouraged a diversity of conditional accesssystems.

Meanwhile. British cable operators havedeclared their support for the IBA's systemby announcing agreement on a data andconditional access standard based on the20.25MHz data rate of C and D -MAC.

The 1987 European Satellite Broadcasting Conferenceand an accompanying exhibition were held at theBrighton Metropole in June. Further details from theorganizers. Online International. on 01 -868 4466.


Novel oscillatorsThree new oscillator discoveries - an active -C tunable

oscillator particularly suited to i.c. manufacture, a singleop -amp capacitorless oscillator, and a current -conveyor

based design with low sensitivity andnon -interactive tuning.

Single -amplifiercapacitorless oscillator

Accurate op -amp modelling produces a single op -ampcapacitorless oscillator that works reliably over a decade

frequency range. It uses at most four resistors.

M. T. ABUELMAATTI and W. A. ALMANSOURY

Anumber of schemes have been de-veloped for realizing capacitorlessoscillators over the past few years 1-10.

They exploit the frequency -dependent gainof practical operational amplifiers.

Analysis of such circuits usually involves alinear two -pole model for frequency -dependent gain of the op -amp. Using thistwo -pole model. which is well accepted forcompensated op -amps, capacitorless oscilla-tors cannot be realized with fewer than twoamplifiers'.

Using a more accurate model, a newcapacitorless oscillator with a single op -ampand at most four resistors can be realized. Anequivalent of the oscillator circuit, Fig. 1. isshown in Fig. 2 with the op -amp representedby its differential input capacitance andresistance" '12.

Fig. 1. Single -amplifier capacitorless oscil-lator designed using a more accuratemodel for the op -amp.

Open -loop gain of the op -amp is repre-sented by the conventional two -pole model

Ms) - AOWaWh

(s+walls+wb)

where Ao is d.c. gain and wa and wb are thefirst and second corner frequenciesrespectively.

Node equations of Fig. 2 are

y-x= x x-zZd R2 RI

x-y= y y-zZd R4 R,

z=Alx-y), 1

Zd=Rdli r.d Juit,

By eliminating x, y and z from the previousthree equations the characteristic equationof Fig. 2 can be shown to be a + (3A=0, where

a =oti +col I +ST)

=14_ +1.111+.11IR! KV

co=1/1+1+1+1Rd kRi 2 3 4,

1= CdRd

1 1

P=R2R3 Itt.

Using the first equation with this one thecharacteristic equation reduces to

-r

A (x -y)

R2

Fig. 2. Equivalent circuit of Fig. 1. Theoperational amplifier equivalent circuitincorporates differential input resistance/capacitance and frequency -dependentgain.

(a1 +a2(1+sTIl(s+walls+wb)-(-1.3A0wawb=0.

By equating the imaginary part of thisexpression to zero, frequency of oscillationcan be shown to be

(002=wawb1+et1+Ct2I 1+

1

Tat 10). wbij

In a similar way, the condition of oscillationis obtained by equating the real part ofthe last but one equation to zero. Note thatfor most practical cases wb is much greaterthan wa and the condition of oscillation canbe shown to be

Ao, a +a2 +6,02(al(1. +4.1)1M+oti)

Rwato,

From the previous two equations it is clearthat by using appropriate resistances pro-duction of sustained oscillation from thecircuit of Fig. 1 is possible.

We built and tested this oscillator using a741 op -amp. Figure 3 shows a typical outputwaveform using 10kf for R12, 68k12 for R3,infinity for R4 and op -amp supplies of ± 15V.

Variation of frequency with R3 is shown inFig. 4. When this resistance was varied from33 to 470kft, frequency varied from 350kHzto 27kHz.

For stable operation, op -amp parameterswa and wb must be stabilized, so we

therefore recommend a temperature -compensated op -amp such as the LM324.Power supply voltage should be stable.


fo 200 kHz

Fig. 3. Typical output waveform from theoscillator of Fig. 2 using R12= 101(11,R3=68kil and ± 15V for the op -amp.

0

0 100 200 300 400 SOD

RESISTANCE R3, (kill

Fig 4. Variation of frequency with R3.

References1. Battacharyya, B.B. and Raut, R.. A simplewide -range active -R oscillator. AEU vol.35. 1981,pp.457-459.

2. Abuelma'atti, M.T. and Alsawaty. A.H.. "Iwo newactive -R oscillators, International Journal of Elec-tronics vol.% 1986. pp.167-173.3. Shah, N.A. and Bhat, C.K.. Operational -amplifier based voltage -controlled oscillator. In-ternational Journal of Electronics vol. 59, 1985,pp. 649-652.4. Drake. J. M. and Michell. J. A., Continuously -tunable sinusoidal R -oscillator with a minimumnumber of elements. International Journal ofElectronics vol.50, 1981, pp. 141-147.5. Li, M.K. and Choi, K. C., On the synthesis ofsecond -order active -R filters, International Jour-nal of Electronics vol.47, 1979, pp.483-489.6. Moni, R. S. and Rao, K. R.. Large -signalinstability in active -R and a class of compensatedactive -RC filters, Proc.IEEE vol.69. 1981.pp.1366-1367.7. Nedungadi. A. and Venkatesrawani, S., General-ised second -order active -R filters, InternationalJournal of Circuit Theory and Applications vol.10,1982, pp.311-322.8. Pyara, V.P. and Jamuar, S. S., Single -elementcontrolled oscillator without external capacitors,Electronics Letters vol.16, 1980. pp. 607-608.9. Siddiqi, MA. and Ahmed, M.T., A low compo-nent oscillator with operational amplifiers andresistors, Proc. 12th ASILOMAR Conference onCircuits, Systems and Computers, 1978. pp. 12-13.

10. Venkatarmani, Y. and Venkatesrawani, S.,Active -R multifunction circuit synthesis, Electro-nics Letters vol.18, 1982, pp.96-97.11. Mandour, 1., Elawady, R. and Abdelati, 1.,

Macromodelling of operational amplifiers and itsapplication to active filters, IEEE 1983 Mediterra-nean Electrotechnical Conference. pp. B1.07/08.12. Treleaven, D.H. and Tofimenkoff. F.N.. Mod-elling operational amplifiers for computer -aidedcircuit analysis, IEEE Transactions on CircuitTheoryvol.CAS-18, 1971, pp. 205-207.

Active -C oscillatorWith low sensitivity and non -interactive electronic tuning,

this resistorless o.t.a. oscillator is particularly suited tointegration.

M. T. ABUELMA'ATTI and R. H. ALMASKATI

0 perational transconductance ampli-fiers have three main advantagesover standard operational ampli-

fiers. They require few if any resistors,tuning of their transfer gain (gm) is electro-

nic and highly linear, and their high -frequency performance is more reliable. Forthese reasons, o.t.as are replacing op -ampsin oscillator and active filter designs.

Recently, a number of o.t.a.-based oscilla-

0

Fig. 1. Proposed resistorless oscillator uses five operational transconductanceamplifiers and two grounded capacitors.

tors have been proposed". Those by Saha etand Nandi2 are not particularly suited to

integration because they include RC com-binations to determine the frequency andcondition of oscillation. Non -interactiveelectronic tuning of the amplifier proposedby Ahmed et al. is impractical because ofinterdependence between the frequency andcondition of oscillation.

Our new o.t.a. active -C oscillator has twogrounded capacitors and is therefore highlysuitable for integration. Moreover it hasnon -interactive electronic tuning and lowsensitivity.

Figure 1 shows the oscillator, whosecharacteristic equation is

s2g.,5C1C2 +S.gm3(gms -gm4)C1

+ gm Igm2gm5 =

where gm, is transconductance gain of theo.t.a. number i. By equating the real andimaginary parts of this expression to zero,the condition and frequency of oscillationcan be shown to be gm4=g,5 and

2 gmlgm2C1C2

From the previous two equations it is clearthat the circuit has non -interactive electro-nic tuning since frequency of oscillation canbe varied without affecting the condition ofoscillation. Moreover by making gm, equalto gm2 and since gm, is directly proportionalto direct bias current of the o.t.a., frequencyof oscillation will change linearly with directbias current (or bias voltage).

Sensitivity figures are calculated from thedefinition

Sw°=LY coo aY

where w° is oscillation frequency and y is theelement of variation. Using this equation, w°sensitivities are

Sdwo = SdW° = -SW0= - SW0= 1/2em I e.m2 C1 C2

which indicates that sensitivity figures arelow.

We built and tested the oscillator using aCA3080E o.t.a. and polystyrene capacitors.Figure 2 shows the variation of oscillationfrequency with direct bias current 1BI =In -When bias current was varied from 10 to110µA the change in frequency was from 5

SO

40-

0-

0 20 40 60 80 100 120

DC BIAS CURRENT 1131=IB2, (pA)

Fig. 2. Variation of oscillation frequencywith direct bias current


to 51kHz. Currents and capacitors used were4.51LA for IB3 and IBS, 27p.A for 164 and 47pF

for C12.

References1. Saha, A.R., Nandi, R., and Nandi, S., Integrabletunable sinusoid oscillator using DVCCS, Electro-nics Letters vol. 19, 1983, pp. 745-746.

2. Nandi. R.. Novel variable -frequency oscillatorwith single -resistor control using DVCCS/DVCVS,Electronic Letters vol. 18, 1982 pp.144-145.3. Ahmed, M.T., and Khan, 1.A., Realization ofintegrable electronically tunable oscillators withoperational transconductance amplifiers, IEEEInternational Conference on Consumer and In-dustrial Electronics and Applications 1984, pp.20-21.

Current-conveyorRC oscillator

One current conveyor makes a low -sensitivity sinewaveoscillator with non -interactive single -capacitor tuning.

M. T. ABUELMA'ATTI and N. A. HUMOOD

Because of its distinct advantages overthe operational amplifier, the currentconveyor has attracted the attention

of many researchers in the field of activefilters and oscillators'. Recently a number ofoscillator circuits using a single currentconveyor have been proposed'. Here weexplore the possibility of a new design forsuch an oscillator.

Y4410-1--)

Y5

Y IX

CC n

IFig. 1. Current -conveyor oscillator formsthe basis of the new oscillator.

Fg. 2. New RC sinusoidal oscillator usingthree grounded resistors and three capaci-tors, two of them also grounded.

Three grounded resistors and three capa-citors are used, two of them grounded andone not. Assuming that the current con-veyor in the oscillator structure of Fig. 1 isideal, with ix= iz, iy= 0 and vx=v, routineanalysis yields the characteristic equation ofthis circuit configuration given by

Y3(Y4-1-Y6+Y6)+Y4(Y6+Y6)=Y4(Y1 +1(2).

By performing all possible permutations toexplore the possibility of oscillation, twooscillator circuits resulted. With Y, =GI,Y2 4-0C2, Y3 =je0C3. = 4.005 andY6 = G6 the oscillator circuit reported by Janaand Nandi is obtained, and with Y, =GI.Y2 =..002, Y3 = G3, Yet =j(0C4, Y5= G5 andY6=jwC6 a new oscillator is obtained, Fig.2.

Oscillation frequency and the condition ofoscillation of the new circuit can be obtainedby equating the real and imaginary parts ofthe resulting characteristic equation to zero.Frequency of oscillation will be

2 G3C5

C4(C2- C6)

and the condition of oscillation will be

G IC4 = G3(C4 C6)+ G5C4.

From these two equations it is clear thatoscillation frequency can be changed with-out affecting the condition of oscillation.This can be done by changing C2.

The various sensitivity figures are calcu-lated using sensitivity definition

Sw°= aw°Y wo. ay

From this equation, oh, sensitivities for Fig.2 were calculated and are given by,

St°°=SW4)= -Srt°°= 1/2,St"= -1/2G3 G5 v4 1C2

( C21--

Swo_ I/2C6 Ica 11

kC6

Using the current conveyor implementationof Fig. 3, the oscillator in Fig.2 was built andtested. Figure 4 shows variation of oscillator

100k 100k

Fig. 3. Current conveyor building-blockused to implement Fig. 2.

80

7,3

>_ 60-

ta..1

ci

u- 40-

20Th

0 20 40 60CAPACITANCE C2, (nF)

Fig. 4. Variation of oscillation frequencywith capacitance C2.

80

frequency with capacitance C2. Componentsused to obtain these results were RI = 1.4k11.R3= 11.2kil, Cs= 100pF, and C6= 500pF.When capacitance C2 was varied from 22 to79nF, oscillation frequency varied from 77to 38kHz.

References1. Kumar. U. and Shukla. S.K., Recent develop-ments in current conveyors and their applica-tions, Microelectronics Journal vol. 16, 1985,pp.47-52, No.1.2. Jana, P.B. and Nandi, R., Single currentconveyor tunable sinewave oscillator, ElectronicsLetters vol.20, 1984, pp.45-46.Naqshabandi, S.F.N., Shah. N.A. and Sharma,R.S., On the realization of a voltage -controlledoscillator. International Journal of Electronicsvol.56. 1984. pp. 359-363.Senani, R., New canonic single -resistance con-trolled sinusoidal oscillator using a single cur-rent conveyor, Electronics Letters vol.15. 1979,pp. 568-569.Soliman, A.M., Realization of frequency -dependent negative -resistance circuits using twocapacitors and a single current conveyor, Proc.lEE vol.125 1978. pp.1336-1337.Soliman, A.M., Simple sinusoidal active RCoscillators, International Jo'.... Ai of Electronicsvol.39. 1975, pp. 455-458.Soliman, A.M.. Novel variable -frequency sinu-soidal oscillator using a single current conveyor,Proc.IEEE vo1.66. 1978, p. 800.

The authors are in the department of elec-trical engineering and computer science,University of Bahrain.


RESEARCH NOTESFluoride for

long -hop fibresOptic -fibre technology is de-veloping so rapidly at the mo-ment that most of us would bevery happy to have a few shares insilica glass. But although thebest mono -mode silica fibres cannow achieve attenuations downto 0.2dB km'. even this makes itimpossible to dispense with re-peaters that are inevitably expen-sive and awkward to power. Theplanned transatlantic fibre linkTAT -8 will, for example, requireboosters every 50km. And whilethe fibres themselves require noelectricity, the boosters certainlydo. necessitating power cables inparallel with the fibres.

What may well oust silica inthe end is one of a new range ofmaterials based on fluorine che-mistry. especially berylliumfluoride and fluorozirconates.The potential advantage offluoride glasses is that they willtransmit the longer wavelengthsthat are increasingly opaque tosilica. Over the last decadethere's been a steady move intothe infra -red part of the spec-trum as the technology - espe-cially laser technology - hasmade it possible. The phe-nomenon of Rayleigh scatteringdictates as long a wavelength aspossible because of a fourth -power loss/frequency rela-tionship. Unfortunately. if thewavelength of transmission getsmuch beyond the current1.55µm maximum. then a newloss mechanism comes into playdue to resonances in the chem-ical lattice of the material. Thisplaces a physical limit on silicafibres. Fluoride glasses have adifferent chemical structure andwill therefore transmit longerwavelengths where Rayleighscattering is less of a problem.

So far. this promise has yet tobe fully realised, since fluoride -glass technology is relativelynew. Attenuation figures of0.7dBkm ' have been achieved inthe laboratory. but there areconsiderable problems in prac-tice because many of the mate-rials are chemically unstable orlacking in durability. Fluorozir-conate glasses are potentiallyprone to devitrify. that is tocrystallize spontaneously and be-come opaque. Berylliumfluoride, on the other hand. is

dangerously poisonous and veryprone to absorb water. Water isnot just a mechanical problembut a limitation on operatingwavelength: even small amountsin any fibre can introduce astrong absorption at about2.9µm.

For the future it looks as if thefluorozirconate glasses willprobably extend the operatingwavelengths down towards about2.5µm. where in theory operat-ing losses are much less thanthey are at I .55p.m. That, inturn. if the manufacturing prob-lems can be solved, heralds thepossibility of long-distance linkswith repeater spacings of severalhundred kilometres. Resear-chers in Japan, the USA and atvarious UK universities and Brit-ish Telecom are now activelypursuing that goal.

New researchin parallelcomputing

A major £10 million internation-al collaborative research project,to spearhead the development ofpowerful parallel computer sys-tems able to integrate numericand symbolic processing, isbeing led by Thorn EMI's CentralResearch Laboratories.

Involving participants fromfive European countries. thethree-year 'SPAN' project isbeing 50% funded by the EECthrough the Esprit programmefor Information Technology Re-search and Development. An im-portant aim of the project is toestablish a de -facto standard forsoftware/hardware interchangebetween parallel computers. Inaddition to Thorn EMI the col-laborators include CISMA SIN-TRA (France). PCS (West Ger-many), University College Lon-don. CTI (Greece) and INESC(Portugal).

The objective of SPAN is tobridge the gap between conven-tional computers supportingarithmetic operations and newertypes of symbolic computerssupporting Artificial Intelligenceapplications. The goal is to de-velop powerful parallel compu-ters able to undertake the rapidlyexpanding range of applicationsinvolving both arithmetic andsymbolic computation. Areas of

application already identified in-clude image interpretation,weather forecasting, and real-time expert systems.

The project will make exten-sive use of up-to-date researchon parallel computers, notablythe concept of paral el -processing building-block com-ponents, pioneered by Inmo in -the transputer family of micro-processor chips, together withthe results of other currentEsprit projects.

Magnetic peaksThe High Field Support Group atthe Clarendon Laboratory, Ox-ford, has achieved a magneticfield of 52 Tesla for 10ms using asolenoid fitted with a specialhigh -strength conductor (Phy-sics Bulletin Vol.38. No 5). Theresearch undertaken in conjunc-tion with the Catholic Universityof Leuven in Belgium used arelatively simple layer -wounddevice precooled in liquid nit-rogen at -196°C. Electrical ener-gy involved was about 80kJ at4kV from a capacitor bank.

The conductor, specially de-veloped at Harwell Laboratory.consists of a rectangular purecopper core surrounded by alayer of stainless steel forstrength. This 'wire', approx-imately 4mm-, was insulatedwith polymide plastic, woundinto a solenoid. vacuum impreg-nated with epoxy resin and givenan outer binding of Kevlar.

So far. the solenoid has suc-cessfully withstood fairlypunishing tests in which fields inexcess of 50T have been regularlycreated. There has been no signof either mechanical or electricalbreakdown. The Clarendongroup are confident that fields inexcess of 60T should be achiev-able with minor modificationsand still greater fields as andwhen more exotic conductorsare employed.

Towards aplastic magnet

With the exception of a few exoticalloys based on metals likesamarium, cobalt and neody-mium. virtually all magnets con-tain iron. So much so that the

term ferromagnetism was coinedto describe the property exhi-bited by all strongly magneticmaterials. Such materials owetheir properties to the fact thatthe atoms have unpaired elec-trons that align' their spins inareas called magnetic domains.When magnetized by an. externalfield. these domains themselvesline up to produce a net overallpermanent magnetism. What isspecial about ferromagneticmaterials is the unpaired elec-tron of the iron atom. Mostatoms and molecules have pairedelectrons with opposite spinswhich means that any magneticmoments cancel out, leavingonly weak magnetic propertiesknown as paramagnetism anddiamagnetism.

In principle. there's no reasonwhy it shouldn't be possible tocreate molecules that have oddnumbers of electrons and whichas a consequence exhibit ferro-magnetism. In practice it's ex-ceedingly difficult because of thechemical reactivity of moleculeswith unpaired electrons. Suchradicals, as they're called, tend tocombine with 'other radicals toeven things out.

Recently, however, there havebeen a number of research re-ports indicating that organic po-lymers and crystalline salts canbe made ferromagnetic, just assome of them can be made con-ductors of electricity. In fact theparallels are remarkably close.and just as some normally in-sulating materials becomesuper -conducting at low temper-atures, some normally non-magnetic materials become fer-romagnetic.

One such material, synthe-sized in the USA. employs twodifferent constituent molecules:decamethylferrocene and tet-racyanoethylene. The first ofthese loses an electron to thesecond. resulting in ions withunpaired electrons that line upalternately to form a latticestructure. At temperatures be-low -268°C the electron spins fallinto line, causing the crystal tobehave like an iron magnet.

Other materials that hold outthe promise of exhibiting ferro-magnetic properties include avariety of polymers developed byresearch groups in 1 -man. theUSA and the USSR.

One intriguing material madeby reacting triaminobenzenewith iodine is claimed to be


RESEARCH NOTESferromagnetic. The problem, asthe researchers from IBM readilyadmit, is that the reaction is

rather unpredictable and the re-sulting black material extremelyunstable. Nevertheless as muchas 2% can consist of a ferro-magnetic polymer that main-tains its properties up to almost400°C.

Rather similar results havebeen achieved in the USSR andJapan by polymerizing an acety-lene derivative. The Russiansfrom the Institute of ChemicalPhysics and Mendeleev's Insti-tute of Chemical Technology(Nature vol.:326, no. 6111) inMoscow ended up with a materialof which about 0.1% was ferro-magnetic. The polymerizationprocess is. however, claimed tohe highly explosive!

Taken overall, these variousattempts to create magneticplastics seem a long way fromany practical application. Never-theless, a cogent theory is nowemerging and it may be - as withsuperconductivity - just a matterof time before some dramaticbreakthrough occurs.

Fastestphotodiode

Workers at the GEC Hirst Re-search Centre and St AndrewsUniversity in Scotland have suc-cessfully fabricated an indiumtin oxide/gallium arsenidephotodiode with a -3dB band-width of 115GHz and an externalquantum efficiency of greaterthan 25%. This, they claim, isnot only the fastest device of itskind but also has the highestresponsivity-bandwidth productever recorded.

The diode employs a trans-parent indium tin oxide layer toform a Schottky barrier and isgrown on a semi -insulating sub-strate to reduce cpacitance. Thisapproach also makes the devicesuch that it can readily be in-corporated into a monolithicchip.

In their paper in ElectronicsLetters (vol.23. no. 10) the au-thors point out that at the high-est frequencies of operation thepackaging of the device is critic-al. The experimental diodes werefitted in a quartz microstrip sub -mount with a millimetric transi-

tion into conventional wave -guide. Testing was accomplishedusing a colliding pulse mode -locked dye laser with a pulseduration of less than 100 fem-toseconds. that is more than tentimes faster than the diode re-sponse.

Copper oxidevalley?

The obscurity of that title isperhaps a measure of the speed atwhich developments are takingplace in a subject that leaves eventhe editors of learned journalsgasping for breath. That subject- superconductivity - has sud-denly taken off in the way semi-conductor physics did in the for-ties and early 1950s.

It began in earnest in April lastyear when two IBM physicistsdiscovered that a compound ofbarium, lanthanum. copper andoxygen would superconduct.that is lose all its electrical resist-ance. at 30 degrees above abso-lute zero. That may sound nogreat achievement when super-conductivity has been known forover 80 years. Until last year.however, the best superconduc-tors were metal alloys thatneeded to be cooled to withinabout 8° of absolute zero.

The race has since hotted up inmore ways than one with a profu-sion of new ceramics that be-come superconducting at liquidnitrogen temperatures, that is77K and above. Paul Chu of theUniversity of Houston and Maw-Kuen Wu of the University ofAlabama of Houston jointly pro-duced a compound only a fewmonths ago that would super -conduct at 98K. Here in Britain,Plessey has demonstrated super-conductivity at around 88°K -the highest temperatures yetannounced by a UK industriallaboratory. Plessey have nowestablished a whole new researchprogramme at Caswell.

The breakthrough in estab-lishing superconductivity atpractical engineering tempera-tures is obviously a great stepforward towards commercial ap-plications. So much so that therace is well and truly on to theultimate goal of a room -temperature superconductorthat requires no cryogenic

equipment. Chemists around theworld are scanning the periodictable, concocting thousands ofnew recipes in a desperateattempt to push the magictemperature ever higher.

All that, however, would havebeen largely academic but foranother discovery announced onMay 10th at the company whereall the excitement began: IBM.Until then virtually all the knownsuperconductors had one seriouspractical snag: they ceased to besuperconducting above a criticalcurrent density. In the best avail-able materials the self -generatedmagnetic field would destroy allsuperconducting propertiesabove a current density of about1000Acm 2. What IBM scientistshave now done is to increase thatlimit by two orders of magnitudeusing a compound of yttrium,barium, copper and oxygen.

Further progress at the prac-tical level is now dependent onthe serendipitous discovery ofnew ceramics and also on bettermethods of fabrication. Many re-searchers believe that a lot willdepend on new methods of grow-ing pure crystals of some of theseexotic materials. Meanwhile in-dustrialists in Britain. the USAand Japan are struggling withpractical ways of applying thenew materials. Some US and UKcompanies have reported successin applying thin films and wiresof superconducting material to avariety of substrates. (See, forexample, July Update, page 762.)

According to Plessey themechanical engineering indus-try could also be a major be-neficiary from the developmentof high -temperature 'hard'superconductors. When used inconjunction with magnets.levitation can be achieved, open-ing up the prospect of friction -free bearings, for example. Smallwonder the call has gone out inthe USA for the creation of a"Copper Oxide Valley" to exploitthe latest discoveries.

A flood of papers has appearedrecently in Nature indicating theextent to which theory is chasingpractice.

A team from the University ofBirmingham for example, hasshown that the electrons makingup the current flowing around asuperconducting ring move inpairs, not singly as they do inroom -temperature metallic conductors. This narrows down tothree the models postulated to

explain superconductivity.Another research group in

Cardiff has been using electronmicroscopy and X-ray diffractiontechniques to examine themolecular structure of some ofthe latest ceramic superconduc-tors. These seem to indicate theexistence of a multi -layeredstructure, which in turn sug-gests that the superconductingmaterials have different electric-al properties in different direc-tions.

An important clue to the ori-gin of high temperature super-conductivity has arisen frommeasurements at the BritishRutherford -Appleton Laboratoryby a team of physicists from theUniversities of Warwick andDurham. By scattering neutronsfrom one of these ceramics basedon lanthanum and barium. theteam has been able to show thatthere are two energies at whichthe crystals most readily absorbenergy from the neutron beam.presumably because vibrationsof the atomic lattice at the ener-gies concerned are coupled withexceptional efficiency to themovement of electrons through-out the system.

It is well understood that allsuperconductivity relies on somekind of coupling between latticevibrations and electron motion.The importance of the newobservations is that they suggesta direct way of studying theconditions favourable for cera-mic superconductivity. Watchthis space....

Note added in proof

Now. Indian physicists at theNational Physical Laboratory inNew Delhi claim to have madethe first room -temperature su-perconductor. In an attempt tocircumvent the delay ofacademic publication, they havereleased brief details of theirachievement. A report (Naturevol. 327 no. 6121) describes aphase of yttrium barium stron-tium coper oxide that is said tobecome superconducting below+26°C. Further details areeagerly awaited.

Research Notes is written byJohn Wilson.


Image localization andinterchannel phase

differenceUsing the wavefront reconstruction approach to predict

image position in stereophonic systems

The use of phase delay in stereophonicsystems as a means of conveying spa-tial impressions is not widespread.

This is probably because subjective experi-ences have shown that a stereophonic imagereproduced by a system having an interchan-nel phase delay is less well defined anddifficult to localize than when interchannelintensity difference is involved 2-14.

The wavefront reconstruction theory instereophony has been developed in a pre-vious paper' and this contribution is in-tended to utilize these principles in provid-ing a new explanation for image localizationin stereophonic sound systems with inter -channel phase difference.

WAVEFRONT RECONSTRUCTION

Consider the stereophonic system geometryin Fig. 1 for speakers having an interchannelphase delay of y radians. The wavefrontgenerated along the x-axis, H(x) is given as'

H(x) =L-exp (jk ri -yl+ -Rexp (jk r2) (1)r1 r2

where L,R are the amplitudes of the left andright channels respectively and K = 27r/X isthe wave constant.

Neglecting the radial amplitude reductionin the divergent wavefronts and assumingspeaker polar diagram variations are smallover the region occupied by the head, wehave

H(x) = L exp(jk ri- y) + R exp (jk r2) (2)

For the listening geometry chosen and fre-quencies to be considered'

ri = S + -x2 + x sin 0 (3)2s

X2r = S + -2- x sin 02s

(4)

Substituting equations (3) and (4) intoequation (2),

F.O. EDEKO

w

H

Fig.l. Stereophonic system geometry.where W=2.3m and H =2m.

x2H(x) = exp Lik(S + F1 I exp (jk x sin 0

-Y) + R exp (-jk x sin 00)). (5)

Neglecting the multiplicative term, as itcarries no directional information,

H(x) = L exp (jkx sin 0-y)

R exp(-jkx sin 00) (6)

Since the system is to provide directionalcues by employing phase delay only, thechannel amplitudes will be assumed equal,so that R=L, and hence

H(x) = L exp(kx sin 00- y) +

L exp(-jkx sin 00)

= 2L exp 1-j -y/211cos (kx sin 0 -

-y/211.

(7)

(8)

If the delay -y = 0, then

H(x) = cos (kx sin 00). (9)

Equation (9) is maximum if kx sin 00 = 0 andthe wavefront is symmetrical about x = 0.

However if -y #0. then the maximumshifts away from the point x = 0. For

equation (8) to be maximum.

kx sin 00- -y/2 = 0.

Therefore

x -1

2 K sin 00

=1 X

2 2-rr sin 00

(10)

From equation (11) it is evident that whenan interchannel phase difference exists be-tween the signals in the left and rightchannels, the reconstructed wavefront inthe region of the listener is displaced to-wards the leading loudspeaker. Dependingon the amount of displacement and thefrequency, the head could be immersed ineither a section of one complete interferencepattern of the wavefront H(x) or in betweentwo interference lobes.

Figure 2 shows the amplitude and phase ofthe wavefront reconstructed by the twoloudspeakers in Fig. 1 along the x-axis over arange of 60cm for a 500Hz signal, when theright loudspeaker leads the left by 90° and180'. Here, H = 200cm, W = 230cm and 00= 30°. The wavefront for the zero delay isalso shown. The computer simulations ofthese wavefronts are based on equation (1)and therefore do not include the later appro-ximations. These graphs show that for adelay of 90°, the wavefront is displaced to theright by 17.15cm and for 180° by 34.3cm asequation (11) would suggest for a 500Hzsignal. In the computer simulations thevelocity of sound in dry air at 20°C was takenas 343 m/s.

Curves C, C' are of particular interest asthey show discontinuity in the phase charac-teristics of the wavefront and a zero of theinterference pattern in the centre. This isexpected when the two speakers are inanti -phase. Under such conditions theperception of a well defined image is highlyunlikely, as the phase characteristics lackthe continous nature normally associatedwith a localized source'.


100

075

050a

025

0

-30 -20 -10 0 10 20 30

180

90

a; 0

-90

-180-,

-30 -20 -10 20 30

Fig.2. Wavefront reconstruction, with f = 500Hz and R -- L.a.a' -amplitude and phase when -y = 0 b,b' -amplitude and phase when -y = 90 c,c' - amplitude and phase when -y - 180

10

07

02

,--'/-8 -6 -4 -2 0 2 6

180

90

11:"

0

-90

180

-8

a

c'd

e'

-6 -2 0 2 4 6

Fig.3. Amplitude and phase of wavefront across the head (500Hz, R= L, right leading)a,a' for -y=0 ; b,b' for -y=30 ; c,c' for -y=60 ; d.d' for -y=90 ; for -y=120 ; for -y=150 ; g,g' for -y = 180

100

0 75

O

050

025

0

-8 -6 -4 -2 0 2 4 6

e

d

b

a

180

90

0

90

-1808 8 6 4 2 0 2 4 6

a'

c'cel:

9'

8

Fig.4. Image localization with interchannel phase delay

In attempting to develop a theory of imagelocalization with interchannel phase differ-ence it is important to examine the form ofwavefront across the head.

IMAGE LOCALIZATION

The wavefront reconstruction theory de-veloped in reference 1 proposed that thewavefront reconstructed by two loudspeak-ers in a region of space occupied by the headof a listener contains the fundamental direc-

tional information of the image. Initially itmight appear that the phase tk(x) of wave -front H(x), equation (8), could provide thedirectional cue in a stereophonic systemwith interchannel phase delay.

However, analysis of equation (8) showsthat the phase slope of H(x) is constant forlow -frequency signals until y is increased toa value where an interference null can besensed by the listener. Therefore, even in thelow -frequency range, where the theory of

phase shift in signals suggest at least that theear/brain combination can still follow thewavefront parameters,' 14 the linear phaseslope of H(x) contains no directional in-formation when only an interchannel phasedifference is present.

There is therefore only one remainingparameter of the wavefront H(x) that canprovide directional information, and this isthe amplitude. To obtain a correct assess-ment of the amplitude of H(x) and its


variations with interchannel phase differ-ence, it is necessary to generate H(x) overthe region head. Figure 3 shows the graphsof computer simulations of the amplitudeand phase of H(x) generated by the twoloudspeakers in Fig. 1 along the x-axis over aregion -xm--5.x <xm occupied by the head fordifferent interchannel phase delays and for a500Hz signal. The software is based onequation (1) and the head width used is 2xm= 14cm.

As expected, the phase slope of the wave -front across the head is zero for all the phasedelays less than 150°, above which dis-continuity in the phase characteristic of thewavefront occurs. The zero phase slopeindicates that the apparent source is locatedat a distant point in front of the listener).There is no indication of image displacementfrom the on -axis position from the phasecharacteristic of the wavefronts.

However, the amplitude of the wavefrontacross the head shows significant variationsfor a given phase delay. This, then, mustprovide directional information. Amplitudedoes not normally provide directional in-formation in a propagation wavefront: this isa role normally associated with the phase.Therefore, the manner in which the ear/brain combination used amplitude variationacross the head to deduce image directionappears to be based on the sound pressurelevels directly present at the two ears. Thismeans that the amplitude values of thewavefront at the points -xm and xm are thecue to directional localization in stereopho-nic systems with interchannel phase differ-ence.

Clark et al derived the followingexpression15

sin a = (IL-R)/(L+ IN sin 0 (14)

where L and R are the average instantaneouspressures at both ears from the left and rightloudspeakers respectively. In wavefront re-construction, the question of crosstalk doesnot arise and therefore the values of L and Rcan be taken as the sound pressures atposition -xm and xm of the reconstructedwavefront H(x). Equation (14) can thereforebe used to predict image location a for lowfrequency signals in a system with inter -channel phase difference.

Equation (14) has been used in conjunc-tion with computer predicted values of theamplitude of the wavefront H(x) at earpositions, -xm and xm to determine imageazimuth angle a for various interchannelphase differences for a 500Hz signal. Thetheoretical curve of image displacementwith interchannel phase difference is shownin Fig. 4.

Several interesting effects are seen here.Firstly, no amount of interchannel phasedifference can completely displace the imageto either of its extreme positions. This is amajor hindrance to the deliberate use ofphase delay in stereophonic sound reproduc-tion. Secondly, small interchannel phasedifferences result in no significant change inimage positions at low frequencies.

The use of equation (14) indicates thatwhen -y = 180°, image position is indeter-minate. This true since L = R. The inter-pretation to this is also evident in the curvesin Fig. 3. When y = 180°. the head is in

-150 -120 -

120

E

-`2- 90

60

30

- 30

-60

90

-120

Theory

Practice

30 60 90 120 150 180

INTERCHANNEL PHASE DIFFERENCE (degrees)

Error marginlby standard deviation)

Fig.5. Amplitude and phase of wavefront across the head (2500Hz, R=L, right leading).The curves have the same meaning as in Fig.3.

contact with two interference lobes of thereconstructed wavefront and the phasevalues of these lobes are actually displacedfrom each other by 180°. This implies that.the sound pressures at the two ears are inanti -phase. This is a most unnatural situa-tion which has no counterpart in normalhearing experience and is generally under-stood to lead to a loss of localization.

At high frequencies, the wavefront H(x)shown in Fig. 5 for a 2500Hz signal, essen-tially also has a constant phase for -y = 0.However, increases in interchannel phasedifference create a shift of the wave patterntowards the leading loudspeaker so that thehead soon becomes immersed in more thanone of the interferences lobes. Image loca-lization in this situation is anticipated to bedifficult as the circumstances are very simi-lar to y > 150° in the 500Hz case. Thereforethe introduction of elementary types ofphase delay into a stereophonic systemwould appear to be unsuitable for highfrequency information.

LISTENING TESTS

Practical tests have been carried out todetermine image positions when interchan-nel phase difference is introduced into thetwo -loud speaker system in Fig. 1. The phasedelays were introduced by using a stereochannel phase control system consisting ofan SQ decoder circuit designed by EMIResearch Laboratories. The tests were car-ried out in an anechoic chamber with areverberation time of less than 0.25 secondsfor all frequencies down to 125Hz. Thesignal used for all subjective tests was 1/3octave pink noise with a central frequency of500Hz produced by a random noise gener-ator in conjunction with a bandpass filter set(Bruel and Kjaer Type 1402 and 1611). Eachloudspeaker cabinet housed a single type 8Punit produced by Goodmans LoudspeakersLimited and ten listeners participated in thetests.

Each subject. occupying the median planeand looking directly towards the stage cen-tre. (Fig. 1) was asked to give the position ofthe sound image as phase delay was gradual-

ly increasd from zero to 180°. The tests wererepeated several times for each participant.The listeners were unaware of the delaysintroduced into the system during the tests.The mean of individual results is shown inFig. 4 as the practical curve of imagedisplacement with interchannel phase delaywith the error margins for the given imageposition shown as vertical bars.

Comparison between theory and practiceindicated that the use of equation (14) topredict image positions provides results thatagree very well with practice for values ofinterchannel phase difference up to 90°. Forphase delays above this value, the theoreticalpredictions begin to increasingly overesti-mate the image positions and the differencebetween theory and practice increases asphase delay approaches 180°. The 90° phasedelay value also marks the stage whenindividual results tend to differ much morewidely as shown by the error bars. It is

interesting to note that during the practicaltests many subjects reported image spreadacross the entire stage width and sometimesbeyond it when interchannel phase differ-ence exceeded 90°.

The increased divergence between theoryand practice when -y > 90° can be attributedto the increased uncertainty of localizationas these images tend to be very wide.

When the two channels were 180° out ofphase. most of the subjects could not local-ize the image and others reported an in -the -head image. This is what was anticipatedfrom the theoretical analysis.

CONCLUSIONS

The wavefront reconstruction theory pro-vides a means of assessing the performanceof stereophonic systems with interchannelphase differences. Image position variations with interchan-nel phase delay can be predicted using thestereo sine law at low frequencies. Providedthe L and R in the expression are taken assignal values in left and right ear respectivelyfor values of -y 90°. No amount of interchannel phase delaycan completely displace the image to either


the left or right speaker. This is a limitationin the use of phase delays in stereo systemsas the stage width is greatly contracted. For the reproduction of a reasonably welldefined image, interchannel phase delayst )uld not exceed 90° and the use of phasedelays should be limited to low -frequencysignals. Phase delays greater than 90° wouldresult in the reproduction of very broadimages.

References1. Benneth, J.C., Barker, K., Edeko, F.O.: A newapproach to the assessment of stereophonic soundsystem performance, Journal of Audio Engineer-ing Society. vol. 33, pp. 314 -321, May 1985.2. de Boer, K.: The Formation of StereophonicImages, Phillips Technical Review, vol. 8, no.2,1946, pp. 51- 56.3. de Boer, K.: Stereophonic Sound ReproductionPhillips Technical Review, vol. 5, 1940. pp. 107 -114.4. Leakey, D.M.: Some Acoustic and Psychologicalaspects of Multichannel sound systems, Reprintfrom Journal of the British Sound RecordingAssociation, vol 60, 1960, pp. 118 -125.5. Leakey, D.M. Some Measurement of the Effectof Interchannel Intensity and Time Differences inTwo Channel Sound Systems, J.A.S.A.. vol. 31,1959.6. Bauer, B.B.: Phasor Analysis of SomeStereophonic Phenomena, J.A.S.A.. vol. 33, 1961,pp. 1136- 1539.7. Snow, W.B.: Effect of Arrival Time onStereophonic Localization, JA.SA.. vol. 26. 1954,pp. 1701 - 1074.8. Hanson, R.L.: Sound Localization. J.A.SA., vol.31, 1959, pp. 830.9. Leakey, D.M.: Further Effects Produced by TwoLoudspeakers in Echo Free Conditions, J.A.SA.,vol. 29,1957, pp. 966.10. Leakey, D.M.: Influence of Noise Upon theEquivalence of Intensity Difference and SmallTime Delays in Two -Loudspeakers Systems,J.A.S.A., vol. 29, 1957, pp. 284 - 286.11. Hanson, R.L., Kock. W.E.: Interesting EffectProduced by Two Loudspeakers under Free SpaceConditions. J.A.S.A., vol. 29,1957, pp. 145.12. Matsudaira, T., Fukami, T.: Phase Differenceand Sound Image Location. Journal of AudioEngineering Society, vol. 21, 1973, pp. 792 - 797.13. Mcfadden, D.. Pasanen. E.G.: Binaural detec-tion at high frequencies with time -delayed wave-forms, J.A.SA.. vol. 63,1978, pp. 1120 - 1130.14. Blauert, J.: Spatial Hearing MIT Press, 1983.15. Clark, H.M., Dutton, G.F.. Vanderlyn, P.B.:The 'Stereosonic' Recording and ReproductionSystem, The Proceedings of The Institution ofElectrical Engineers, vol. 104, Part B, 1957, pp.417- 430.

F.O. Edeko obtained an M.Sc degree inSound Electronics from Leningrad's Insti-tute of Motion Picture Engineers in 1979and in 1985 obtained a Ph.D degree fromSheffield University. He is currently a Lec-turer with the University of Benin, Nigeria.His research interest involves the wavefrontanalysis of stereophonic systems - an areawhere he has many publications to hiscredit. He is an Associate Member of theInstitution of Electrical Engineers.

Top brains' exodusThe brain drain is really about the loss of skillsrather than numbers: senior researchers going

abroad are not being replaced.

Britain's best brains are leaving thecountry because they are fed up withbeing mistreated, poorly paid and

having their research grants cut year afteryear. This is the verdict of a report pub-lished this month by the Royal Society onthe brain drain.

According to the report "the effects ofcurrent policies are likely to be seen inyears to come as one of the most disastrousexamples of this country's post -Victorianreluctance to invest in the future but ratherto gobble up the present."

The Royal Society's report is the firstquantitative study into the brain drain formany years. Questionnaires were sent to750 heads of research departments inuniversities, government laboratories andindustry. They were asked who had left orentered the UK between 1975 and 1985 andthe reasons why they moved.

"The continued emigration of talentedscientists and engineers from the UK is amatter for concern and undoubtedly repre-sents a loss of research talent," said SirDavid Smith, Secretary of the RoyalSociety.

The survey found that the number ofscientists leaving the country to be relative-ly small, about 931. And this outflow waslargely made up by the influx of 685scientists from overseas. On the face of itseems hardly like a dire situation at all.

But the study reveals those leaving thecountry are more senior, more experiencedand leaving long-term posts which will bedifficult to fill. The scientists arriving fromoverseas are less experienced and 80% ofthem have taken short-term posts of lessthan three years after which they may leaveand take knowledge earned here with them.According to the report we are seeing theemigration of "a few key staff which ishaving a debilitating effect of UK research."

Over 68% of respondents from universi-ties said that it was difficult to find replace-ments for those that left. In electricalengineering departments the task wasfound to be even harder with 90% ofrespondents saying posts were difficult tofill.

Of those leaving industry to go overseas19% were senior people of professor orreader rank. Of those coming from overseasto fill their shoes only 4% were of seniorrank.

The report concentrates on five keyresearch areas: electronics, physics, che-mistry, biochemistry and earth sciences.Electronics is singled out as a particulararea for concern. Over 70% of electronicsengineers leaving Britain left senior long-term posts which will be difficult to fill.

"The problem in electronics is how tofind people prepared to work in universi-ties" said one of the respondents. "We usedto rely on overseas Ph.Ds, but this supply is

drying up as other countries offer moreattractive prospects. The total number ofBritish graduates is far too low to satisfydemand."

The vast majority of disgruntled electro-nics engineers, over 70%, go -to the US andCanada. Their next most favourite haven isWestern Europe, with 16% going there.

The main reason for all types of resear-chers both in industry and universitiesleaving the country is because they feeltheir careers in the UK are limited andbelieve that better opportunities are to behad abroad. The next most popular reasonfor leaving is a desire to widen experience,followed by rates of pay. University resear-chers put level of equipment and researchfacilities fourth on their list, just behindpay, whereas those in industry put researchfacilities sixth on their list.

The most worrying part of the report isthe section on future trends. Already thereare fewer young talented students choosingto carry on research in the UK than therewere 25 years ago. Over half of the 314 mostrecent Ph.Ds chose to continue theircareers overseas.

"In the present climate it is difficult tounderstand why any new Ph.D studentshould wish to enter scientific research inthis country" comments one respondent"Scientists have low status relative to otherprofessional groups and are relatively poor-ly paid in comparison to other professionalgroups."

Many replying to the questionnaires saidthe number of overseas offers being made tothem were increasing. Physicists, electro-nics and computing people particularlywere being seduced to the States with plumstar wars contacts. Worse still they said thatgiven the current climate of financial res-traint in Britain they were more likely toaccept such offers.

The Brain drain is about the loss of skills rather thannumbers of people. Senior researchers leaving Britishindustry (19%) are not being replaced by thosecoming from overseas.

Seniority of from UK to UKemigrants Number per cent Number per centSenior 22 19 1 4Middle 40 35 12 43Junior 45 39 6 21Others 7 6 9 32

Senior professors. readers. Middle: senior lecturers. Junior:lecturers. assistants. Others. research fellows.

Reasons for leaving UniversityNumber per cent

Rates of pay 95 14Status of science 33 5Scientific vigour 44 6Standard of living 41 6Political climate 2 21Facilities 92 13Personal 30 4Career opportunities 125 18Carrer limitations 122 18Working conditions 14 2Widen experience 86 12Others 7 1


-FCIRCUIT IDEASDouble-sideband s.c. demodulator for 455kHzThis suppressed -earner double-sidehand de-tector - an enhanced version of the p.I.I.detector shown on page 35 of the Januaryissue - is suitable for an i.f. of 455kHz.

Two comparators at the input form theduty -shaper block. The first comparator actsas a high -gain half -wave rectifier. Resistor/capacitor loading on its open -collector out-put turns the signal into triangular pulseswith a long rise time and short fall time. Byvarying the potentiometer at the secondcomparator input, squared pulses drivingthe p.I.I. can be adjusted for 25% duty cycle.

Free -running frequency of the v.c.o. mustbe 910kHz. From the data sheet, f= VitiRCwhere value R=10011 but my circuit usedR=13011 which produces a C,, of 680pF.Output of the v.c.o. is delayed and invertedby a third comparator, the potentiometer atits input being adjusted for maximum audiooutput.

As in the previous detector, the synchro-nized signal is taken from a JK bistable i.c.driving an analogue gate.Kerim FahmeAleppoSyria

220p

100k

The

22k

50k

3.c3 p

64148R12 13 '0

NE564,

g(t) ti

4066

Oscilloscope

3k3 3k 3

T470 p T470 p

402-

T

1 k <9

.g(t) cos 2nftVdcf .455 kHz

gOl=cD and Ig011ma4.2V

I g(fIlmax<Vdc. 2 5V

Electromechanical d.c. AhAn ampere -hour meter is useful for deter-mining a battery's state of charge providedthat battery age and temperature are takeninto account. This meter is relatively inex-pensive.

The shunt (not shown) is selected to drop100mV at full-scale load current which withthe values shown is 500A. Output of the x 10

Shuntin

12 V -9-

10k 1%

amplifier feeds a voltage -to -frequency con-verter, divider and pulse stretcher.

At 500A converter frequency is 2275.56Hzand frequency at the divider output is

0.1388Hz (i.e. 500/3600). which is slowenough for the electromechanical counter.Pulses are stretched to suit the counterusing a simple timer i.c.

\>.....0p -amp offset

.> 100k

1n

100k 1*/.

Frequency adjust

10Ck

12V

Converteroffset 470k

10k6

14

9400

10k

10

1M 1%

7

1

100

stagenary

To calibrate the meter, connect the shuntinput terminals together and adjust theop -amp offset potentiometer for OV at theop -amp output, pin 6. Still with the inputshorted, but with the frequency -adjustpotentiometer set to maximum, set theconverter offset potentiometer so that pin 10of the v -to -f converter is 0Hz. Finally with100mV on the shunt input set the frequency -adjust potentiometer so that frequency atoutput of the converter is 2275.56Hz.

used an electromechanical counter be-cause it inherently provides non-volatilememory but the circuit can easily be mod-ified to suit other counters and full-scalereadings: v.c.o. frequency must not exceed10kHz.Dennis EichenbergParma HeightsOhio

6

in

10r. 9--

""'100n

480

3VA eie tro--imechanical

counter

10n

2N3906

ELECTRONICS & WIRELESS WORLD so3

FCIRCUIT IDEASSingle -referencewindow comparatorTwo open -collector comparators from aquad i.c. form a window comparator needingonly one reference voltage, which is particu-larly advantageous in variable -thresholdapplications.

When input voltage is within the window,outputs of both comparators are off andoutput is positive. Comparator output -sinkcapability is 16mA so leds or small relays canbe driven directly.

Addition of feedback resistors and diodesshown dotted introduces hysteresis that canbe varied without altering the trip level.Modifications shown in the second circuitprovide high. low and pass indications.S. MurugesanISRO Satellite CentreBangalore, India

Auto -callingtwo-way telephonesThese circuits are for a simple automaticcaller using two surplus BT706 telephones.One circuit is a simple d.c. driven oscillatorproviding d.c. ring signal for intercom -typeapplications.

Switching for a simple two-way telephonesystem with automatic calling is shown inthe second circuit. When the caller's handsetis lifted, the other telephone rings.

In the switching circuit, two relay coilsand two impedances form a central battery -type telephone -transmission bridge. Whenthe first station's handset is lifted, a loop isextended causing relay RLI to operate. Alter-nating ring current passes to the second

station through contacts RI,21, and RLIncausing the second station's bell to ring.

When station two answers, relay RI,disconnects the station's a.c. ring currentand contact RL2a operates to prevent ringcurrent from reaching station one. Capaci-tors C1,9 link the two loops together, com-pleting the audio path.

N. Cook -AbbottIpswichSuffolk

8 red ESTI BT4

0\1'7

r0/1-6t 01\ T5

14 010

C.--+1.

01.17

T19

240V

18V

18'

116BT2

118 wh

R1.1 RL2 0:RL2a RL1b

RLia RL2b LX

180 820

17000u T000y

IC) Remove lot, T5 and Ton each telephone

0 Connect tonic T1) T,BIce two wee

lennonatan

0 Remove Mks B to CXand T4 to T, lotseparate bell/call

system Id c ono)

Connect CX to T4 andd.ode 05 For

ac nnong D5 esreplaced by a Ink

BT3 BT6

Z=100 oZ=100[Li412

706 telephone

18

Remove links14/5E716m

1k

T9

76

T5

14

116

6_1).117

119 T18

Bell

To oscillatorunit

Bell

20


,CIRCUIT IDEASA -to -D converterusing ditherAdding dither, i.e. an impressed signal ataround one 1.s.b., to the input of an a -to -dconverter allows resolution to be improvedby averaging. This circuit was developed foran Apple II with 6821 interface chip.

Band -limited noise from a noisy zenerdiode - the reverse -biased e -b junction of aBC184 - provides the dither signal. Noisedither prevents possible aliasing with thesampling or signal frequencies.

Resolution of the 8703 converter is eightbits but as the graphs show, dither producesconsiderable improvements in resolutionwith slowly -varying signals.C.P. OrmistonLutonBedfordshire

Bridge amplifierwith common driverReplacing the differential pair commonlyfound in bridge circuits with an input tripleof transistors reduces the number of compo-nents required.

Extra transistor Tr3 provides common -mode feedback which, when R4= R5, keepsdifferential outputs X and Y symmetricalabout ground. Emitter resistor ratios arechosen as RI =122=2R3 so that differentialgain (times Va-Vb) is equal to single -endedgain (times Vd.

Mirrors multiply collector currents byfactor M and drive the active current -sourceloads as in a conventional amplifier (othercurrent gain stages could replace the mir-rors). Unity -gain voltage followers providecurrent gain. In the feedback network. Rfwith Rh= R,nIIR1gives voltage gain of approx-imately 2Rf/R,n when suitable compensationcomponents (not shown) are used.

In some respects. this driver is similar toone described by Linsley Hood (80-100W

26 way -12V

ribbon connecto I -5V

O 0 0 0 0 0 0 0

O 0 0 0 0 0 0 O 0

C82

CA 2

ex100k

10V fullscale

12V5V

1

4078

14

1MLaw, High

1M In1

Input resistance

LF351

Use decoupling to;round on pins 17109

270k

777j .127143

"\.A.7

10

15C 0 //c.

10k-471-.5vBC 184

Zero crossing

-12V1k2 6V2 220).

1 -/VVI/V\.--"ThVs0-0 7Sk 30k 15k 15k 7k 5 3x

+12V Gainx5 x10 x20 S 1n .1.

tOk

-6.- -12 V

mosfet audio amplifier WW July 1982, p65).In that design the current source of astandard -input differential pair was mod-ulated by d.c. feedback from one of thedifferential outputs and overall negative

Vin

feedback was taken from the other. Mydifferential -triple configuration is simplerand. I believe, more elegant.T.C. ThomasManchester

Tr

Ry

IV5

1

L.

2

/cc

-v«

M10 /4


CIRCUIT IDEASPower-fet shunt regulator for 50volt outputModern power lets, with their nigh g, andinput impedance, make simple but effectiveshunt regulators. Load regulation of thisexample for shunting up to 100mA at 50V is0.1%; nominal d.c. input is 60V.

Since g, of an IRF9520 is typically 0.9Sonly very small changes in the bipolarcollector, and hence zener current, occur asthe load is varied. Off load. 100mA letcurrent results in dissipation of around 5Wso a heat sink is needed.G.C. LovedayUckfieldEast Sussex

Instrument supplySupply requirements of a portable instru-ment were ± 12V at 500mA each and 5V at 2Abut only a -48 to -52V telephony supplywas available. An LT1070 in flyback modewith a 50W core and post regulators providesthese supplies with excellent performance.

Windings should he bifilar: for the threeoutput voltages above. Li has n turns, L., has1.8n turns and L3 has 2.8n turns. Any fixedor variable regulator can be used, providedthat there is a 3V or more drop across it. Fastswitching diodes and a V. of at least 60V forthe p -n -p level -shifting transistor areneeded.G.R. NimmoBasingstokeHampshire

Unregulatedinput

200)

0-0-0

50mH

L3

1000p

000 -06 -Le

n S

LT 1070

-9,2

+ 50V100mA

Ck2

1k2

... 1000p

Mains control interface gives 256 phase stepsCrider control ul latched eight -hit micro-processor ports, counters and comparatorsproduce 39ps pulses whose phase can bevaried relative to the mains cycle. Havingphase delay proportional to the binary valueon the output port, these pulses are ideal fortriac driving through opto-isolators inmains control applications.

Outputport 1

Downcounter

Reset2ck t,Dpc

4

2S trTsc

Reset

Pulse delay to give Taste- di n

Triac drive 1*

A=B

A Comparator

B

100Hz

Mainssync

Mains hall cycles are split into 256 steps bya down counter and a comparator for eachtriac compares the output port value withthat of the counter. An FF output -port valuefires the triac near the start of the mainshalf -cycle whereas zero leaves the triac off.

Delaying phase of the 100Hz mains signaleither under computer control or using a

12V 0 SA

12V 05A

5V 2A

monostable multivibrator. produces a

master -dim function. If comparator A<Band A>B outputs are available they couldprovide a p.w.m. output.

Equivalent software flow suitable for a

high-speed digital signal processor is shown.Graham HardyNottingham

Triac drive 2!..

-0 -Delay100Hz --sit-Interruptsync

Outputport 2

A Comparator

B

Othercomparators

Set counter to FF

Read port 1Compare with counter

Output result to tnac 1

*Ina( drives are opto ,solaced -39ps(- 2ps perport for

Read port 2Compare with counter

16 ports) Output result to triac 2

Decrement counter


Simple voice -controlfor the IBM PC

Taking advantage of a useful, undocumentedone -bit i/o line to switch keyboard mode with simple 'stop'

'go' commands.

My word processor (New Word-whichis just like Wordstar) requires exten-sive use of the control key in com-

bination with other keys on the left hand sideof the keyboard, which greatly slows typing.I cured this by writing a program to allow afootswitch to function as an additional con-trol key. Later I replaced the switch with asimple voice recognizer that changes thekeyboard action according to the last com-mand spoken.

This simple circuitry allows four commoni.cs to detect and remember the differencebetween two spoken words. Saying "stop"puts my word processing keyboard into thecontrol mode, where it remains until I say"go" and resume normal alphabetic typing.

The switch -responding program has to be

the background while the word processor is

B.J. SOKOL

also running. As PC -DOS does not allowproper multi -tasking, my program uses thetimer tick interrupt implemented by DOS onIBM PCs and compatibles. Eighteen timeseach second this hardware interrupt passescontrol to a routine which is pointed to bythe vector stored at memory locations00070-73. PC -DOS initially points this vec-tor to an interrupt return instruction. Myprogram changes the vector to point to codethat scans a switch, reacts, and then returnscontrol to the main program flow.

When it is first run my program musttherefore change the interrupt vector topoint to code in memory, and it must installthis code in memory in a way that preservesit against overwriting after the program isterminated. PC -DOS provides functions todo these tasks, allowing the creation of t.s.rs.or 'terminate and stay resident' programs.

This code reads a spare line on the printer port. In use, it isprotected by DOS.

PAIS

POPT

1.1/4L

DATA

segment AT DODO

oto 406h Word at 0000 0400 an shsolute mem,label wordorg 412h .,,tIrgonlInlabel byte stores leybosrd flag- bit 2 is cntr:ends

..'.."'assume cs CODE se STACY

Executable initialising code

change eeeeee upt vector

R[ ]SIEn**** t° '""" "" "°'of feet MNIDLER... us

3Then free memory 411ncated for copy of DOS ***** cement

pap:: 00 .,-rn;°1," -^t toES 101.2Chl segment addles. of nnnnn dod environment

nor AN 4.1 prepare to deallocate memoryint an do it

KANIDEPpushPosh

IAD

acePirhPoP

int

PoPPoPPDPPoP\rat

PPFSSi cmp1n

or

100T db[00F ends

STACK

nnnnnn TSR

AN Illi prepare to terminate 6 *lay residentAL 01 dummy return code7%20h resid::tjartuIre men paragraphs

Resident ,.tern* pt handltng code

OXgsCX

Al 0000 .to PORT

D X DS PORTD X /Al. DX

:ILIEWEDCS FOOT 1

ENOCS FOOT 0DS FLAG 0.111.CX 6101Ih

AM 1

10h *aging int ter rLpt

CX restore registets and fini

segmentisme s STACKdw duel '1end

end START

me. for bat I lain 14 m*rttedlblf closed

"erc:IOnd clittnee caror shape

tiv creak last stateeither end since no /hang.

or if -retitled 1 tit e lest scanuse ...deb to set cnt r : bitand change cursor shape

using in ttttt Pt

atores al.@ of ewitch

Foot = 0

Set CTRLbit off

The program is called vox.exe. It is assem-bled from the source code shown in thelisting, and should be executed before anyother t.s.r. is installed (usually from anAUTOEXEC BAT file). Some t.s.rs interact badly,but i.ox seems to work happily together withthe Turbo Lightning spelling checker t.s.r.which I normally use, and also with thepopular t.s.r. Sidekick.

Once installed vox responds to switchclosure by changing the cursor shape as wellas changing the key meanings, so that youcan see which of the two possible states thekeyboard is in. It also allows normal use ofthe control key: the program flow thatachieves this is shown in Fig.l.

There is a way to connect a foot switch orvoice switch to a p.c. (IBM or near -clone)without the need for any special hardware toplug into the i/o channel or backplane. This

Timer tickinterrupt

Find port

Recd bit 1

Set thincursor

Foot =1

Set CTRLbit on

Set fatcursor

Interruptreturn

Fig.1. In this resident code flow, the flag Foot controls thefunction of the Ctrl key. Initially it is set to O.


is because a spare line is available on theparallel printer port. It is only necessary toconnect two wires to the p.c. end of a parallelprinter cable, one to pin 14 of the DB25conneector and the other to an earthed pin(any one, 18 to 25). Be warned, though, thatsome multifunction cards do not connectpin 14 at all, although IBM printer cards, andmonochrome -video -plus -printer cards, do.

The data state at pin 14 of the DB25appears in the PC in inverted form at bit 1 ofthe i/o port numbered two above the baseaddress of the parallel printer interface. Thebase address can always be found at absolutememory location 00408 (for the first printer,LPT1: - any other printers follow at 40A,40C, 40E). By the way, the bit in question

Pre -amplifier

FilterThreshold

andzero crossing

Thresholdond

zero crossing

can be written as well as read by i/o instruc-tions, so this trick makes available an un-documented one -bit interface to any PC.

The speech recognition circuit can re-place the open -circuit switch connected topin 14. It distinguishes the words "stop" and"go" or "yes" and "no" to produce two logicstates. These word pairs are differentiated bymeans of frequencies and the duration ofspeech energy. Sibilant sounds in "stop" or"yes" or "system" produce energy in the highfrequency portion of the spectrum. Theproblem is that this energy is noise diffusedacross a wide band and has a very lowamplitude compared with the average ampli-tude of speech. Moreover the transient por-tion of vowel sounds, and some vowels'

RetriggerableCounter

I 4 one-shot/

P

Clock

Outputdominated

by -set"Set

Latch

Reset

Fig.2. Essentially the function of the circuit is to differentiate between words containingsibliants and those without

;/ 150n

>100k

Filter

0lm

.12k7 15k 15k

BC 109

5rn V

56k

3 k

4093

4040

390k

'"100n it

I'-MLrJ

10ms

2k7

higher overtones (especially the higher so-called formants of the sound "ee" whenpronounced by women or children), putconsiderable energy into the sibilant region.

The problem is worse if we require thecircuit to work with cheap electret mic-rophones that lose part of the high end of thespectrum and also have no noise -cancellingcapabilities to eliminate background tran-sients. Nevertheless a reliable circuit usingsuch microphones is possible (Fig.2).

Sibilant energy is first emphasized by amulti -pole high-pass filter cutting off below5.5kHz. The output of the filter is passed to alevel detector and digitizer, and then to afrequency counter. The count is assessedover 10ms to eliminate the effect of brieftransients passing through the filter.

The counter is realized very simply be-cause only a logic output indicating an inputstream at greater than some frequency isrequired; the high -order bit of the count isused to indicate whether a sufficient counthas been reached. This bit actuates a retrig-gerable pulse stretcher and so any overrun ofthe count (by more than a factor of two) thatwould reset the high bit is ignored, eliminat-ing the need for an input inhibit at max-imum count; and the effect of vowel energyfollowing sibilant detection is locked out fora certain period after the sibilance subsides.

5V

1M

BC107

"''470

Fig.3. In this speech detector the one -bit output duplicates the action of the Control key: itcould equally well duplicate the Shift or Alt Keys.

3

:n

33e

33k

24k

33k

33k

Fig.4. Filter for thecircuit of Fig.3. Allop -amps are 074 types.

104k

33k

206e8

yK. 206r.8

19k4

33k

175k

Out

Pin14

Footswitch

074 powered by .! 5V

40408,4093 5V

The second action is achieved because themonostable drives a set -reset flip-floparranged to favour the sibilant state: if avowel and a sibilant are detected simul-taneously, the sibilant wins.

Circuitry is minimal (Fig.3). The voweldetector uses the base junction of a transis-tor as a level detector. The output of this odddetector is properly digitized thanks to theSchmitt trigger action of the 4093 inputs,and the cross -coupled set -reset latch ignoresthe irregular stream of pulses produced by it.

Another transistor, a resistor and capaci-tor. plus another section of the 4093 verysimply produce the required retriggerablemonostable action. Finally, an elementarylevel shifter composed of a diode and resistorplus the Schmitt action at the input of the4040 counter eases the design of the sibilantenergy digitizer.

Jerry Sokol started a design consultancy inthe U.S. in the early 1960s. He also lecturesin English renaissance literature at Gold-smiths'College, London University.


All about curls and divsA gentle introduction for those who missed out on vectorfield theory as undergraduates, which should help dispel

the mystery surrounding vector field equations.

In an earlier discussion' I covered a littleof James Clerk Maxwell's remarkablework on electric and magnetic fields and

how he related his model to the propagationof light. After reading it, a student friendsoon took me to task and said that althoughhe had learned that James himself hadconjured up the terms curl and convergence(nowadays oppositely directed as diverg-ence), he still couldn't see the wood for thetrees. "You see", he went on, "I'm none thewiser about what the curl and div - to saynothing of grad - really mean" In answer Isaid, "It is all to do with vector and scalarfields; they are the basis." "Oh no!" hereplied, "we had a ghastly maths courseabout them. That course is still a poor one,you know." I knew this observation could bevery true, as maths teaching is now some-what grim in our educational system. No-body seems to care enough about it.

But what of curl and so on? You do needsome knowledge of differential and integralcalculus, but most 0 -level syllabuses con-tain a little about these topics now, so it isn'ttoo frightening. The only little bit extra youneed is a nodding aquaintance with partialdifferential coefficients2.

SCALARS AND VECTORS

Nearly everyone knows the difference be-tween a scalar quantity and a directednumber, or vector. You would get some

zdrag ,

11

\ / // / //I /v n \ /

\ A I I, Iso-surface'

I,1/\\ f'\\/ \I

Fig.1....invisible surfaces all over your room.

'JOULES WATT'

raised eyebrows if in the grocers someoneasked for, "A pound of tomatoes, due Northplease. Or if elsewhere you heard, "I drovemy car at 60 mile/h - quite exhilarating." Towhich a comment was, "Where to?" and youheard the reply, "Oh, anywhere, I just closemy eyes and go - it's just the speed thatmatters!"

Imagine we have entered a region ofspace; a room, a pond or river, a box in alaboratory - anywhere, to make measure-ments on some quantity permeating theregion. Typically, your measurements mightapply to a draught in a room, or thetemperature distribution, or the water -flowpattern in the river, or again, the electro-magnetic radiation in the box. We call anysuch region a field. It might be small, likethe box, bounded by walls of some sort. Itcould be vast with undefined boundaries, or"go off to infinity" as a mathematician mightsay.

As an example, consider measuring thetemperature in the room. Point by point werecord the thermometer reading. Figure 1shows what might be happening. Eventuallythe data would apply to the whole volume.We don't say at some position that it is "22°CEast of North," or any such thing. Thetemperature is a scalar quality and the wholedistribution of our measuring pointsthroughout the volume is the appropriatescalar field for this measurement.

Fig.2....crawling about with a vector field roaring about your ears.,

The points are distanced apart and weinterpolate in between, so that we imaginethe field smoothly varying around the re-gion. In fact, the same conditions of 'con-tinuity' and 'differentiability' that interestmathematicians regarding other functionsapply here also.

Light a candle. With it we can nowinvestigate .the cold draught cutting acrossour feet. The candle flame bends over point-ing to where the draught is going. We judgehow strong the draught is by noting theguttering of the flame. So as we crawl aboutthe room, we end up with some idea of howstrong the draughts are and the directions inwhich they are blowing. Plotting all this out,point by point yields the vector field of thedraught distribution, as Fig.2 shows.

You can think of this type of field as aroom full of lines, some crammed togetherindicating high intensity regions, otherswidely spaced in the weaker regions. They allstream along in the various directions of'flow' that meander from point to point.These imaginary lines are the stream lines offlux in such dynamic vector fields as fluidflow systems. Victorian river and estuaryexplorers had a fine old time plotting sourcesand flows. Modern wind -tunnel technolo-gists engage in the same practices.

James Clerk Maxwell appreciated MichaelFaraday's genius in visualizing the 'fieldlines' in the regions around electrically

`Vector (draught)streamlines

ELECTRONICS & WIRELESS WORLD S09

charged bodies and around magnets. And asI discussed earlier', the whole edifice of thehydrodynamicists with their 'sources andsinks', 'fluxes', 'pressure gradients' (or'forces'), 'stream functions' and so on, ar-rived on the electromagnetic field scene withvery few changes in terminology. We stillhave sources and sinks, and fluxes -although no actual flows of electric ormagnetic field occur. The one real "flux"situation in our subject is the vector field ofconduction current flow.

GRADIENT

If we return to measure the (scalar) tempera-ture distribution all over our imaginaryroom. we soon discover sets of points atwhich the thermometer reads the sametemperature. We find these points lie on a'surface' which we might therefore call anisothermal surface, which Fig.1 also shows.The iso-surfaces cannot intersect anywhere,or we would have the impossible situation oftwo different 'constant' values at the samepoint.

These constant surfaces characterize sca-lar fields. Another observation soon showsthat the temperature changes most rapidlywhen we move off the isotherm at rightangles to it. We could quickly plot all the'streamlines' of greatest rate of change oftemperature. and find they all cross theiso-surfaces at right angles. A vector fieldresults from all this plotting. We have foundthe gradient of the scalar field. We write it as

grade = A

Alternatively as

V tp = A.

Here pp is the scalar field point function. Inother words, it is the magnitude of the fieldqualtity measured point by point. A is thederived vector field distribution point bypoint. As you now see, we visualize A asimaginary stream lines pervading the entireregion where the gradient of the scalar fieldexists. Remember, they arise from the direc-tions of the greatest rate of change in thescalar field.

V is the Hamiltonian operator, which hashad a chequered history in all the names thatpeople have proposed for it. 'Nabla' was one,another is 'del'. It arose in the theory ofquaternions that Hamilton and Tait were sokeen on last century. Quaternions have diedout because not much use for them arose inapplied mathematics3. As a vector operatorV has proved very useful. The main thing toremember about it is that it has all theproperties of a vector, but also differentiateswith respect to length, any variables uponwhich it operates on its right hand side.

VECTOR FIELDS: DIV

Vector stream lines might arise on somesource and disappear at a sink. Think of thedraught blowing in a room. Streams of coldair might be found issuing out of the keyholeand under the door - and disappearing, say,up the chimney. We could imagine a surfacesurrounding a 'source'. count the number ofstream lines coming out of it and subtractany going into it. The result measures the

Streomlines of AThe totalequals DIVA

Surfacecontaining

unit volume

Fig.3. The excess of vector field linesissuing out of unit volume at a point in afield over those going in, is the measure ofdivergence.

Fig.4. Completelyclosed streamlines still means avector field - butone in which thereare no sources orsinks.

Curl 9

Unit are:

Fig.5. The measure (per unit area) of thecirculation of the field lines at a point is thecurl of the vector.

41,

strength of the source of the vector. Thegenerating source 'material' is a scalar. Itmight be concentrated at definite locationsor distributed around the region as a scalarfield. Measuring the vector field lines gener-ated in this way coming from unit volume atany point is called finding the divergence ofthe vector field. and it is written

divD = p or V .D = p

where D is the vector'and p is the sourcescalar quantity. V is the differential vectoroperator again. V .D measures how rapidlyD is appearing per unit volume at a point, inother words how concentrated the generat-ing source quantity is there.

VECTOR FIELDS: CURL

Vectors can possess a different property. Asecond vector field can be derived from thefirst one. Imagine we are in the draughtyroom again. All the keyholes, fireplaces,door surrounds might be well and trulyblocked, yet here is a huge draught roaringpast our ears. Then we notice the large fanwhirling the air round and round. We soonrealise that the flow lines form completeclosed paths or loops, like those in Fig.4.They do not start and stop on any sources.Further investigations show that there arelittle closed flow loops distributed all overthe points of the region. At some points thecirculation whirls vigorously, at othersrather weakly - or hardly at all. These

vortices rotate in planes whose orientationsvary from point to point.

We can draw an imaginary line throughevery little plane loop according to thestrength of the vortex per unit area at eachlocation. The lines point away at right anglesto the circulation planes in such directionsthat the rotations go round them in thesense of a corkscrew, as Fig.5 shows. Joiningall these new lines point by point givesanother vector field. This is the curl of thefirst one. It measures the vorticity per unitarea:

curIB=A or V x B =A.

You might notice the significant resultabout the curl airising from this discussion.A and B are always at right angles every-where.

The divergence operation gives a scalarfield from a vector, whereas the curl givesanother vector field from a vector. Thegeneral vector field consists of combinationsof these two extremes. Separating the gener-al field into its 'curl' part and its 'div' part iscalled Helmholtz's Theorem.

The main point I make now is that theequivalent to scalar products in the vectorfield situation is the div operation, while thecurl is a vector product operation. Themeanings of div and curl might now appear alittle less daunting.

You might have already noticed that ascurl fields start and end on themselves, youcan never have a divergence of such a field.


Vectors

A -B ABcos e.work done

We can write this as

div curIA

Again, div fields, and grad of scalar fieldsnever have lines that loop round and end onthemselves. This mean that

curl divD = 0

or curl gradp = 0

always.

VECTOR DIFFERENTIAL OPERATOR

The vector differential operator V , is admit-tedly hard to visualize. It obeys the vectorrules for products etc., but also it is as wehave seen, a differential operator and workson variable and functions 'to the right'. Inother words, V A is not the same as A. V ,where in the second expression V is lookingfor something to operate on to the right of it.V might be asked to operate on (that is,differentiate) a product - such as V .IA x B)or V x (Ax B). The rule for differentiating aproduct has to be obeyed and the rules forscalar and vector product expansion at thesame time. For examples

curl curIA = grad divA - V2A

where V2 is a second -order or doubledifferentiation. Or again

div(E x H) = H.curIE - E.curIH

Both these expansions give valuable andconcise results and descriptions in themathematical modelling of the electric andmagnetic vector field distributions thatoccur in e.m. theory.

CONCISE NOTATION

The value of vector notation lies in itsconciseness. A number of results, like theabove two, give a succinct view of what isgoing on. Nevertheless, we usually return toold Rene Descartes 'scaffolding' - the x. y andz axes, when doing real life problems. Thisapplies especially to engineering situationswith rectangular symmetry, for examplerectangular wavegu ides. If the waveguide isa circular one, the coordinates might be thecylindrical set r. 0. z. If an aerial is radiatinginto a sphere, we might choose the sphericalcoordinates r. 0, p.

All of these coordinate systems have axeswhich are at right angles to each other at any

Fig.6. Work doneby the componentof a force shows atypical example ofthe scalar productof two vectors.

Fig.7.The turning force at the end of anarm - like that on the corkscrew shown.constitutes an example of the vector pro-duct.

Fig.8. The cartesian axes form the mostcommon 'scaffolding' upon which we erectthe dimer sions and boundaries of realtechnical and engineering problems.When we want to show the vector prop-erties in particular, the unit vectors i. I and k

are usec to point the way.

Represents H

Fig.9. A though the iron filings in thisfamous old experiment are not curlH.nevertheless they lie along the lines ofcirculation of the magnetic field. Then ifyou imagine the amount of this circulationper unh: area in the region. tien you havethe idea of curl' I there.

Running away into the above discussion. I

have put the cart before the horse, at least byassLming everybody knows about ordinaryvectors and how to multiply them. Quite anumber of people probably don't

Adding two vectors is simple. We use the'parallelogram' or 'triangle' rule that mostpeople no doubt did learn. Products ofvectors give rise to a little more thought Theresults all depend on how we define themuhiplication and this can be done in anumber of ways. Over the years, it has allboiled down to two different products - oneproducing a scalar result, the other giving anew vector. These products have survived.indeed have developed into 'standards' be-cause they fit well into how physics, en-gineering etc. describes things.

Tie first kind, Fig.6, known as the scalarproduct - or 'dot' product as some peoplecall it is simply

A B ABcosii

where A and B are the two vectors withmagnitudes A and B and an angle() betweenthen. This mens that if the vectors are atright angles there is no product, becausecos30 is zero. Of course, this product is themultiplication of one magnitude by the mag-nitude of the resolved component of theother vector along the direction of the firstFor example, also in Fig.6, it gives the workdone by a force (A, say) moving along B. (Thedirection of A in general being at angle () toB). However. the greatest value of A B occurswhen the vectors are parallel.

The other product, illustrated in Fig.7, wecall the vector or 'cross' product It is

A B NABsino

Now you see that the resolution takes theother component, the 'sine' one and theresult is greatest for normal, or 'orthogonal'vectors this time. N is the direction vector orunit vector such that if A is turned towards B.N goes in the direction of an ordinarycorkscrew.

If we consider B A instead, then N pointsthe other way. This means A B is not equalto BxA. In other words, A and B do notcommute in the cross product Often this isthe first time students come across a non -commutative algebra, (and some of themcome to grief at first). In fact

Ax8= -BxAAt his stage in proceedings. if you look into abock dealing with vectors'', you can have anerrertaining time working out some of themL. Itiple products like

A (BxC)

or A.(B x

Nobody has give a meaning to division by avector, so you won't come across that opera-tioi. This means that the denominator ofdifferentiations always contain scalars, orscalar components of a vector. and so on.

It is normal practice to typecast vectors inhold face or. as an alternative to underlinethe characters. as in our drawings. Here theyare in light face as the text is in bold.

ELECTRONICS & WIRELESS WORLDSII

point in space: in other words they formorthogonal sets. Admittedly, as Fig.8 mightremind you, the rectangular set x, y, z is thesimplest.

If we put the properties of scalar andvector fields into cartesian coordinates, weget the three sets of equations that describethe three components of each field.

V is equivalent to 14+ j4+(IX ay az

in cartesians. This shows the three compo-nents explicitly. i, j and k are the unit vectorsalong the x, y, and z axes.

Therefore gradtp, which is Vtp, has thethree components

it it itI ay ' dz

in cartesians and is a vector.

divA, also written V.A. is

(1ax jiff y+j4+ Ic41.(iAx+jA, +Irk)

which, when we notice Li= I and i.j= 0 etc.,is

i(Ax Ly. + AAz

(IX (6/ (iz

curlB, alternatively V x B, is

(. . . .r: + j=+ Ic7- X (111,, +JB,+kB,)X (ly (12

Keep a sharp eye on the definition of thevector product and you will see that i x i etc= 0, but i xj = k, xk = -j and so on, whichyields

dBi i16. (aB aBz)+IC. (aBx aa,)(ly (12 ilZ ilX ax ay

for the three components of the vectorcurlB.

The curl expansion, definitely the mostcomplicated, has a determinant expressionso that you can remember it easily

curlB i j k

a it it

ax ay az

13,, B,. B,

MAXWELL'S EQUATIONS ANDE M WAVES

With our little nibble at vector fields, wearrive at one of the most famous sets ofequations to grace the table of science andengineering. Oersted showed that a magne-tic field H amps per metre sprang up round acurrent flow in a conductor. You may re-member the science master demonstratingthis with a card and iron filings. Fig.9.

Ampere wrote down a mathematical state-ment about what might be going on. Usingour modern notation and units. togetherwith the idea of the current flow paths as avector field distribution, like that in Fig.10,Ampere's description amounts to saying thatthe curl of the magnetic force at any point isequal to the current density J streamingthrough

curIH = J amps m

Michael Faraday found the inverse effect,namely his Law of Electromagnetic Induc-

tion. A changing magnetic flux field linkinga circuit causes a voltage round it. In vectorfield terms we can write this as a changingflux density B webers per square metres atany point sets up the curl of an electromotiveforce field E volts per metre to circulateround it. So Faraday's Law says that the curlof E is equal to the negative of the magneticdisplacement current density threading theregion, see Fig.11

aBcurIE= -7t volt m-2

An electric force field E sets up an electricflux field D, usually called the electricdisplacement, in a kind of 'stress and strain'relationship.

D = a coulomb m-2

where e is the well known permittivity of themedium in which the fields occur.

The units of e are farads per metre. Youcan see that in the above, farads per metretimes volts per metre is equal to farads timesvolts per square metre. You might recall thata farad multiple by a volt is a coulomb.Similarly the magentic force H sets up astrain, the magnetic flux density B, so that

B = RH weber m-2 or tesla

where p. is the permeability in henrys permetre.

The e and p. look like 'stress and strain'moduli of some kind, relating electric andmagnetic forces with their fluxes. This israther like Young's modulus in mechanicalstress and strain. No wonder the Victoriansstruggled to invent an Aether in which allthese goings-on could occur. Many a genera-tion of students say how perplexing to havefour vectors describing EM fields, but if wethink of them as a force field together with aflux field in each case (the 'stress andstrain), it does help.

In our modern units, even a vacuum hasvalues for E and p.. They are

Eft3671.x 109 farad m I

14)=-47r X 107 henry m I

Electric charge gives rise to the flux field D.If there is a concentration p coulombs perunit volume distributed in a region, then

divD = p coulomb m-3.

Nobody has found isolated magnetic chargeyet, although people are looking. Therefore

divB = 0.

This means magnetic vector flux lines alwaysclose on themselves.

We say they are solenoidal. And in somediscussions, authors write the magnetic fluxdensity in terms of another vector as B =curIA, where A is called the vector potential.Because of this, the only magnetic currentthat can flow is magnetic displacementcurrent - as we saw in Faraday's Law.

We have one last result, not usuallyincluded in Maxwell's equations, which re-lates the current flow lines coming out of aregion to the rate of change of charge per

'The S.I. Committee asks us to call a weher m 2. a lesla.

Fig.10. CurIH goes round flux of current Jsomething like this.

Curt L

- - - -- -4 --

-65

/ `61

Fig.11. Similarly for the rate of change offlux density B. But notice the direction ofcurIF in this case.

ek

14b.

b.

Fig.12. The EM wave advances somewhatlike this diagram suggests. The E andfield lines cross at right angles to eachother and with the direction of travel. For asingle frequency, the lines distribute alongthe propagation axis in strength and withreversal's depicted in the lower part of thediagram.

unit volume at each point in it. This is the'equation of continuity'

divJ = -at coulomb m-35-1 (or amp m -31

Now here is an interesting conundrum, aproblem Maxwell solved although not quitethis way. From your knowledge of vectors,and the equation of continuity. you mighttry getting the continuity result from thecurlH equation by taking its divergence

div cur1H=divJ= -at

This looks alright - but hold on, div of a curlis always zero. Apparently we can never havea charge changing with time! Maxwell gotout of this difficulty by looking at the other


curl equation (the one for E) and by analogyadding on to curlH a separate flux changeterm which he called the electric displace-ment current. So this equation with Max-well's revolutionary bit of addition, reads

aDcurIH=J+7 amp m -2.

Now try the continuity equation

div cur1H=div(.1+1)-0

divJ = -div(2-12at

The order of differentiation doesn't matter

.*. divJ = - Dat

and from the fact that divD= -p we havedirectly

divi= -at

So at last we have Maxwell's equations.For example, out in space, away from cur-rents J and charges p. the equations are

curIH=-aD

curIE= - -aB withat

divD =0divB =0

f D=E,,EB = RoH

The two curl equations tell quite a story.If, say, the electric field changes with time,there must be an accompanying magneticfield at right angles to it. If this is changingwith time also. then it sets up anotherelectric field - again at right angles. And soon indefinitely. It looks as though E and Hcan support each other in space. In otherwords, a wave might exist. Notice the twofields cross everywhere at right angles be-cause of the curl property relating them. Thefields also lie across the line of travel, whichmeans that the wave is a transverse type andwill polarize. Therefore a plane electro-magnetic wave travelling along the x axismight look something like that shown inFig.12, if we could actually see the lines. Thisset of results shows Maxwell had writtendown the transmission line equations forfree space.

What we do with them now is to take thecurl of the first equation

curl curIH = curl-aDat '

As the order of differentiation doesn't matterand is a constant,

curl curl H=-at curlD=EoacurIE.

But from the second curl equation.Curl E= - aBiat, so insert it

aHcurl curIH= -14E042.

We know all about expanding curl curl of a

vector from our previous discussions, there-fore the piece de resistance:

aHcurl curIH= grad divH -V 2H

= RoEii-at2

aH.. \.72H= poE11at2.

This is the famous wave equation. Here itdescribes the magnetic field part of theelectromagnetic waves Maxwell predicted.The electric field has a similar wave equa-tion. The multiplier of the right hand, or'time' term is always in this type of equation,which. by the way. is called d'Alembert'sequation. where c is the velocity of the wave.

This means that for our derivation withMaxwell's equations. And what is more, cworks out to be very nearly 300 000 000metres per second from the measured valuesof E and p.. This is the velocity of light in avacuum.

References1. 'Joules Watt'. Maxwell's EM theory revisited. E

July 1987.2. I have found a very good simple introduction to partial

differentiation in chapter :1 land ch. 91 of "TeachYourself Further Calculus-. by F.L. Westwater. En-glish Universities Press Ltd..11966)

:i. 0. Heaviside. "Electromagnetic Theory' Vol. 1. Chap.Ill "The Electrician" Publishing Co. Ltd.. 118931 Ormodern reprint.

A. Ref. 3. is pithy and interesting. For a good moderntreatment. see: -Field and Wave Electromagnetics".Chap. 2. by U.K. Cheng. Addison-Wesley. 119831.

25 years of space communications

Aquarter of a century ago this month the era ofinternational communications through space began.Telstar, the world's first commercial communica-

tions satellite, was launched by NASA into low earth orbiton 10 July, 1962, and within days it had successfullyrelayed television, telephone and facsimile signals acrossthe Atlantic in both directions.

Built in the United States by Bell Telephone Laborator-ies, Telstar weighed just 77kg and it carried a 2.5Wtransmitter capable of handling 600 telephone circuits or asingle television channel: the uplink frequency was6.39GHz and the downlink 4.17GHz. The spacecraftorbited the earth once every 156 minutes. at a. heightvarying between 910 and 5875km.

Communication between the US and Europe waspossible only when both ground -stations could see thesatellite: about three or four useful orbits occurred inevery 24 hours. offering transmission periods averagingabout 30 minutes. And it was to maximize these that theGeneral Post Office (British Telecom's predecessor)erected its dish close to the western extremity of the UKmainland, at Coonhilly Down.

Ceostationary satellites, which would later allow trans-

mission for extended periods, were to come much later -though their principle was by then well known, havingfirst appeared in print in Wireless World in February 1945.in a letter from science writer Arthur C. Clarke.

The first night's tv transmissions from Telstar weremarred by poor reception: through a misunderstanding,engineers on either side had adopted opposite planes ofpolarization. But by the following evening, staff atGoonhilly had made the necessary adjustments and livepictures were exchanged between the BBC and USnetworks. Colour pictures from a 525 -line slide scannerprovided by the BBC Research Department were sent on16 July: The US station at Andover. Maine, reported themas "excellent".

Today there are some 130 communications satellites inservice, handling traffic between 160 countries. Satelliteprogramme feeds are routine events on radio and televi-sion, and the era of direct satellite broadcasting into thehome is almost with us. Coonhilly now has ten aerialsworking to satellites over the Atlantic and Indian oceans;the first of them, the dish used for Telstar, is stilloperational.

ELECTRONICS & WIRELESS WORLD S 3

01-208 1177 TI.CHNOMA IFIK LID 01-208 1177BBC Computer & Econet Referral Centre

AMB15 88(, MASTER I fro, --.011ware £395 (a)AMC06 Turbo 1650 02) Expansion Module

ADC08 512 ProcessorADF14 Horn CartridgeADJ22 Rel Manual Part I

(195(b)

E13 (b)

(14(c)

ADJ24 Advanced Het ManualADF10 cone? ModuleADJ23 fret Manual Part IIBBC Master Uust Cover

BBC MASTER COMPACTFree bundled software within tau r.SYSTEM 1 128K Single 6408 Drive not bundled software £299 (a)SYSTEM 2 System 1 with a 12 Hi Res HGB Monitor £349 (a)SYSTEM 3 System 1 with a 14 Med Res AGO Monitor E489 (a)Second Dove Kit £99 (e) Extension Cable for eat 5 25 dove £12.50 (0)

View 3 0 User Guide £10BBC Dust Cover £4.50 (0)ADE S ROM (tor B with 1770 DFS 6 B Plus) £26 (d)ACORN ZEIO 2nd Processors £329 (a) ,

MULTIFORM /80 2nd Processor £289 (b)TORCH Z80 2nd Processor ZEP ICOTZDP 240 ZEP 100 with Technomanc PD800P dual drive with built-in monitor stand

41.4812 BBC MASTER F. E315 la)£99(b)

E19 50 (c)

E41 (c)(14(c)

E4 75 (d)

Viewsheet User Guide C10 (d)1770 DFS Upgrade for Model B (43.50 (d)

120S ROM 05(0)Universal 2nd Processor Adaptor En (b)

ACORN IEEE Interlace £265 (a)£229(a)E439 (a)

META Version III - The only package available in the micro market that willassemble 27 different processors at the price offered. Supplied on two 16Kroms and two discs and fully compatible with all BBC models. Please phonefor comprehensive leaflet £145 (b).

We stock the full range of ACORN hardware and tirmw3ri ind a very wide range of otherperipherals for the BBC For detailed specifications and pm ,r end for our leaflet

EIISC5.25" Single Drives 40/50 switchabie:

DRIVESTS400 4001(640K £114 (b)

£129 (b)PS400 400K/640K with integral mains power supply5.25" Dual Drives 40/80 switchable:TD800 800KJ1280K £199 (a)PD800 8008/ 1280K with integral mains power supply f229 (a)PD800P 800K/1280K with integral mains power supply and monitor stand £249 (a)3.5 801 DS Drives:TS351 Single 40010640K C75 (b)PS351 Single 400K/640K with integral mains power supply 0119 (b)TD352 Dual 800K/1280K E135 (b)PD352 Dual 8001Q1280K with integral mains power supply £187 (b)PD853 Combo Dual 5.2573.5- drive with p.s.0 £229 (a)

PRINTERS & PLOTTERSEPSONLX800 £229 (a) STAR NL I 0 (Parallel Interface) f 209 (a)FX800 C329 (a) STAR NLIO (Serial Interface I f279 (a)FX1000 £449 (a)EX800 £409(a) BROTHER HR20 £349(a)L0800 (80 col) £439 (a)L01000 £589 (a) COLOUR PRINTERS

Integrex Jet Printer f 549 (alTAXANKP815 (160 cps) C269 (a) Dotprint Plus NLO Rom forKP915 (180 cps) £369 (a) Epson versions for FX. RX, MXJUKI and GLP (28(d)6100 (Daisy Wheel) £319 (a)

PLOTTERSNATIONAL PANASONIC Hitachi 672 f 495 (a)KX P1081 (80 col) £159 (a) Plotmate A45 f370 (a)KX P 3131 (37K buffer) £249 (a) Plotmate A4SM f450 (a)KX P 3132 (32K buffer) f249 (a) Plotmate A3M £569 (a)

PRINTER ACCESSORIESWe hold a wide range of printer attachments (sheet feeders, tractor feeds etc)in stock. Serial, parallel, IEEE and other interfaces also available. Ribbonsavailable for all above plotters. Pens with a variety of tips and colours alsoavailable. Please phone for details and prices.Plain Fanfold Paper with extra fine perforation (Clean Edge):2000 sheets 9.5' x 11" £13(b) 2000 sheets 14.5' x 11' f 8.50(b)Labels per 1000s Single Row 3' x 1 7/16' £5.25(d) Triple Row 2-7/16' X 1 7/16' £5.00(d)

Single Disc Cable f6 (d)10 Disc Library Case £1.80 (d)50 5'2 Disc Lock** Box £9.00 (c)

3M FLOPPY DISCSIndustry Standard floppy discs with a lifetime guarantee. Discs in packs of 10

51/4" Discs 3i5 Discs40 T SS DD C10.00 d 40 T DS DD £12.00 d 80 T SS DD C.00 d80 T SS DD 0(

4.50 (d))(

80 T DS DD £15.50 (d)) 80 T DS DD C2520

(.00 (d))

FLOPPICLENE DRIVEHEAD CLEANING KITFLOPPICLENE Disc Head Cleaning Kit with 28 disposable cleaning discsensures continued optimum performance of the drives. 51/4" £12.50 (d)

31/2" £14.00 (d)

DRIVE ACCESSORIESDual Disc Cable £8.50 (d)30 x 5' 2 Disc Storage Box £6 (e)100 5'2 Disc Lockable Box C13 (c)

MONITORSRGB 14"1431 Std Res C179 (a)1451 Med Res C225 (s)441 Hi Res 0365 (a)

MICROVITEC 14' RGB/PAL/Audio1431AP Std Res £199 (a)1451AP Std Res 0259 (a)All above monitors available in plastic ormetal case.

TAXAN SUPERVISION 62012" - Hi Res with amber/green options.IBM compatible £279 (a)Taxan Supervision 625 C329 (a)

MITSUBISHIXC1404 14- Med Res RGB. IBM 8 BBCcompatible £219 (a)

MONOCHROMETAXAN 12" HI-RESKX1201G green screen £85(a)KX1203A amber screen £95 (a)

PHILIPS 12" HI-RESBM7502 green screen £75 (a)BM7522 amber screen 079 (a)BM7542 white screen £79 (a)

ACCESSORIESMicrovitec Swivel Base £20 (e)Taxan Mono Swivel Base withclock £22 (C)Philips Swivel Base £14BBC RGB Cable f5 (dMicrovitec C3.50 (dTaxan £5 (d) Monochrome £3.50 (d)Touchtec - 501 0239 (b)

MODEMSAll modems carry a full BT approval

MIRACLE TECHNOLOGY WS Range

WS4000 V21/23 (Hayes Compatible,Intelligent, Auto Dial/Auto Answer) £135 (b)

WS3000 V21/23 Professional As WS4000and with BELL standards and battery backupfor memory £295 (b)

WS3000 V22 Professional As WS300 V21/23but with 1200 baud full duplex £495 (a)

WS3000 V22 bis Professional As V22 and2400 baud full duplex C650 (a)

The price of WS4000 and WS3000modems includes a COMMSTAR 11 ROMand BBC Data Cable.

WS2000 V21 V23 Manual Modem £95 (b)

DATA Cable for WS series/PC or XT £10 (d)

Auto Dial Card £27 (d)

Auto Answer Card £27 (d)

WS2000 SK I Kit. £5 (d)

This offer can only apply if It is specifled at thetime of placing your order for the modem.

RT256 3 PORT SWITCHOVERSERIAL INTERFACE3 input 1 output or 1 input.3 outputmanual channel selection Inputoutput baud rates independentlyselectable 7 bitt bit. oddeven noneparity Hardware or softwarehandshake 256K buffer mainspowered (375(b)

PB BUFFERInternal buffer for most Epsonprinters Easy to install InstsuppliedPB1211128K (c)

UVERASERSuv1T Eraser with built-in timer and mains indicatorBuilt-in safety interlock to avoid accidental exposureto the harmful UV raysit can handle up to 5eproms at a time with an averageerasing time of about 20 mins £59 + £2 pip.UV1 as above but without the timer £47 + £2 pip.For Industrial Users. we offer UV140 8 UV141 era-sers with handling capacity of 14 eproms UV141 hasa built in timer Both offer 'all built in safety featuresUV140 E69, UV141 £85. pip £2.50.

EXT SERIALiPARALLELCONVERTERS

Mains powered convertersSena) to ParallelParallel to Serial £48 cBidirectional Converter £105 b

Serial Test CableSerial Cable switchable at both endsallowing pin options to be re-routed orlinked at either end - making it possibleto produce almost any cableconfiguration on siteAvailable as M/M or M 'F £24.75 (d)

Serial Mini Patch BoxAllows an easy method toreconfigure pin functionswithout rewiring the cableassay Jumpers can be usedand reused [22 (d)

Serial Mini TestMonitors RS232C and CCITTV24 Transmissions.indicating status with dualcolour LEDs on 7 moststgnificant lines Connects inLine 122.50 (d(

CONNECTOR SYSTEMS

I.D. CONNECTORS(Speadidock Typal

No of Header Recep Edgeways Plug 'ac le Conn

10 90p Up 120020 145p 125p 195p26 175p 50p 240p34 200P IWO 320p40 220p 190p 340p50 235p 2000 390p

SOFTY IIThis low cost intelligent eprom programmer can program 2716 25162532 2732 and wan an adaptor 2564 and 2764 Displays 512 bytepage on TV has a serial and par-allel I/O routines Can be used as an emulator cassette interfaceSofty II C195.00(b)Adaptor fnr 276412564 (25.00

SPECIAL OFFER264-25 £3.20 (d);

27128-25 £3.50 (d)27256 £5.50 (d);27512 £9.00 (d);

6264 LP -15 £2.80 (d).

D CONNECTORSNo of Ways

9 15 25 37MALE:Ang Pins 120 180 230 350Solder 60 85 125 170113C 175 275 325 -FEMALE:St Pin 100 140 210 380Ang Pins 160 210 275 440

der 90 130 t95 290195 325 375 -

Hood 90 95 100 120,ew t30 150 175 -

EDGE

CONNECTORS2 6 -was Icommodore,2x 10way2 x I2way 2012x 18 way2 23 way 12881.2x 25 way2 x 28 way (S0e,num2x 36 way

way2. 22 way2x 43 way

77 way2 . 50 wfay1S100conn,

'56- 3000SOp - - 350p- 140p

i ?Sp 220p225p 2200200p -250p -260p -190p -395p400p 500-p600p -

AMPHENOLCONNECTORS

36 way plug Centronicssolder 500p (IDCI 475p

36 way skt Centronicssolder) 550p ( IDCI 500p

24 way plug IEEE (solder)475p 479p24 way ski IEEE IsOlder1500p IDC 500pPCB Mtg Skt Ang Pin24 way 700p 36 way 750p

10 -way

16 way

20 -way

26 -way

RIBBON CABLEgrey met,

4op 34-4a,Epp 40 -way

55p 50 way120p 64 -war

160p

180p200p280p

TEXTOOL ZIF50CxE15 24 -pin E761). 9 10 40 pmE12 10

EURO CONNECTORSDIN 41612 Plug SIC2 x 32 way St Pin 230p 275p2 x 32 way Ang Pin 275p 320p3 x 32 way St Pin 260p 300p3 X 32 way Ang Pin 375p 400pIDC Skt A + B 400pIDC51414+C 400p

For 2 x 32 way please specifyspacing (A + 8. A Cl.

GENDER CHANGERS25 way D type

Male to MaleMale to FemaleFemale to Female

£10

010

DIL HEADERSSolder IDC

14 pin 40p 100p16 pin Sop 110p18 pin 60p20 ptn 75p24 pin 100p 150p28 pin 160p 200p40 pin 200p 225p

RS 232 JUMPERS125 way Di

24 Single end Male24 Single end Female24 Female Female24 Male Male24 Male Female

C500£5 25

f 10 00f 9 50

£9 SO

MISC CONNS21 pin Scar! Connector 200p8 pin Video Connector 200p

DIL SWITCHES4 -way 90p 6 -way 10508 -way 120p 10 -was 150p

ATTENTIONAll prices in this double pageadvertisment are subject to

change without noticeALL PRICES EXCLUDE VAT

Please add carriage 50punless indicated as follows(a) £8 (b) £2.50 (c) Et 50 (d)

£1.00

TECHNOLINE VIEWDATA SYSTEM. TEL: 01-450 9764Using 'Prestel' type protocols. For informationand orders - 24 hour service, 7 days a week

ti 1 4 ELECTRONICS & WIRELESS WORLD

74 SERIES

740174027403'404'4057406'407'408'409'41074117412'413-414'416-4,7'420'421'422'423:4257426'427-4213743074327433743774387439744074417442474434744474457446474474744874507451745374547460747074727473747474757476748074817483A74844748574867489749047491749247493474947495A7496'497-4100'4107-4109-4110741117411674118-41194120'4121'4122-4123'4125'4126741287413274136'4141'4142'4143-4144'414574147741487415074151474153741547415574156'4157'4159-4160'4161'4162"4163'4164'4165'4166'4167'4170'4172-4173'4174741757417674178-4179'4180-4181'4182'4184'41854'4190

.1191'4192-419374194'41954I96'419774198'4199742217425174259:4265

0.300.300.300.300.360.300.400.400.300.300300.300.300.500.700.360.400.300.600.360.360400400320.430.300.360.300.300400.400.400.900.701.001.100.701.001.001.000.360.350.380.380.550.500.450.450.500.600.450651.801.051.251.100.422.100.550.700.700.551.10

0.600.802901.900.500.750.750.551.701.101.701.000.5S0.700.600.650.550.550.750.700.902.501.302.701.101.701.401.750.700.001.400.500.900.802.251.100.901.101101.201101.404.002.000.20IA01101.051.001.501.501.003.401.401.801.613

1.301.301101151.10

0.801.301102 202.201.101.001.500.80

74273 2.0074276 1.4074278 1.7074279 0.9074283 11574742 3.2074290 0.9074293 0.9074296 1.8074351 2.0074365A 0.80743664 0.80743674 0.8074376 1.6074390 1.1074393 1.2074190 1.40

74LS SERIES

74LS00741S0174LS02741503741504741505741508741509741S10741511741S137415147,115157,11520741521741522741524741.526741527741528741530741532741532741S33741.537741538741540741_542741543741548741549741S51741S54

0.240.240.240.240240.240.240.240.240.240.340.500.240.240.240.240.500.260.240.240,240.240.240.240.240.240.240.501.500101.000.240.24

741555 0.24741.573A 0.307415744 0.35741575 0.457415764 0.36741578 0.427415834 0.70741S85 0.7574L S86 0.35741 S90 0.48741591 0.90741592 0.35741593 0 54741.595B 0 75741596 0.9074L5107 0.407015109 0.407415112 0457415113 0457415114 0.457415122 0.70741S123 9.807415125 0.50741S126 0.507415132 0.65741S133 0.55741S136 0.457415138 0.557415139 0.557415145 0.957415147 1.757415148 1.40741S151 0.657415152 2.007415153 015741S154 1.60741S155 0.6574LS156 0.6574LS157 0.507415158 06574151609 0.65"4LS1614 0.7574151624 0.7574151634 0 75741S164 0.75741S1654 1107451664 1.507415168 1.307415169 1.007415170 1.4074151734 1107415174 0.75741S175 0157415181 2007415183 1.90741 5190 075741S191 0.7574LS192 0.8074151944 0.7574151954 0757415196 0.807415197 0.607415221 0.9074LS240 0807415241 0.807415242 0 9074LS243 0.907415244 0.70741.5245 0.9074L5247 110741.5248 1107415249 1.10741S251 0.75741S253 0.757415256 0.9074524574 0.7074152584 0707415259 1.20741 7260 0.75747. 0.60

7415273 1.25741S279 0.707415280 1.907415283 0.807415290 0.80741529214007415293 0607415295 1.40741529)14.007415298 1.007415299 2.207415321 3.70741S3228 3907015323 3.007415324 120741S348 2.00741S352 1.207415353 1.207415356 2.107415363 1.807415354 1.807415365 0507415366 0.50741S367 0.527415368 0.507415373 0.707415374 0.70741.5375 0.757415377 1.307415378 0.957415379 1.307415381 4.507415385 3.25741S390 0.60741S393 1.0014153954 1.00

741S399 1.407415445 11074LS465 1.20741S467 1.2074LS490 1107415540 1.007415541 1.007415608 7.007415610 25.007455612 25.0074LS624 3.507415626 2.25741S628 2.257415629 1.257415640 2.007415640-1 3.00741S641 1.507415642 2.50741S642.1 300741S643 2.50745643-1 30074LS644 3.50741S645 2.007415645-1 4 007415668 0.907415669 0.907415670 1.707415682 2.5074L5683 3.0074L5684 3.507415687 15074LS688 1507415763 16.00

74C SERIES

74C00 0.7074C04 0.5074C08 0.7074010 0.7074C14 0.5074C20 0.7074C32 1.0074C42 1.5074C48 1.5074C73 1.0074C74 1.2074C76 1.0074003 2.0074C85 2.2574C86 0.5074C90 1.9074C93 11074C95 1.6074C107 1.0074C150 5.0074C151 2.0074C157 2.5074C 160 1.8074C 161 1.8074C 162 1.8074C163 ..9074C173 1.0074C174 1.5074C175 1.5074C193 1.5074C194 15074C195 1.5074C221 2.5074C244 2.0074C245 2.2574C373 2.2574C374 2.2574C902 1.2074C911 9.0074C912 4.5074C922 6.0074C923 6.5074C925 6.5074C926 7.50

74ALS SERIES

7441500 0.4574411502 0.457441.504 0.507441508 0.507441.510 0.4574ALS20 0.4574ALS32 0.457441574 0.70'4615138 1.50-4615139 110'4415244 4.00'445245 4.75-4415573 2.6074415574 4.5074415580 2.60

74S SERIES

74500745027450474505745087451074511745207452274S3074S32745377453874540745517456474S74745857458674511274511374511474512174513274513374513874513974514074515174515374515774515874516374516974517474517574518874518974519474519574519674520074520174522574S24074524174524474525174S25774525874526074526174528374528774528874528974529974S373745374-4,387

0500.500.500.500.500.500.750.500.500.500.600.600100.500.600.450.705,501001.501.201.203.001.000.601101101.001.501.502002.003.005.503.003.201101101003.003.504.503./05.204.001.004002.502.502.501 003.002.702.752.002.254.504.00100225

4000 SERIES

4000400140024006400740084009401040114012401340144015401640174018401940204021402240234024402540264027402840294030403140324033403440354036403740384040404140424043404440454046404740484049405040513052405340544055405640604063406640674068406940704071407240734075

0.200.240.250.700.250.600.450.600.240.250.360600.700.360.550.600600.800.600.700300.460.240100100.600.750.351.251.001.252500.702.501.101.000600.550 500600 601 000 600 600 550 360 350 650 600 600800800850.700.850.402.300.250.240.240.240.240.240.24

407640774078408140824085

408640894093409440954096409740984099450145024503450445054506450745084510451145124513451445154516451745184519452045214522452645274528452945314532453445364538453945414543455145534555455645574560456645684569457245834584458547241441114412144161441914490144951450001459922100221012210240014400854009740098401004010140102401034010440105401064010740108401094011040114401474016340173401744017540192401934019440244402454025740373403740009580C9780098

0.650.250.250.240.250.600.751.20

0.350100.950.902.700.750.990.360.550.360151600.900.351.200.550.550.551.501.101.10

0.552.200480.320.601.15

0.800.700.800.65too0.750.651802.500.750.750.900201.002100.360.50

.2.401.401.402.401.700.450.900.480.601.507.507.503002.604.204.506.502003.507.007100.481.200.360.401501.251.34)

2.001.201.500.480.553.200.602.252.252.601.001 201.001.001.001.001.001 so1.501.S011401.600.750.750.75

LINEAR ICs. COMPUTER COMPONENTSA 7581 124000808 1190441791000 29 pp49103 2.00AN -I-5050 10040.3.1350 3.5047.18910 49048.18912 5.00CA30194 1.00CA3020 3.50CA3028A 110CA3046 0.70CA3059 3.25CA3060 3.50CA3080E 0.70CA3085 1.50CA3086 060CA3089E 2.50CA309040 3.75CA3130E 010CA31301 1.30CA3140E 0.45CA31401 1.00CA3146 2.25CA3150E 1.50CA3161E 2.00CA3162E 6.00CA3189E 2.70CA3240E 1.50CA3280G 3.0007002 6.00DAC1408.13 3.00DAC0800 100DAC0808 3.00OG308 3.0064I365 1.90CL7106 6.75

17611 095CL7650 400CL7660 2.50

18030 4.00CM7555 090CM7556 1.40LC7120 3.00LC7130 3.00LC7131 3.50LC7137 3.5071-347 1.20LF351 0.60LF353 0.90I.F 355 0.90LF 356N 1.10LF357 1.00LF398 4.001.1010C1H 4.501583018 0.30114307 0.45158306041 0.75LM310 2.25LM3II 0.60LM318 1.50158319 1.80114324 0.451143342 1.15LM3352 1.30LM336 1.60LM339 0.40114348 0 601.113580 0.50LM377 3.00158341014.8 1 50

1143809 1.50LM381491.70161381N 100LM383 3.25LM384 2 2011.138674-1 1 00

LM387 2 70LM391 1 801M3925 1.10LM393 085LA13940- 4.00LM709 0.35

LA.4710 0.48758711 1.00711723 0.607117250N 3.00

44733 0.6511741 0.221447747 0.701,4748 0.301%11011 4.807581014 1.507611801 3.007581830 2.507581871 3.007141872 3.00_M1886 6.00911889 4.50142917 3.00

1813302 0.901583900 0.801143909 1.00L813911 1.807543914 3.50113915 3.40143916 3.40114I3600 1.5014515131 2.30M515161 4.S014E13712 2.00kt,310P 1.501413 0.75MCI456 0.45MC1495 3.00N4C1496 0.700C33400 2.005803401 0.70MC3403 0.65MF 10CN 4.10MX502409.0041902 5.00581922 4.0095862218 3.00NE529NE531NE544NE555NE556NE564NE565NE56611E567NE570NE571NE592NE55327

2.201.201.900220.604.001.201501.254.003.000.90ISO

NE55337 110NE5534P 120NE553447 110OP-07EP 3.50P11024 5.001104136 0.55RC4151 2.009C4195 150RC4558 0.5S350240 9.005441900 16.00SF F96364 $.0051490 3.0094760139 300iN7602324 3.00'15760335 30011476115N 2.15514764894.003976495 4.003976660 1.20370256412 710308515 7.50'87120 1.20'47130 1.40'47204 110147205 0.90047222 1.50147310 1.50

184231 1.20

184800 0E0184870 0.9010420 080T13482014 0.75TBA920 2.00TBA950 2.25TC9109 5.00TCA210 3.501CA220 3.50TC4270 3.50TC4940 1.7510810044 5.000041010 2.25704I022 4.50TDA1024 11010411705 300TDA2002 125TDA2003 1101042004 2.40TDA2006 3.20TDA2020 3.20TDA2030 2.50TDA2593 5.00TDA2653 7.001043560 7.501043810 7.50TDA 7000 3.500E41002 7.0011061CP 0.40T106211064T1071T1072T1074T10811108211083TL08411094114300U4759 3.20U42240 1.20

vt-1 6.00011422014 0.7501920024 0/501520034 0/5019120044 0.75U192068 2.90ULN2802 1.9011192803 1.80ULN2804 1/0UPC575 2.75UPC592N 200UPC11561-1 3.00URCINI5N 50009210 4.00092206 4.50002207 3.750112211 5.75X52216 6.7SX52240 1.20ZN404 10029414 0.80294190 1.7529423E 1.30Z94246 1.30ZN425E8 3.502942668 100ZN427E8 6.00ZN428E84.50IN429E8 225ZN447E 9.00ZN448 I 7.50ZN449E 100ZN450E 7.50224459CP 310Z91034E 2101941040 6.60

m!--1 21001194234E9.50

0.600100.400.701.10

0.350.550751.002.001.20

c Pu

1802CE 6.5026508 16506502 4.5065CO2-28811,

12.006502A 6.5065028 8.006800 2.506502 3.005809 6.506809E 10.0068809 10.0068B09E 120068000 LB 36.00

13035 3.50800 35 6.008039 4.2080C39 7 0080008 7408085A 3.00800854 7.508086 22.00

8087808' 8808887418748'V'-7

020.6V30.81802803Z800Z804

E120E16017.5015.0012.0012 0014501800100010001502.901507.50

trizu

VOLTAGE REGULATORS

1/1 FIXED VOLTAGE PLASTIC 70220+VE -VE

5V 7805 015 7905 0.506V 7806 050 7906 0.506V 7808 0.50 7908 0.50128 7812 0.45 7912 0.50159 7815 050 7915 0.15188 7818 0.50 7918 0.50248 7824 0.50 7924 0.50

1A FIXED VOLTAGE PLASTIC 10925V 78105 0.30 5V 79105 0.456V 78106 0.30 128 79112 0.508V 78108 0.30 158 79115 0.50128 78112 0.30158 78115 0.30

26512816 3032423245652065226522A65326551A

682168021682968406884068506885068526854688546875

815481558156

82058212821682248226822882438250

120020008.004.503.003.505.504.806.00

2.5012.503.756.001 802 502.506 508005008.503803 80

2.252.001.60

P0.44.255.502.601200

82514 3.258253C-5 3.5082554C-5 3.208256 18.008257C-5 54.00

8259C.5 4.00

8275 29008279C-5 4.808282 4.03

8284 4.608287 3.8082880 6.5087554 16.00

T1454500 14.0011459901 5.001059901 5.0015859911 111.01

15859914 14.10

160110 2.00ZBOAPIO 225/80C TC 2.50ZACIACTC 2751300481 6.501,304048T 7.40

2800MA 7.00/9040144 7'A

Z90451001,97,400

18013110 5.140

111013CTC 5.90Z306DART 980

2101 4.00

21078 5.002111A-354.002114 1.50

21143 250'

2147 CPO

4116 2.904116.20 1.90

41256.15 3.00

4164.1501 3.00

4416-15 3.504532.20 2..40

5101/501 4.00

55145114 4.005516 4005517AP 4.00

61160-3 3.50611610.3

6264L715 3.60

6514.45 4.006810 2110745189 1.60745289 2.25

93415 610

93425 6.30

PROM,

28122 411024510 2.50185030 2.00185A030 2.00746188 1.80745287 2.25745288 1.80'45387 2.2582523 150825123 1.50825129 1.75

EPROM,

2516.5V 3.50251635 5.502532 450253230 5.502564 11 OP

2708 4.502716+5V 3.50271635 5.502732 4.502732002 9.50

2732830 60027324-35 5.502764-25 2.27C6429 6.

27128-25126027256.25 4.00

27512 P.O.A27512 25140015802716 S.00

CONTROLLER

CR T502' 1810CRT 5037 12.00CRT6545 9.00EF9364 8.00EF9365 25.00EF9366 25.00EF9367 38.00EF9369 12.00MC6845 6.50MC684557 6.50MC6847 6.50SFF96364 8.00TMS99213 10.00

EE PROM

2816-30 15.00210+8 30.009306 25661151160161 4.80

OTHER REGULATORS

FIXED REGULATORS154309K IA5V1583230 34 53.,781+050C 54 5v78412 54 QV78705 104 5V

VARIABLE REGULATORS158305441583171 10.220LM317K 7031M3371LA.13501 104+841115839613 vAli158723978HHO5KC 545V78HGKC SA+VAI178GUIC 1A.1.,A14794-IGKC 54-VAR..79GLAC -.843n

SWITCHING REGULATIONSICL 7660SG3524IL 4941149778540RC 4

TECHNOMATIC LTDM AIL ORDERS TO: 17 BURNLEY ROAD, LONDON NW10 lED

SHOPS AT:17 BURNLEY ROAD. LONDON NWIOTel: 01-723 0233 4 lines. Telex: 922800

305 EDGWARE ROAD. LONDON W2

1 403 505 406 40900

2 501202.402 25

15.000.505.756502.256.752.50

2.50100

2.252501 50

OTHERS

295777 0.50BPX25 1.80BPW21 2.80OCP71 1.80ORP12 1.20ORP60 1.20ORP61 1.20SF H205 1.00TiL32 0.55T1L78 0.55TiL31B 1.2011181 1.20111100 0.75

INTERFACEIC2

41)7581 15.00ADC0808 HIM405613 20.0041425510 3.50414251525213.50

48129.515383.50A14261531 1.20AM261532 11043479920C 25.00CAC8003, .

28006.003.503.501.401.501102.251.502.258.000.60060

MC3446 2.50MC3459 4.50MC53470 1.7544C3480 1550MC3486 2.25MC3487 2.25MC4024 150MC4044 5.50MC6883 16.00,MC 14411 7.50MC14412 7.5075107 0.9075108 0.9075109 1.2075110 0.90

1 75112 1.6075113 1.2075114 1.4075115 14075121 1.4075122 1.4075150P 1.20

DM8131DP8304DS36911358830D58831D58832DS8833OS8836058838D7002MC 1488MC 1489

LEDs

0.125'RED1112091312GRN 111211C16DEL 111112 020Rect LEDs119.0771 0.30COO ,13, c010.r

1.0010 LEDBar GraphRed 225Green 2.25

0.2-111220 015111222 011111226 012

'5154

.160151

,162'72182158189165

',450451

754527545375454754807549175492

1.202205.006.507.504.000.900600601 500800.500.500.700.701.500.650.65

8126 1.208728 1.208195 1.208196 1.208197 1.208198 1.20811595 1.40811596 1.40811597 1.4081 598 1.40881.5120 4.50

9602 3.009636A 1.609637AP 160

963P 1.90

.7E 150

CONTROLLERICI

765A 10006843 8 00

FD1771 20.00001791 20.00F01793 20.00FD1797 22.004C:-. 2400WD1691 15.00WD2143 12 00W02793 2700W0279' 27 00

v 804.0ENCODE Rs

CHARACTERGENERATORS

TDECODER

SA45020 6.00SAA5030 7.005445041 16.00SAA5050 9.00

A 11 50

740922 5.00740923 6.00

&ANA90

6 507 50

4,31015P 3.0049,51013P 3.0COM8017 3.001646402 4.50

100OULATORS

6MHz 3.75BMHz 4.50

SOUND &VISION

12MHz 12.00

CRYSTALS

31 7681010 1.0016432MHz 2.252 00MHz 2.252 45760MH71 L

2.002 45760Mhz1S.

2.502 5MHz 2.503 2761467 1.503 57951467 1.004 00M114 1.5041941411, 2104 43MHz 1.004 9152MHz 2.50S 00MHz 1.5C

06814117 1/56 00644, 1.406 144MH41.467 001,0810 1.507 161MHz 1.758 00MHz 1508 867MHz 17510 00MHz 1/910 50MHz 2.5010 70MHz 1.50n 00MHz 3.0071 002110 1.5014 9048: 1/514 31911, 11014 756MHz 2.5015 00MHz 2.0016 00MHz 2.0017 734MHz 15118 00MHz 1.5018 43210Hz 1.5019 969M1-17 1.5020 000440, 1.5024 0001.462 1 7548 711161.461 2.5P101000 1216

COUNTERS

74C925 6.5074C926 6.5074928 6.50

1941040 6.70

DISPLAYS

50357 1.00150500 111730

1.00F N0507 111729

1.009457' 01707

1.00MAN3640 17SMAP -44640 2.00

MAN6610200NSB58815,70111311 6.50111729 1.001117300 100MAN8910 1.50MAN8941 2.50

DISPLAYDRIVERS

9368 4.509370 4.S09374 3.50

LM3914 3.50LM3915 3.50L1143916 150UDN6118 3.20UDN6184 3.20ULN2003 0.90ULN2004 0.90ULN2068 2.90U1842802 1.90111520831.50ULN2804 1.9075491 0.7075492 0.70

Please note.

All prices are subject tochange without notice.

Only Current prime gradecomponents stocked.We also stock a widerange of Transistors.

()odes. Triacs Plastic.Bridge Rectifiers.

Thyristors and Zenors.Please phone for details.

OPTO-ELECTRONICS3.003.003.003.001.001.001.002.002.001.205.701.201.001.200.900.901.201.206.50

OPT(>ISOLATORS

BPX25BPX34BPW2100021F ND357MAN74/31.704MAN71 D1707MAN4640MAN6610MAN8910 0 8NSB588IORP12SF H305TIL31411L32TIL78TIL8IT11.100TIL311

11074 1.30MC726 1.00MC52400 1.90MOC3020 1.5011074 2.20

Nu,111112T111131111166913769139

0.700700.700701601 75

LOW PROFILE SOCKETS BY TI

8p,n 9p 113pm 169 24p,n 24p14pm 10p 200,1 18p 280n 26p16pin 119 22pm 209 40PT 30p

WIRE WRAP SOCKETS By Ti

8pM 25p 18p.n 50p 24pm 70p14pm 35p 20DM GOp 25Pm 63P16pm 40p 22p,n G5p 40PM 100p

TURNED PINLOW PROFILE SKIS

8p,n 259 16p.n 359 20pn 45p 28pin 65p14;,,n 30p 113pm 40p 24p.n 55p 40PT 110P

PLEASE ADD 50p p&p & 15°7o VAT(Export: BO AT. p&p ai Cosi)

Orders from Government Depts & Colleges etc. welcomeDetailed Price List on request

Stock items are normally by return of postMinimum ILicphonC ()flier L5

ealIZMO


VME 16 -BitPerform a n reat STE/STD/G64 8 -bit prices Performance/Price The exciting new range of PSI VME single -heightEurocards produce a system which is hard to match at any price!

O System Architecture Distributed intelligence provides for high localactivity and low global activity thereby minimising bus access andinterrupt latency. All cards come with their own on -board driversproviding a high level systems interface. This architecture offers Real -Time Event generation with the absolute minimum of system CPUintervention.

Surface Mount Extensive use of SMDs allow 5J to produce double -height VME functionality in a single -height package offering amultitude of cost -benefits to the user.

Signal Conditioning Industrial I/O is bus -independent and may beaccessed via a high speed parallel link or a remote serial link. The powerconnections to external plant are on seperate standard connectors at therear of the signal conditioning boards.

The PSI Range The extensive 5J range of processor independentEurocards are available for development and target applications. Therange includes CPU, hard and floppy disc interfaces, terminal interface,analogue and digital interfaces...

...and it's ALL BRITISH!ENTER 33 ON REPLY CARD

Board Featured: PSI I/O Local Processor

For more information contact our UK distributors:Cambridge Microprocessor SystemsBrookfield Industrial CentreTwentypence Road CottenhamCambridge CB4 4PS

lit 0954-51122

R. WITHERS COMMUNICATIONS LTD rManufacturers, importers and suppliers of work! famous communications products584 HAGLEY ROAD WEST OLDBURY, WARLEY, BIRMINGHAM B68 OBS021-421 8201 23 CELLNET 0660 323056. PRESTEL MBX 214216216 FAX 0215614074amateur Roo. Business Redo. Radio Teelphones Saies Service Accmsones and antenna systems

rigom' YAESU `11$' THE TECHNICALLY ORIENTATEDRADIO COMMUNICATIONS

100Khz to 950MHz IN ONE BOX! SPECIALISTS.

THE YAESU FRG9600/RWC Mk3 SCANNING RECEIVERNow at last a World First' from Ray Withers Communications Ltd. Here aresome of the unique features only available on this fine receiver

All Band Capar,,s., ;lb VHF Low Am -band 137-Satellme HEN -' Low UHF highMarine Atnaleut Business Radro Me, . up to 950MHu Tr,

a,y as all bands channel steps are now .7101 .1,4

All Mode Capabd.ty aide and narrow modes) LSB USB Video opt.r available sNESC PAL 5 56411-12

Soteclable Frequen, Steps S. modes have searlable tunIng SIO$35 winch allow last and eas.frequency (Mafiosi, st turitng,

IRO MeRtertatt.Scanninq .1, I nequency in the range can be entered stored 000 a memory anselected .Canned at balm- tuna.° allows monttontig of tmportant channel frequency white yedrors;°Met bands channels Mode, sti stored in memory

Fully Computer Compatible. ',item's CAT system is a standard teatime and a RS232 interface is availabhplus other interlaces for popular HOME compute's Watt the feature memones can be expanded and the unit co,PP,ernotel,ntroiled via a modern

Clock Function ,,codt tunclpn alkars tuna lowang and auro switch On Oa 0, the ',OW,

Nigh Receiver SensitivIty. In the range P0-950MM, a typical IlVeIVOI Sensanatv Is / Soyfor 1208 SINAI,filf) ran, 6MH2-60000MH: 2u1' for :009 SINAD (FM, below I 600M192 senst.q. nornonally 6uvmeter function .1100.5 nominal ,old sftenulh measurement

13.8 Volt DC Operation. The unit operates at 138V DC at IA ma. which -Awes mobole portable baseOfavation

Matching Antenna Options Matcfmnq antennas for portable nxlvte tared adrOoSO.Ons arc available drSrOfrera pdroglana. ban., closr,e,

Please ash for lull colour brochure and detailed specohcattons MA1 versions are available ovithcoverage 60-950M112 all untls now have N connectors Idled for VHF UHF and S0239 connectorsfor LF HE

The YAESU Mk2.'3 RECEIVER is only available horn HWC lid earlier models have the e.gende,,bands an.*IF rrso011e fined and receiver sonsrtiviry S metro .mprosed ple.ise end La re to, nlort, ,14 and to WC'S

Export enquiries welcomeNee. Tek 021 421 0201 eft. ereephe.)

T111107C 334303 0 TXAGWIA

ENTER 140N REPLY

Toroidal & E.I.TransformersAs manufacturers we are able to offer arange of quality toroidal and laminatedtransformers at highly competitive prices

Toroidal Mail Order Price Listprices inclusive of VAT 8 Postage15va 7.95. 30va 9.18, 50va 10.16. 80va 11.36. 120va 12.07. 160va 14.20, 225va15.21, 300va 17.04. 500va 22.10, 625va 24.66. 750va 28.75. 1000va 44.82.Also available 1 k2. 1 k5, 2k. 2k5. 3k. Prices on request.

Available from stock in the following voltages: 6-0-6, 9-0-9, 12-0-12. 15-0-15, 18-0-18, 22-0-22, 25.0.25.30-0-30, 35-0-35, 40-0-40. 45-0-45. 50-0-50. 110. 220. 240.Primary 240 volt

Quantity prices and delivery on requestAm -42&. Air Link Transformers

LINK Unit 6, The Maltings. Station Road,Sawbridgeworth, Herts. Tel: 0279 724425


QUALITY QUARTZCRYSTALS QUICKLY

Our frequency ranges are

10k Hz 50kHz 100kHz 500kHz 1MHzProtessiondl Lrystiis

also supply quartz crystal filtersoscillators of all types andcommuntcation antennae.

VV

100MHz 250MHz 360MHz

Webster Electronicsx.18NSTER. SOMERSET TA19 90A. ENGLAND

TEL:10460)57166 TELEX:46571 EMMET GFAX 10160) 57665



Electronic organgenerator system

Microcomputer -controlled design scans eproms to obtainhigh quality waveforms.

S. .1. KL\RLEY

n organ design was required twomanuals, a full set of pedals and ateast 12 stops, that was capable of

playing classical pipe organ repertoire. Thepedal notes were to stand out against the restof the sound, and 'mutation' stops to rein-force the harmonics of the note being playedwere to be provided.

Conventional pipe organs contain hun-dreds of pipes, one per note per stop. If thesewere simulated by the same number ofelectronic oscillators, the resulting instru-ment would contain many components andwould be difficult to tune.

From one point of view, this is an ex-tremely inefficient way of producing music.It is most unusual for a piece of classical soloorgan music to have more than 10 notesplaying simultaneously and since a twomanual organ has about 150 keys, more than90% of the generators are idle for the stopsselected and all are idle for those stops notselected. The number of generators can bereduced to ten if they are capable of produc-ing notes at any pitch in all the tone coloursrequired. These generators will naturally bemore complicated and their linking to thekeyboard contacts is not straightforward;nevertheless this system has the potential tocut down on wiring and components.Moreover, the smaller number of generatorswould make it easier to obtain silence whenno note was being played, and make itcheaper to apply measures to avoid start-upthumps and clicks.

A microcomputer would be the best way ofinterfacing the keyboard contacts to thLgenerators; keyboard contact wiring couldbe reduced by connecting them into a diodematrix. This would be scanned by the micro-computer which would decide which gener-ators should operate at which frequency, andsend them appropriate commands. In thisdesign the microcomputer also provides thecoupling between manuals and controls theautomatic tuning system. A piggy -back ver-sion of a single -chip microcomputer leads toa very compact computer board.

Turning back to the ten generators. agreat effort was made to simplify the circuitsas far as possible, in the knowledge that eachcomponent would be multiplied by ten.

The first method is familiar from theelectromechanical Hammond and Comptonorgans and the Wireless World design by A.D. Ryder; usually harmonics up to the tenthare provided. The technique gives naturalsounding flutes and diapasons but insuffi-cient harmonics are available for reed or

0 512 1024 1538 2048

OctaveOctave 2 1 it ;we s

1024 1152 1280 1408 1536

601

40

20

0

1

C65 Flute

I

C65 Diapason

1 1 1 1 1 11

C65 Reed

7 9 1

C262 Flue

C262 Diapason

FluteC1047

C1047 Diapason

111111 1 1 _ 111

C262 Reed 01047 Reed

I5 7

Fig.1. Spectra of pipe organ stops at different frequencies show that the harmoniccontent changes with frequency, and that diapason and reed stops contain moreharmonics the flutes, giving a 'keener' sound.

Pedal Great

Total eprom contents

...---Swell-..t.-Solo16' le 8' 4' 8' 4' 2' i'lp 8' 8' 4' 2' E' 8' SI 113'

02048 61444096 8192

14' Principal -..1-,--16.13ourdons---1-16'Basset hor1-4-.--

A//' VVW\i\ Eprom contents(4 pedal stops)

Eprom contents(1 8' stop)

Fig.2. Map of eprom contents, progressively zooming in. There is only room for thepositive half of the 16' waveform, so this is inverted for the negative half cycle. Adifferenct waveform can be stored for each octa%e.

string tone at the lower frequencies. Furth-ermore. the simpler forms of the techniquesynthesize waveforms which have the sameharmonic content throughout their fre-quency range. which is uncharacteristic ofboth organ reed stops and orchestral reedinstruments. Figure 1 shows the harmonicspectrum of typical flute. diapason and reedstops for three different notes recorded froma church organ.

The subtractive filter technique startswith a waveform rich in harmonics andattempts to use filters to cut the harmoniccontent down to the required amount. It isdifficult to make the circuit work over a wide

frequency range, since the low-pass filtersneeded to remove the harmonics from lowfrequency notes severely attenuate the highfrequency notes. Although it is claimed thatthe change in harmonic content with fre-quency enables effective reed stops to besimulated with the aid of resonant filters. theauthor's experiments with this techniquehave met with little success.

Nowadays it is quite practicable to gener-ate the required waveforms by storing asuccession of data points in ram or rom andsending them to a d -a converter at a ratewhich determines the frequency of the notegenerated. The fidelity with which the

ELECTRONICS 6: WIRELESS WORLD 817

00

N

6 0 23 Lr) 0.) CO 41) ,..,.. _ C C --I . -- C I. r.:E 7, rt3 -= E 2 "-f0. ,.1) an 0 3 = 2 _-0 ..., 7.- t...

--E 1- 0 .:1,C-) CO >., an .

i, a_ o Ir..i. - 0 ,7)""

0.)C = 6) ,, 2 '.0 .., an - CL -.V - n5 an C1)(f)

E o ,.7. = co g.) .-c 0d uo

mci' .= ccD .8 E - =;:- c ..r., a)0 _D 0 0 = cv 0LL Ual) co - .,_, cr

7-1

0 00

0

O

0

0V ca.

E

P

0

.11

CD

0 a 10 Va

waveshape will be reproduced depends onthe number of points and their resolution:compromises must inevitably be made tokeep the cost of the design within bounds.

Further complications result from thenecessity to provide realistic attack anddecay for the notes; organ notes take severalcycles to build up from zero to their max-imum amplitude, and some contain startingtransients, such as the renowned 'chiff ofsome flutes. This problem may be solved bystoring suitable starts and finishes for eachnote and 'splicing' these to the steady statewaveform, hut this is expensive in memoryspace and needs high resolution.

A more common technique is to multiplythe steady-state waveform by the requiredstart and finish envelopes by either digital oranalogue means. If the first. a high resolu-tion d -a converter is required. so I chose atechnique employing a multiplying d -a con-verter. The digital code from the eprommodulates the reference current, whichincreases slowly from zero at the start of thenote to a steady value, then decays back tozero when the note ends. The attack anddecay times are made proportional to theperiod of the note Ilow notes build up moreslowly than high notes). This agrees quitewell with measurements of the attack anddecay envelopes of organ notes taken with astorage oscilloscope, and gives a much su-perior effect to circuits which have the sameattack/decay times for all frequencies. Theeffect is important for the very lowest notes.where a slow build up is essential for realism.Three different attack/decay characteristicsare available in this design, and a diodematrix enables the constructor to choosewhether each stop will be slow, medium orfast.

The output current from the d -a convertercontains the waveforms for all 16 stops; thiscurrent is steered sequentially to the 16different stop busbars. Each busbar output isfed to the appropriate amplifier through astop switch which enables the performer toselect which stops he wants, and a potentio-meter which adjusts the relative volume ofeach stop.

The frequency of the note coming fromthe d -a converter depends both on thenumber of points per cycle of waveformstored in the memory and also on the speedwith which they are clocked. In my design,the clock speed is selected from 12 masterfrequencies corresponding to the notes C.C . . . B: the different octaves in which anote may sound are obtained by varying thenumber of points/cycle from 256 in thelowest octave to 16 in the top octave.Different waveforms may be stored for eachoctave if desired.

The different possible clock speeds meansthat the sampling frequency of the waveformvaries from about 16 to 32kHz. Unwantedheat frequencies of the waveform with thesampling frequency and the clock frequencyitself are filtered out by a low-pass filterrolling off at 7kHz.

WAVEFORM GENERATION

The organ output waveforms are generatedfrom a series of eight -bit values stored inmemory. I decided that the cheapest suitable


Highfrequency C15 a6 7

Binary counter

Multiplexer

0-10 all

DIS

113

CS

Binary counter Low frequercy

Fig.3. Binary counters increment the eprom address lines. Octave control lines C4 to C6adjust the high -order addresses to find the section of eprom with the appropriatewaveform and to force repetition of codes after the right number of steps.

Table 3 Data selection for eproms

Octave C4 C5 C6 A5 A6 A7 A8

1 0 0 0 Q10 Q11 Q12 HI

2 1 0 0 Q10 Q11 HI LO

3 0 1 0 Q10 HI LO LO

4 1 1 0 HI LO LO LO

5 0 0 1 LO LO LO LO

memory was the 2764 eprom which contains8K x 8 -bit words. The number of pointsavailable per cycle is proportional to theeprom size, but inversely proportional to thesampling speed and the number of stops.The provision of different waveforms for thesame stop at different frequencies also fillsup the memory, but I was determined toinclude this, as it is a characteristic of thepipe organs which the design attempts tomimic. Finally, the scanning speed must beslow enough to enable the samples to heaccessed reliably.

After juggling with various combinations.

R502KR,

5ti4k7

KR2 Key returns

Masterreset

R501 10k/ R500

470

81 15

.0 E t

39

Address range

256-511128-255

64-12732-630-31

I decided that 16 different waveforms couldbe stored in the 8K memory and scanned atspeeds of between 16 and 32 kHz. Thisimplies an upper cut-off frequency for theorgan of about 7kHz, which is acceptable. As16 waveforms are to be sequentially scanned.the eprom data rate is 16 times the waveformscanning rate; this gives a maximum ofabout 500kHz, well within the access timecapabilities of unselected 2764s.

Fig.5. In the microcontroller circuit theswitches control which tuning system isselected.

Keyboard control - - Frequency control -X500 4MH: Hl LO

Key0_7 keyboard scan FA -DSEL1 SEL27A___ c_n_.\

371 i i 341 221 I 1251 1113311111112610 7_Port 1 7 Xtal 0 RV, 5

10500 MK 38P70 with 2716:,orr)

61

IC501 /4 LS1313

1 I I I I -I '1

51 1 3

62 02 A B

10502 74 LS 138

C

-1 I I I I I I I

SELO- 15Synth board select

The problem now is to fit five differentwaveforms into the 512 words available foreach stop. This is achieved by progressivelyreducing the number of data points stored asthe waveform frequency increases. Since thecircuit has a fixed upper cut-off frequency,less harmonics are available for high fre-quency notes than low ones, so nothing islost by the reduction in data points. Thewaveform and the number of points arechanged for each octave, implying five dif-fent waveforms for the five octaves of eachstop.

256 points are assigned to the lowestoctave of each stop, 128 to the next lowestand so on down to 16 for the highest. For 8'stops, which sound at normal pitch, onecomplete cycle of the desired waveform isstored. For 4' stops, one octave above pitch.two cycles are stored, and 21/2, 2', 11/2' and11/3' stops required 3, 4, 5. and 6 cyclesrespectively; 16' stops present a greaterproblem as they sound one octave belowpitch, and require twice the number of datapoints. This is a problem faced by conven-tional organ builders who cannot find spacefor 16' pipes 1so called because the lowest isabout 16' long.) Their solution is to stop oneend of the pipe, getting away with a half-length pipe, at the expense of only being ableto generate waveforms with odd harmonics.and I chose an analogous technique. Half acycle of the waveform is generated, thenrepeated while the ouput is inverted tocomplete the cycle. Again, the drawback isthat only odd harmonics can be generated.In this design, 16' tone is restricted to thefirst two pedal stops by the hardware.

Fig 2 gives a graphical representation ofthe waveforms in the eprom together withtheir addresses.

The next problem is to obtain suitable datapoints to store in the eprom. A computerprogram reconstructed pipe organ wave-forms from their harmonic analyses. The

R503-508

3.71 6

11

Start Stop CK

Synth control

STPorto

!III 1 114

476

6 C503

7408

Data RJ)

Reset DO-

FreqLencycontrol

20

SEL3

t 5V

1 151 2 4 6 10 12 16

10kx6

%%.*

S500-505

DIS 1- -

10504 MC1450

01 06

31 51 71 91111131


spectra of various pipe organ stops weremeasured at different frequencies, and acomputer programmed to calculate thevalue of the waveform at the sample pointsby summing sine waves with appropriateamplitudes. For example, if the spectrum ofa flute stop is measured as fundamental80dB (to an arbitrary reference), the dBs arefirst converted to linear form (fundamental10,000 units. third harmonic 1,000 unitsand fifth harmonic 100 units. The computeris instructed to calculate the value of10,000sinx + 1,000sin 3x +100sin 5x foreach data point x in the cycle. The answer isthen scaled and offset so as to fit into the8 -bit data word. In this circuit the outputcodes vary from 0 to 255 with zero output at128, maximum positive output at 255 andmaximum negative output at 0.

The scaling should ideally ensure that thewaveform with the greatest peak amplitudeexercises the full range of the 8 -bit d -aconverter, as this will give the best possibleresolution for the system. The computerprogram finds the greatest data point valuefor the five octaves of a particular stop; if thisis k, then all the values are multiplied by127/k, and 128 added. The result is roundedto the nearest integer for the eprom.

The routine described preserves the rela-tive amplitudes of the waveform at differentparts of the frequency compass. It is usualfor the amplitude of a pipe organ stop toreduce towards the top of its range. Aproblem occurs for the 16' stops, wheremeasurements showed that the acousticpower of the pipe organ stops increasedgreatly in the bottom octave, probably tocompensate for the reduced sensitivity of theear at low frequencies. When the computerprogram above was applied strictly, theamplitude of the waveform in the lowestoctave was so great that very little resolutionwas obtained at higher frequencies. In addi-tion, a noticeable step in amplitude occuredbetween and adjacent notes B in the lowestoctave and C in the next octave. In this case Iadjusted the lowest octave amplitude to avalue nearer to the amplitude at higherfrequencies, and used bass boost in the pedalamplifier to increase the power at very lowfrequencies.

It must be stressed that there is insuffi-cient resolution in the eight -bit converter tocope with the amplitude differences betweendifferent stops. The relative balance betweeneach stop is adjusted by potentiometers onthe audio output of each channel. Becausethe dynamic range desirable for a church orpublic building is not usually suited toprivate homes, these potentiometers shouldbe adjusted to suit the application.

Even with the aid of a computer forcalculation, the described procedure forwaveform synthesis is time consuming andlaborious. Experimenters with eight -bit orbetter digital transient recorders might beable to record the waveforms and obtain therequired values directly. In this case, therecorder sampling clock should be adjusteduntil the required number of samples percycle of waveform is obtained, and thesystem gain set for each stop until the notein the octave with the greatest amplitudeoccupies the full range.

The technique of progressively halvingthe number of points per cycle for eachoctave, as well as reducing the size ofmemory required, has the further advantagethat the contents of the eprom may beclocked at the same speed for each octave.For example, assuming one complete cycleof waveform is stored, the output frequencyin the bottom octave (256 points) will be1/256 times the clock frequency. The nextoctave contains one cycle in 128 points, sothe output frequency is 1/128 times theclock frequency, so is twice as high; and soon up to the top octave. This makes possiblethe simple data selection circuit of Fig. 3 andtable 3, which permits octave selection withtwo dual 1 of 4 multiplexers, the fifth octavebeing obtained by disabling the multiplex-ers, under which conditions addresses 0-31will be scanned. Only 16 points are requiredhere, so two identical cycles of waveformmust be stored.

MICROCOMPUTER CONTROL

A block diagram of the microcomputercontrol system shows that 34 input/outputlines are required, Fig.4. Almost any eight -bit microcomputer system with a minimumof 64 words ram, 1K rom, timer interruptcapability and 21.is or better is adequate. Iused a Mostek single -chip 'piggy back'MK38P70 with the program held in a 2716eprom.

The main program scans the keyboard andcoupling switch matrix; eight output linespermit up to 256 switches to be monitored.One input tests whether the selected switchis closed and the other whether any one of agroup of eight is closed. Four output lines(decoded to 1 of 16) select one of 16synthesizer boards, for start or stop com-mands. Four output lines run in parallel toall synthesiser boards. providing start, stop.data and data clock to the selected board andone input monitors whether a selected boardis ready or busy playing a note.

A second program operates under timerinterrupt at 4ms intervals. This uses fouroutput lines to sequentially select 1 to 12master oscillators for period measurement.A reset output line resets the period counterat the start of the measurement, whichresults in a 12 -bit word which is fed to sixinput lines in two words. Three output linespermit the computer to select the low orderor high order of the period, or six settingswitches, three of which define the tuningsystem. The period of the master oscillator iscompared with the desired value, and anoutput, Hilo, derived which drives the vol-tage on the appropriate sample and hold in adirection which corrects the frequencyerror.

Fig. 5 shows a 3870 -based solution. The3870 ports can be programmed as inputs oroutputs, so most ports have measures ofinputs and outputs. A strobe pulse is auto-matically generated every time a port -output operation is carried out, and this isused as a data clock. Circuit IC503 buffersthose outputs which drive several ports, asthe computer outputs are relatively lowpower, while IC504 is a tristate buffer for thesetting switches, its output paralleled with

two buffers on the frequency locking board.The piggy -back single -chip microcomputer,while relatively expensive. results in a simplelayout which operates reliably without spe-cial wiring precautions.

COUPLERS

Pipe organs normally have couplers wherebystops from one keyboard may be coupled toanother. These are unidirectional, i.e. Greatto Pedal implies that the Great stops will beplayed by operating pedal switches but notvice versa. In this design, the 16 stops arearranged in four groups of 4 - Pedal (con-trolled by foot pedals), Great (lowerkeyboard), Swell (upper keyboard) and Solo.The last-mentioned is 'floating' in that it hasno keyboard but may be coupled to any of theother three.

When a note is selected, the computerdecides which of the four groups should beactivated by observing which manual switchhas been operated and which couplers havebeen selected. The final four bits of the serialdata stream select the required group. If acoupler switch is altered while notes are stillsounding, these notes will not be affected bythe change, but succeeding notes will be.

Pipe organs usually have octave couplerswhich add in the pipes an octave above thepipes being played and sometimes sub -octave couplers which sound one octavebelow. It would not be difficult to add theseto the microcomputer control system, buteach note would then tie up two synthesizerboards, and it is likely that, with ten boardsas in the prototype, the system would soonrun out of available boards when chordswere played.

Another possibility is automatic tr,position, whereby a tune in one key can b,caused to sound higher or lower. All theseeffects are obtained by adding a fixed dis-placement to the code for the key depressed.For example, the lowest D switch of thepedalboard has a code of 61. To transpose itdown one semitone, 1 is added to the code; totranspose it up, 1 is subtracted. To transposeit up one octave, 12 is subtracted and totranspose it up two octaves, 24 is subtracted.

The code for the high Pedal C is 32; whentransposed up by two octaves, 24 is sub-tracted, leaving eight, which is the code forcoupler switch. However, there is no con-flict, since the pedal switches are onlyscanned from 32 to 63.1f the keyboards wer,scanned in the opposite direction (lowestnotes = lowest codes) either the low pedalnotes would conflict with the coupler switchcodes, or the double octave of high pedalnotes would have codes exceeding 63 andwould play in the Great range.

A novel tuning scheme that allows theselection of different schemes for tuning isdescribed separately, together with detailsof suitable generator boards. (Availablefrom the editorial office in return for an A4envelope marked "Organ design".)

Steve Kearley. who is an honours graduatein electronics from Manchester's UM/ST. isresearch officer at the Electricity Council'sresearch centre, and specializes in instru-mentation for monitoring imperfections inthe mains supply.


Encryption is the process that turnsdata into secret form. The originaldata is known as the plaintext or

cleartext and the encrypted data as theciphertext. If the plaintext is P and theciphertext is C, then the encryption opera-tion is described by the equation

Ek(P)=C

where K is a parameter called the key whichis used to vary E. The effect of a change of keyis to generate a different C for the same P. Auser chooses a particular key from a set ofpossible keys (K) and encrypts the plaintext.The ciphertext is stored or transmitted overa channel as illustrated in Fig.1 to a receiver.One must know the particular key in use torecover the plaintext using the inverse ordecryption operation

E-IK(C)=15K(C) =P.

From the point of view of maintainingsecrecy, it must be assumed that a crypt -analyst (hacker) would have an unlimitedciphertext and would probably know themethod of encryption which means E with-out knowing the particular K. To ensure realsecrecy, two main tests to assess crypto-systems have evolved over the years:

the set of possible keys must be largeenough to make a search using each keyin turn in the decryption operation im-practicable. This is called an exhaustivesearch and is based on the fact that onlythe true key will produce meaningfultext; deduction of the key from knownplaintext/ciphertext pairs should be im-possible except by an exhaustive search.This is called a known plaintext analysis.

A function is said to be linear it if satisfies thefollowing two conditions:

f(x+y = f(x)+fly)flax) =a0x)

where a is a constant. Otherwise the func-tion is said to be non-linear. It has beenstated that linearity is the friend of crypta-nalysts and the enemy of cryptographers.This point should become clear in the nextsections.Euler's Totient Function tp(n) is the numberof integers less than n that are relativelyprime to n, that is have no common factors.For example, 0(311= 30.

STREAM ENCRYPTION

In electric terms, the plaintext is a binarysequence which consists of is and Os. Theencryption operation is a Boolean exclusive -Or logic operation. where a bit of the plain -text P, and a key bit K, are combined in anexclusive -Or gate to generate Cr The op-eration is described mathematically by theequation

C, = (Pj + KJ) mod 2.

The reasons for using this operation asopposed to an Or or an And are that C, is 0 or1 with equal frequency and that the decryp-tion operation only requires the same opera-tion as follows:

P, = (C, + mod 2

Stream encryptionA stream cryptosystem generates key sequences

containing roughly equal numbers of ones and zeros withvery long periods to appear random to an observer.

BRIAN P. McARDLE

The following example for encrypting/decrypting 5 bits explains the procedure.

Encryption DecryptionP, : 1 0 1 0 1 C, : 1 1 0 0 1

: 0 1 1 0 0 Kj : 0 1 1 0 0

P; +K; : 1 1 2 0 1 + : 1 2 1 0 1

(1:%=-1c) (q+Ki)mod2 : 1 1 0 0 1 mod2 : 1 01 0 1

To summarize, a stream cryptosystem isessentially a deterministic process for gener-ating a key sequence (lid which should haveapproximately equal numbers of is and Osand have a very long period, so as to appearrandom to an observer.

LINEAR FEEDBACK SHIFT REGISTERS

A linear -feedback shift register is illustratedin Fig. 2. Each stage is a J -K flip-flop, theoutput being taken from the last stage. Thefeedback function generates successivestates from an initial state or seed as follows:

= AKJ-1.1C)-2,..,K,) for j > r

which fora linear -feedback arrangement canbe rewritten as

K, = Liki mod 2 for j>r

1=1

Li = 1 for latch i closed= 0 for latch open

To ensure that K, has approximately thesame number of is and Os, latch r is usuallyleft closed. Consider an example for r=5with L3 open and the other latches closed.The linearity of f can be shown by thefollowing results:

f(I,0,1,0,1) = 0111,1,1,1,01 = 1f(0,1,0,1,1) = 1

The third state on the left can be generatedfrom a modulo -2 addition or an exclusive -Orbetween the other two states. The sameresult can be obtained on the right side.Thus the arrangement satisfies the firstcondition for a linear function. There is noneed to consider the second, since a is 0 or 1in modulo -2 arithmetic.

In the operation of the shift register, eachstate has a unique predecessor and a succes-sor. The total number of possible states for rstages is 2', but the state "all Os" is neverused, which means that the maximum possi-ble period is 2"--1. In the previous example,the key sequence from a seed 11.0.1.0.1) is(1.0.1.0.1.0.0.1.0.0.0.1.0.1.1.1.1.1.0.1.1.0.0.1.1.1.0.0.0.0.1) and has a period 31, whichmeans that the choice of latches results inmaximum period. Obviously, every arrange-ment of the latches does not produce max-

imum period. There are (p(2" -11/r differentarrangement where tp is Euler's Totientfunction. For r=5. this gives 0125-1115=6.

In encryption applications. the questionof secrecy must be examined. If a crypt -analyst knows the value of r and the last 10bits of the key, then

0 = (1.1.1.0.0)0 = (0,1,1,1,0)0 = (0.0.1.1.1)0 = (0.0,0,1,1)1 = (0,0,0,0,1)

and the five equations

0 = (L, + L2 + L3) mod 20 = (1,2 + L3 + 1,41 mod 20 = (L3 + L4 + 1.5) mod 20 = (L4 + Ls + Lti) mod 2

1 = L, mod 2

Fig.1. The encryption/decryption operation

Fig.2. Linear feedback shift register

HKj

Non l inear feedback logic

K;- K;-2 K-rt

Fig.3. Non-linear feedback shift register

K

Non linear feedforward logic

J-1 K J 2 -r+1 r

Linear feedback logic

K

Fig.4 Non-linear feedback shift registerwith ion -linear feedforward

ELECTRONICS & WIRELESS WORLD S2 I

From the last equation. L5=1. Therefore theprevious equation can be re -written as

0=1[1+1) mod 2

Continuing in this manner gives = 0 andLi = L, = 1. The mathematical problem canalso be formulated in matrix form such thatthe solution requires the calculation of theinverse matrix. This method is more suitablefor large values of r because a cryptanalystcould make use of computer packages. Butonce a latch arrangement and a state for theshift register are known, the full sequencecan be generated. Consequently, the linearfunction, irrespective of the value of r, is toopredictable for secrecy and does not satisfyTest 2.

SHIFT REGISTERS WITHNON-LINEAR LOGIC

A shift register with non-linear feedbacklogic is illustrated in Fig. 4. The incominghit is generated according to:

= f(K, 1, K, 2 , K, mod 2

where f is a non-linear function. Normally fis chosen such that

KJ =1'1K, 1,K, 2,...K, I + K,.,1 mod 2

which ensures that (K,) has approximatelyequal numbers of Is and Os. The newfunction f' is also non-linear. Consider thefollowing example.

K,=11(,_11(,.4+ K, 2K,.3+ Ko) mod 2

The non -linearity is easily demonstrated:

R1.0.1.0.11= 1f(0,1,0,1,0) = 0f(1,1,1,1.1) = 0

An exclusive -Or between the first and secondstates on the left gives the third, but thesame operation on the right gives 1 insteadof 0. Thus, the feedback function is non-linear. It appears to be more complicatedthan the linear arrangement and a readerwould be justified in assuming at first glancethat it is also more secure. For a seed(1,0.1,0,11, the key sequence for the first 31terms is: (1,0,1,0,1,1.1,1,1,0,1,0,1,1,1.1,1,0,1,0,1,1,1,1,1,0,1,0,1,1,11. But the period isshort and the sequence repeats more thanonce. A more detailed study shows that thesequence of states has six cycles:

L 00.00-,..00-11000-01100-00110-00011-00001-1

L 00010-10001-01000-00100

4L 10101-11010-11101-11110-11111-01111-1011 -01011-1

L 01101-10110-11011i

L 00101-10010-01001-10100-01010

L moo-oino-ooni-10011-11001

An interesting point is that the state(0,0,0,0,0) does not generate itself. From thepoint of secrecy, the key sequence has period8 and repeats three times within the 31terms. Thus, non-linear functions are notnecessarily more secure than the linearvariety. But no such cryptosystem, irrespec-tive of the period, offers any real secrecy. If acryptanalyst knows the feedback function,he can generate the balance of the sequencefollowing known bits. Consequently, thisarrangement also fails Test 2.

The arrangement of Fig. 5 attempts toeliminate the two main weaknesses. Thelinear feedback function generates the in-coming bits to the shift register and conse-quently determines the period of the entiresystem. The non-linear feedforward functiongenerates the key bits and protects the shiftregister from an analysis using known keybits. The two previous examples combinedresult in:

K,=1K,.,41ii... +K,4.3K0.2+Ki) mod 2K,4.5=11(i.,4+Ko.3+Ki4.1+K,) mod 2.

For the seed (1,0,1,0,1) the key sequence is(1,0,1,0,1,0,1,0,0.0,1,1,1,0,1,0,0,1,0,1,1,0,0,1,0.0,1,0,0,1,1) and has period 31. Thesecrecy depends on not being able to deduceK, to Kj+4 from K', to K'; +4 (assuming aworst -case analysis where both functions areknown). In this simple example, such adeduction is turgid rather than impossible.But an actual commercial system wouldhave r = 128 or 256. In designing such asystem, the feedback logic would be one ofp(21. -11/r latch arrangements that produce amaximum period of 2r-1. The non-linearlogic could be chosen using a variety ofmethods, but the following is simple andstraightforward. The output or inverted out-put of such stage (flip-flop) is connected to

K j_

Fig.5. Non-linear feedforward loop for r 8

an And gate with the constraints:- each gate has only two inputs;- the span of the inputs does not exceed- the number of stages.

For example, the pairs (1,4). (2,6), (3,8)and (5,7) satisfy these conditions. The re-sulting function is not necessarily non-linear in the strict mathematical sense, butfor a large value of r would be quite compli-cated. A cryptanalyst would have to discovera method of solving equations which havenon-linear operations or use an exhaustivesearch. For r=256 this would be impracti-cal. Therefore this arrangement is reason-ably secure. A useful hint for designers is toinclude a facility to vary the logic arrange-ment such that the same functions are notused too often.

CONGRUENCE GENERATORS

This method is essentially a computer algor-ithm for generating a sequence of numbersfrom a seed X. as follows.

Number sequence Xi = (AX, +B) mod MBinary sequence KJ= X, mod 2

Thus K, is 0 or 1, depending on whether Xi iseven or odd. The modulus M is the largestprime that can be fitted to the processor'sword size and A and B are integral powers ofsome prime factor. The generated sequencesare periodic because, once the seed is re-produced, the complete succession of resultsafter the seed also repeats. For example, X.= 8191, A = 13077, B = -6925 and M =32767 has period 1050. But the simplicity ofthe operation is also its weakness. Generally,A, B and M are machine constraints, whichmeans that X. is the only parameter to bevaried. In addition, some seeds have shortperiods which, in turn, limits the choice ofseed. In applying Test 2, it must be assumedthat a cryptanalyst would have all the detailsexcept the seed. Thus, the secret depends onthe difficulty of deducing X, from K,. Forexample, if K, = 1, then for a 16 -bit word X,is an uneven number between 1 and 216 -1.An examination which would try each possi-ble number in turn would not be impractic-al. Consequently, this system implementedon the standard desk -top would not satisfyTest 2.

X2 MOD PQ GENERATOR

A sequence of numbers (Xi) is generatedfrom a seed X. as follows.

Number sequence X, = Xi_i mod NBinary sequence KJ = mod 2

The modulus N is the product of two largeprimes P and Q that are congruent to 3 mod4 and mod (Xe, N) = 1. Therefore is 0 or 1,depending on whether X, is even or odd. Theset of possible seeds has 0(N) = (P-1) (Q-11elements, where .p is Euler's Totient Func-tion. The main difficulty is implementationof desk -top microcomputers which normallyhandle 12 digit numbers. For secrecy, P andmust be very large primes in the order of 100digits. The method is therefore considered tobe suitable only for Public -key Encryption,which is not covered in this article.

S22 ELECTRONICS & WIRELESS WORLD

Data capturefor Fourier analysis

In coverage of computer FFT analysis, little has been saidabout getting data into memory. This capture system

digitizes in 10p.s.

Biomedical engineers use Fourier trans-forms to characterize and analyse avariety of signals such as electro-

cardiograms, electroencephalograms orrespiratory waveforms. If the transform andits reverse can be performed rapidly, thenselective attenuation of frequency com-ponents provides a versatile filter.

Dedicated components are available forthe Fourier transform, but microcomputerscan also perform satisfactorily once the datais in digital form. Speed limitations of thesoftware option necessitate use of the fastFourier -transform algorithm.

The widely used fast -Fourier -transformalgorithm requires a predefined number ofsamples of 2"; n is the number of computa-tion arrays, and hence passes through thedata, produced by the calculation'. Runningmachine language on an eight -bit computerit is convenient to have n=8, i.e. 256samples.

Another requirement is for an integralnumber of waveforms in the sample windowto be analysed. If this condition is not met, adiscontinuity will be introduced and this canhave a profound effect upon the frequencyspectrum, causing erroneous non-existentcomponents'. Furthermore, sampling fre-quency must be high enough to avoidaliasing.

These constraints make real-time sam-pling of biological signals difficult since theperiodicity of the waveform may not beknown prior to sampling, and indeed maychange from beat to beat. It may not provepossible to set the sampling interval in orderto generate the required number of datapoints in an integral number of waveformsprior to sampling. This data -capture systemfor a BBC B microcomputer was developedto sample real-time electrocardiographicsignals so that they can be analysed infrequency domain.

HARDWARE

Circuit Fig. 1 gives you control over thetrigger point by setting signal polarity, gainand offset with a choice of peak or leveldetection. The number of cycles of thewaveform can be selected and the startingpoint for sampling may be delayed from thetrigger point. An input to the data capturesystem starts sampling, PB, and an output

PETER FINCH and DAVID TAYLER

In

inverterFollower

12k

12k

&finalpdarity

TL064

10k

81.2

Offset

*VV\o--276

27k e-'N.A.",--

V

Ilk

18k

1

-5V

27k.1: 27k

Peak defector

.5v

106 '3

12

Pea I level r.gg,

1(92

4093

41 14

iC2

4047

3 1M

270n

11

'2

14

*Starting point delay

16

1C3

4018

10

5V

UserportBBC

Fig. 1. Counter and analogue circuits fordata capture. Line no from the computerstarts sampling and input FBI flags theconversion beginning and end.

from it, PB,, flags the sample waveform'sbeginning and end.

Conversion time of the ZN448 analogue -to converter is 10p.s: it is clocked by the1MHz bus and software controlled, Fig. 2.

SOFTWARE

Four menu options are provided by thesoftware. In order to free the maximumamount of memory for a data buffer, theprogram runs in mode seven and has asimple screen display, indicating the optionselected by a solid box next to the menuitem. Options are;

- setting sampling interval, in- manual setting of PB high for setting

trigger level and data -capture systemsensitivity

- waveform sampling- saving the compressed sampled waveform

as a disc file.

Waveforms are sampled by the a -to -d con-verter at a rate dermined by timer one of the6522 versatile interface adaptor. All inter-rupts on the microcomputer are disabledduring sampling (by setting the interruptflag), which avoids intermittent stretchingof the sampling interval by unexpectedinterrupts. Free running mode is used forthe timer, and the time interval isloaded into the timer latches. By continuallypolling the appropriate bit of the interrupt -flag register (set for no interrupts), a time-out is detected whereupon a sample point isread from the a -to -d converter latch ataddress FCF3. Another conversion is im-mediately initiated by a write operation toaddress FCF5 so that the data may be readyat the next time out.

Data capture requires two control lines,provided by the two low -order biti of port Bof the computer's 6522 v.i.a. When you


Digital board

LLD

5V

1

AO-

3

443

6

7FCr7-

74LS138

5V

16

15 FCF014

13

12

11

10

A analogue ground

D digital ground

decide to sample the waveform Plif, is drivenhigh, enabling At the trigger point of thewaveform selected, PB1 goes high and stayshigh until sampling is finished. During thistime the a -to -d converter is sampled at thepreset sample interval and data from it is

stored in memory. Data is now held in a rambuffer, the start address of which is constantbut the size of which can be up to 20Kbytedepending on the sampling interval andperiod.

When PB1 is driven low by IC3 to indicatethe end of sampling, PK, is programmed lowagain ready for another sampling cycle, anddata in the ram buffer is crammed into a 256byte page buffer so that it can be saved ondisc for later Fourier analysis. This requirestaking every n/256 sample point where n isthe total number of sample points stored inthe ram buffer. List 1 shows how this can bedone.

LIST 1. Basic and its machine -language equivalent10 REM Waveform in RAM between BUFFSTART and

BUFFEND15 REM Result in buffer from PAGE to PAGE +25620 Y 0 :REM LDY a025 Index = buffstart :REM LDA buffstart

:LDX buffstart +130 :REM STA index

STX index + 135 counter - buffstart :REM STA counter

:STX counter +140 POKE (page `1). :REM Assign LDX #0

PEEK (index) :LDA (index.X)(MBasic]

45 REM (page)?Y -?(index) (BBC Basic]

50 index = index+ 1

55 counter=counter +256

60 If counter- buffendTHEN 50

65

70

75 counter counterbuffend buffstart

80

85 Y=Y+190 IF Y-3256 THEN 40

:REM STA (page).Y

:REM .1p INC index:BNE skp :INC index +1:REM .skp INC counter + 1

:REM SEC :LDA counter:SBC buffend:REM PHA:LDA counter+ 1:SBC buffend + 1:REM TAX :PLA:BCC 1p: CLC:REM ADC buffstartSTA counter :TXA

:REM ADCbuffstart + 1

:STA counter -3 1:REM INY:REM BNE assign

a - to -d mnverter

1MHz

18

2 17

3 16

4 15

82k-5v -A., 5 ZN448

DO

14

13

12

11D7

105V

741

Fig. 2. Analogue -to -digital conversion is

partly controlled by the BBC computer1MHz bus. Conversion sampling rate is

under control of the 6522 v.i.a.

To allow repeated sampling, the pagebuffer can be saved on disc under any name.Up to six characters for the filename aretaken from the keyboard using the routineOSRDCH (&FFE0 ) and stored in a

command -line buffer containing "*SAVExxxxxx 7800 7900". Command line interpre-ter routine OSCLI (&FFF7) is then pointedat the command -line buffer and called.

We have used this data capture system onboth BBC model B and Master computers togenerate 256byte arrays of real time electro-cardiogram waveforms, which we haveanalysed using a machine -language imple-mentation of the fast Fourier transformwritten by Peter Finch. The program usesfloating-point arithmetic, executes in typi-cally three to four seconds, and is capable ofreverse transforms. Using our implementa-tion of the FFT and its reverse transform,we have demonstrated the advantages oflinear -phase high-pass filtering to stabilizethe baseline of electrocardiograms withoutintroducing insignificant waveform dis-tortion -1.

References1 Oran Brigham. E.. The Fast Fourier Transform.Prentice -Hall 1970.2 Omer. W.. Faster Fourier transforms. Electro-nics and Wireless World. June 1986 pp.23-5.3 Tayler, D.. Finch P. and Vincent. R.. Linear -phase filtering - a new approach to distortionfree electrocardiograms. Computers in Cardio-logy. IEEE 1985. pp.283-6.

David Tayler B.Sc.. B.M. is a G.P. and clinicalassistant in cardiology in Sheffield. current-ly investigating the effects of electrical dis-tortion on e.c.g. waveforms. Peter FinchB.M.. B.S. is registrar in gastroenterology inLiverpool with an interest in programmingfor gastroenterology data management.

CONTROL SOFTWARE

Peter Finch and David Tayler's listingsfor grabbing data from the a -to -d conver-ter, drawing a saved page and dumping amode -four screen are available. Sendeither an Acorn -format disc (preferably80 -track double sided) and return post-age or a large s.a.e. to E&WW Editorial,Quadrant House, The Quadrant, Sutton,Surrey SM2 5AS. Please mark FFT-TF onyour envelope.

Risc card for PCsBesides launching the Archimedes compu-ter (see page 839). Acorn has also made itsrisc architecture available in a single -boardcomputer to plug into the expansion slots ofIBM PC -compatibles. The so-called Spring-board contains the full risc chip set with theexception of the display controller andcomes with 1Mbyte of ram (at £1000) or4Mbyte (at 12000). The processor's designavoids the need for costly high-speed staticram.

Up to four Springboards can be usedsimultaneously, each running independent-ly of the others and of DOS, though the PCcan be used as a terminal. A bus connectoris provided on the board to eliminate delaysin passing i/o via the host computer's bus.Development software for the card includesan assembler, utilities and other tools.High-level languages are an Arm C compil-er, Fortran, Prolog and Lisp. Details fromAcorn Computers on 0223 214411.


TRANSFORM YOUR CONVENTIONAL SCOPESWITH THE SCOPADAPTOR INTO A

2CH. 20KHz FFT ANALYSER

+DIGITAL STORAGE SCOPE4 MODES OF OPERATION02 CH. FFTA.00-50Hz Span to 0-20kHz Span.OFull ANTI -ALIAS filtering.0200 line resolution.040 dB dynamic range.OSelectable LOG/LIN Amplitude.0Hanning/Rectangular weighting.OFast update speed.OUp to 128 averages + peak.00verload indicators.

02 CH. D.S.O.050kHz sampling rate (max.)00.5% vertical resolution.OTimebase from 5 sec/div to 1 msec/div.0Comprehensive trigger facilities.OSingle shot.01% or 25% pre -trigger0512 point horizontal resolution.ODisplay 'freeze' control.

OSPLIT MODEOSimultaneous display of frequency and time

domain.0100 line resolution for both frequency and

time traces.

ODUAL MODEOTime and frequency traces correlated.OSame samples used for both domains.OSingle shot mode.OSample rate up to 100kHz.0200 line resolution for both traces.

11111111111111111.1.1

411,sr- ar 4r, Imo Ems el

armIKIP

a

SCOPADAPTOR L

JelF1111PPIpir

A REVOLUTIONARY NEW CONCEPTIN TEST AND MEASUREMENTOTrue real time F.F.T. analysis.0Full Anti -Alias Filtering.OD.S.O. with extensive triggering.OPeriodic and transient waveform

analysis.0 RS423/232 out.ODirect connection for plotter,

printer or computer.

DATA ACQUISITIONLIMITEDElectron House. Higher Hiiigate.Stockport. Cheshire SKI 300Tel: 061-477 3888 Telex: 666839DEVELOPED ANDMANUFACTURED IN THE U.K.

ENTER 30 ON REPLY CARO

TAYLORVESTIGIAL SIDEBAND

TELEVISION MOD JLATOR C.C.I.R/ 3

Power RequirementVideo InputAudio InputF M Sound Sub. CarrwrModulationI F VisionI.F.SoundSound Pre -EmphasisRipple on I F Saw FilterOutput lany channel 47860MHzIVision to Sound Power RatioIntermodulationSpurious Harmonic Output

C C I 8/3 SPECIFICATION

- 240V 8 Watt lavailable in other voltagesl- IV Pk Pk 75 Ohm

8V 600 Ohm- 6MH, lavailable 5 5MENI- Negative

38 9MHz32 9MHz (available 33 4MH7I50us

- 6dB- 6dBmV 12mVI 75 Ohm- 5 to I

- Equal or less than 60dB- - 40413180118 if fitted with TCFL I Filter or combaled via TCFL4 Combiner Leveler

CCIR 3 1 Specihcalion as abort- . .: 60dtim Vt000MV

CRYSTAL CONTROLLED OSCILLATOR19r" RACK MOUNTING, lu HIGH, 205mm DEEP

(..HANNE c COMBINER FILTER/LEVELLERto combine mama, of modulators

TM. 2TCFL47500

2 Channel Filter'Cornbiner/Leveller Insertion Loss 3.5d84 Channel Filte,Cornbiner/Levellet Insertion Loss 3 548Enables up to 4 ,TCFL2 or TCFL4 to be constant

TAYLOR BROS (OLDHAM) LTDBISLEY STREET WORKS, LEE STREET,OLDHAM, ENGLAND.TELEPHONE: 061 652 3221 TELEX: 669911

1.:NTER 190N in.:IrLy crimp


LANGREX SUPPLIES LTDClimax House, Fallsbrook Rd., Streatham, London SW16 6ED

RST Tel: 01-677 2424 Telex: 946708 RSTSEMICONDA A 1 )9 01044530 01744713 03044715 03044717 030A( 107 0 554(175 0 3543176 0 35A(177 040A( 118 0354(141 0354(141K 045A( 147 040A( 1 471( 0454(176 035A( 187 0 35A( 188 0354(517 7254(518 1 554(519 180AC520 150A( Y71 I 554(739 40040149 100AD 161 05040167 06040711 12 SO40717 I? 50AF106 0604314 350Af 1 1 5 350Af 116 35041117 40041139 055Af 186 0 75Af ?39 06541711 3 75417 1? 50006,770 40

UCTORSA5116 .20045717 I 60A5770 4 SOA5721 4 75AUY10 350BA 1 45 01384148 01584154 006BA 1 55 01184156 006BAW62 00584x13 005MA16 0068(107 01?B(108 0 13B(109 014B(113 017B(114 0178(115 0128(116 019B( I 17 0748(118 030B(175 02513076 076B(135 0188(136 0188(137 0 728(147 0178(148 0178( 149 01?8(157 01739158 0138( 159 0173( 167 0108(170 0098(171 0118(172 0098(173 0098(177 0158( I 78 0 788(179 015

B(188(1118(184B(717 0118(713 0118(214 0 1 1

B(737 0098(738 0098(301 03613(303 0368(307 0098(308 0098(377 0098(378 0098(337 0098(338 009B(030 75013(531 7508(037 7508(533 7508(034 7508(539 3608(540 31308(542 0328(543 0458(558 0758(570 0118(071 02133(571 0718(711 35080115 03580123 23080174 75080131 0426013? 04280135 07780136 02780137 03080138 03080139 030BDI 40 0308014480181

09100

740I 5)I 5)0300 160 190170200700 360300 450311031)0300 300 750300 300 3003(101501501601503.101'0170 3503u0300 300 3u030300a 00

H3598 0308rW10 1 04BrWl 1 10181084 07881 085 0 78131580133088313554)81058305?!(106.1

0 780 780 780 78018036065

9 0270770 77

,6 165VOR 303

I:0-5 1 70

200042015015017

010

1300

011

164

40 0640 0 '

60 0,3.m 311

0105416139GM

Mlr 34038)33703811371

r41(520381521

04137965

6,1133055

3838 103

34P3104

MR 105MP54063.41.3456

MP54101

04111.51.106

91115U56

9355561/1.101681403661404045047041004470670067904810685049004910495047(004707

42i1

'101

0 790 731050 750 732252000 550 550 55017017

I II1 33171

04540)3504001 750 750 55015on021021

tX260(780(79035 4000(36 400IX41 1:00(47 1500(43 1500(44 1250(45 0860(71 0650(77 7200(73 1450(74 1400(75 1400(76 1600(77 2750(81 0900(817 400(1(311 0950(83 140MBA 1400(172 6500(173 6500(139 17000(40 18000 141 1800

170 4.0171 440AXI 400

5201 550. 107 550

203 550204 7(A)205 1000706 850107 1800

.071 2500017 740

120088 1 95'..1x)9 2S

1

55144(1

11030A

TIP314

11537411P3.34

TIP344110714

11042411P110TIP1 17

11P175 035IIPI30 045

110131 04511013? 04311P135 045115137 048110140 085TWIN OBS111,147 0851157956 060111119551 0451153055 060TiP30555 0 451515) 07515170 07175178 05175277 07325778 0 67113107 012713108 0 177191097943002943017113077314303 014710304 014718311 0 13718314 0 25'two 014

0 750030030040040010040040040050060010 160 1 t

004011.0 I?1(517,1 Su

1511

031:0360357502107202'0350351 909330.1030.12(.

458S

0303.07:0720707 50

073023074075

. 5,4 030030030

6 On^17 072

017072017

63 030'4 0 55,5 070

060III 075

1006 614 500

0I0101I

0I01017

'13 0 IQ1,

6 00 0101

III5 190S 010

36 01064058 0 17

.944059 0107144060 017764061 017.944062 0 15264174 0132141176 0 137647413 025764/88 0 IS264789 01?7144400 0172144401 01?7144407 017785457 04S7915458 040765459 04075017 160025019 750025024 350075075 350075016 400015103 25025307 550253113 55015372 50025325 40075701 1750757454 1 75

757464 175

VALVES

.15047293 16 0047476 35 00414:1521 75 0071629130 15 0043343 445004731 7 754741 7601310448 1 14 9083484 165 008590 580085810 60 00BT 5 58 95811 7 185 008119 44 058179 349 15BT69 354 808196 17990(8131 400(133 400(Y31 300( IK 70 00(34 22 00(3.14 27 000441 25 00DA41 18 70D AF91 175DAf 96 175DE177 35 00D1774 49 000191 175Df 96 17501191 175131197 700DK96 1 750192 7000( 94 175D196 17501510 14 0001515 12 50O 1516 17 5001519 12 50DM70 700DM71 700DM 160 4 75D987 1 5009807 1 SO1'551 50 00910(( 1517038000 1127NOT 14 191801 13 73fHICC 8003611 13 27187« 8 10193(( 8401860 876188( 825388(( 10 33(9(01 947191)3 9901919 6 25397(( 8 751991 9 III

1.301 18 50

E180(( 10 501180 17 05

182(( 13 75(1868 11 501188(0 89117801 77 513283(C 17 003788C( 17 5018103 35 481452 110 00(476 750FA8(80 1 7514(91 3503/1147 750141801 7009841 4009391 15018(33 75018(41 75018(81 1 50(8(90 1 75(8180 15098183 17518189 1 6018131 100(11(90 175

1(92 1751(I57 46700FC(33 450R(35 450fC(40 550EC(81 175EC(82 175FCC83 1 75(((84 1 75FC(85 175(((86 250(((88 3 50FC(89 700(((91 893FC( 189 7 101((807 I 751C(8438 5005(180 1501(387 I 50F(086 75010335 300R0141 3501(0181 300R1483 7501(084 7501(180 1501(181 1751(182 I 501(183 300R184 I 50F(185 15)1(186 1 7510374 5001139 7 751140 500154 1 350(147 450

5() 2 501(5431551160(183

1(85Ef 86 3',.,ETV 2 SO1191 7 951197 637EF93 1501194 750EF95 599FF98 7001( 183 700(1 184 700118045 12 00108055 15 001(490 1751690 I 50937 750(133 400113444111 400936 750Flat 2801147 7501181 5 751(83 600984 725(186 2 751190 3 251191 7391195 2001(156 30001(360 18 5011500 30031509 1000((871 13 005187? 139731180 75 001M34 8006M80 2501548 I 7 SO1M84 700FM85 2 751M87 750(3437 75 451DWI 6511997 6961551 7 75(081 750(083 750(Y84 9 24(0136 175E588 175(55004 30015807 1751735 I 75T740 7501741 2501780 1501781 1 50F790 3001W450 3501/644 800 350GI 371 K 35 00G551K 17 0001802M 17 506740 70 1? 006400 11 I 750(.N4 900,,N4A 900

16151

15 350030

75 40700070 003004764054754475so,

15 01)

7444437761

9041

2044)

7 562 5,

"4 173.1.1.) 87

98.'969

4,3 107510 43

66 6868 I(815800

4) 10 3(:I6 11 75

7 17 75I 618

6 Si)8 75

11 7X11'0 650

9 35II 75

61 875"1 500

10 85

7 0511 37600630450

14 03115 00

250'9 6500' I 86(X)15 61 50.1 18 30.1 150 00

41 175 0019 5026 40

oU164 0078 SOIS 00325? 506 (X)7504 35

46 7X11.1

5

114

'01

01()DJ074 3 5.P(86 7510(88 2 93P(95 175P(97 I 750(9(KI 1 75P5(84 150PI 1115 1501 .8 700

H9 1 7589 750

3;)5 1 60306 160

132 2(K)30 7 00

1501,16 750

4.7 700I ..00 3 75

01 3 25301 7506.07 7 50

. 3061 70

1,506 110308 170

(I1 03 300

200415 750

1H6 75030571(5 2 5000 600,4, 4061 45 00200 750

1 1 16 2 SO0481 175P1814 700P487 1 50P683 7 50P184 700P1504 6 750P1508 550P1509 600P1519 600PI 801 1 50P1802 600P533 7 50P581 1 50P587 150P583 175P588 700P55004 400P5800 157PY80 1 15005451 87 SO005107 6 38 00

00,103 10 76 7500503 700

00906 401

001107 50

00206

4838

1600

7665

67 701XI37 17 SO

OVI

. 165 603 175004 750 71

084 400(7(5 SOO 7I4,005 mows

5

1171.16 10 4 .

RIO 1.

017018

0)9170RG3 1750RG3 7506RG4 1750RG4 303012113 750 4RR3 1750 41

51305130P6197813 4) It

510700 81)50415U47 ID1003 101003 J1115 5.

1171 3'112111100 6161 175194 40o114 50105 50106 80106 5ar0A

766 500083,

196 5000,7757.

167 6000461

197 6000W5/.

3140111820U19025076U37U48(8000341684108(4i081890(84111(8511(1800102IA 1181

1)(18714183

u

45(890410843.480,(41.185

5631,,1.7500

.5 500,2 6103

.81 16004

1 37(X/341 32004.R1

165130. "1170 395 00" 1740 58860

'33031.1 7500'765 7500"611001 800'411020 900,111070 903'511072 900'M1(123 900'511040 19 36'50011 16 66'381042 17 77

.03GT 300,14 7500. 364 4800

6400I 751 75

14 1 75.4515 II 50

394 woo43 7000

'21 325126 875147 9800

'155 265001704 3831101708 33600

'475 17500,6007 11500'007 10000

AS 30014 11 00

978 15353179 1005,17.103 17 50.824138 17 501(73 2500,f74 1700

8100.454700

1129 4500154 700'44 (75

654 60001/SA 60001506 8000/,00 N70,

836040050451S(10.11(1113751114

825538547?'11803'11460`41.16 .1111.1

13468 2503'6/46 250'8361 25073 100574/ 250U4G1 7506 3026,084 1756487 3(5164(7 30060144 4156461 317064146 500641(5 59964146 7506415 I 5064.14/ 40064145 96964486 60768145 4 7567.684 3506405 3 256456 8666457G 8756416 12564031 50064386 750640564 4506496 150641561 303667 375693 3256046 1506847 50068484 4006864 4066406 1568146 2 566816 2256804 4 50686 11500681761 4506896 115 0008346 70D68374 35068(7 60068138A 3516857 60068W6 60068W? I50681761 50068:6 7 756(064864

9664

153047'30,

7 51,

72'30

S'.I'89'47'7I2 7'314

30'1301,

5035

(Al,5 7'3(1

75075'25,30,4103 7'

100307779311'37

3030134402 7'39,

30,1 7,

5,40,75,1)3,

35I.

70,50

17

17

4

401,5

. A

t).

1 75300

ITT 780065001000

1700047 5067 75

7007002001 60

7 138180

150200700750750300180

15 I801 100

1506054507 506 45

20002000603

54

15 IS2000700084S650

353156(X)

35001250025004500

3 75183347516500193 163503700060)18 50.100350751

300003E0001100015000170001650D

900300

50

7800917 600

6696 4505718 7 505775 5505776 II 375777 7055749 7505751 4005763 45058144 4005810 4005847 1211D

5876A 31 505879 SOO

5886 17 505963 2505965 3506005 7 756021 4 SO6057 12 256058 17346059 6006061 6006063 3 756061 10 756067 11 756077 6006080 140060974 7350060978 7000361464 170)61468 1201361698 19006189 9506701 11 406447 20006550 1000681330 12506973 7507076 15070774 8007651 6157586 15007587 73007609 56007868 6037895 17008068 16508136 7508417 60018017 II 5318045 1019131046 II 53

Irslad

0.04(tomt4( 07508 850(0lu(cr stud.,

Pt*, UVapp1.0000

BASES81, 00554 551

0411874.51'1.041

0 So894 thrtlif

0111894 610 tell

(1 50890 0 55Do (9 101 0401..101 0 SSNov.6110 181,...

00Volvg ..... 101$ 8011111 ..1/0, O.11)

CRTs203,1?BPI3110)31)P1394131 P73(5P131161

'((P731P73KP1.1RP I

3WPI .1011,54051 S50064P1 100)

'APIA 4110.5P1 SA 1501ill07 75 001X.7 5 63 3:'D47 31 58 07067 37 58 071867 36 65 (X,0113 91 56 830117 II 113 I?8(8138 1 110

V(121386 12 508(8)39A 31 00

V(125178 10110V(R5I7( 1000

A01...1%

06 11

11111 11131

put 101,Ion lop

INTEG74(10 0 167400 0.157401 0366407 () 367403 0367404 0477405 0 477406 0.187407 0 5574011 0367409 0 31,7410 0367411 040731? 0477413 036

RATED CIRC741674177470747774237474747774787430743?74337437/438/4J97.140

UITS'441 048/441 175'150 0.10'151 030746.4 1130/164 11 10

'$601 0 10'470 048'177 0 10'473 0 40/474 036475 1165'$76 11.11'480 0,1:'483 II 1-

/4156,I,74,1 0 367497 0 S47493 0 647494 0 7117495 0 657496 061,7497 3 9117,3 III 08.1'1 07 036

0 540410 511511 ?II

tl0600600540 55

II) 0 10II 046I I 13015 0 7717 150

140160036140

600 ix

74159 175741 70 1 70741 77 400741 73 0 77741 74 060741 76 (X)741 76 0074180 4874190 7074191 7074197 1074193 06074194 11074195 0 7274196 10074197 10074198 77074199 770

TAA570TA/66341S144 7(0)7844800T845700184530TBA55001845600784673TBA7130184 7000TBA7500184800184920T849901(47700T(476134

1 751 IS301

SO

60So7575755075500075500075

1.44.76110.01,.. t \111 16 04.0642 and pat mg s.41( Jual lnl.und0aul, 6I1111 16 1 01a1.4 ( RI. 1.1 41 1061105 . 1 34141 !-he.. 011096 .11 1004 01 (1 541.0511

10 5.005 1,11, A0111.101 .11)11 I .11, 10 111.111100w 4ti,,e10,(1 1'10, .1 \kilt II itt 111 t s.6110111111J9 11111\ .1\ AIL 11141. ill dill I \ 11111111111t. k gt, \ III ( 12, .1.011,01,110. 11 ,111m 40,10 old( 1.

( 11.1 10 10111 I\1,..1.1 I\ k. anti lllll ttlititititt,111411\1 (Attlitlitt1,1111 M.\ \ Ittt, 11.14,1 5 \

Telephone: 01-677 2424/7Telex: 946708E. & O.E.Open to callers Monday -Friday 9 a.m. - 5 p.m. WW



114EcsgamerrropicsQuadripartitedigital cellular

agreementThe UK, West Germany, Franceand Italy reached agreement onthe introduction of a pan-European digital cellular radioservice. At a meeting in Bonn,West Germany, on 19 May, theMinisters responsible for tele-communications and their rep-resentatives signed a declarationsetting dates and targets for thenew service which is to be estab-lished in these four countriesfrom 1991. British TelecomsMobile Communications andRacal were among the co-signatories to the QuadripartiteAgreement.

An out!Me specification hasalready been prepared and thereare not thought to be anysticking -points. Consequently, itis expected that the specificationwill be "firmed -up" within thenext 12 months. The total equip-ment market for Western Europealone is estimated to be between£500m and £600m per annum.These revenues being dividedroughly equally between infra-structure equipment and themobile phones.

Coopers &Lybrand lead

networksecurity study

A number of the leading Euro-pean companies in informationand telecommunications tech-nology have agreed to participatein an EEC study of the security ofnetwork systems. The motiva-tion being the increasing recog-nition of the increasing depend-ence on information technology.The urgent need being for sys-tems in which owners, operatorsand users can have confidence interms of the confidentiality ofdata; the integrity of data asregards error and fraud; and theavailability of the system.

The Europe -wide study, with abudget in excess of £700,000,has been awarded to a consor-tium of five European memberfirms of Coopers & Lybrand andAdmiral Computing. The resultsare scheduled for this Septem-

ber. The study is being funded bythe European Commission,sponsors, Case -study organiza-tions and C&L itself. Sponsorsinclude the Italian and DutchPTTs, GEC, Plessey, Olivetti, Sie-mens, DEC and Hewlett-Packard.

DTI ban oncordless phonesThe Department of Trade andIndustry has announced legisla-tion to make the import, manu-facture, sale or possession ofcertain unapproved cordlesstelephones, operating at fre-quencies below 853MHz, illegal.This will remove the anomalywhereby it was illegal to use suchinstruments, but legal to sellthem.

Unlawful equipment can beseized and offenders face fines ofup to £2000. By outlawing thesetelephones, the DTI aims to re-move a considerable source ofinterference with legitimateradio users, including theemergency services.

BT opto spendin the City

British Telecom is investing £40million on the second phase of itsCity Fibre Network (CFN). It willprovide major customers with asingle access path for all servicesand will reduce the time taken toprovide private circuits to cus-tomers. Extra circuits can beswitched in immediately fromthe control centre. Ultimately,control centre staff will be able toreconfigure existing circuitsequally quickly and so allow cus-tomes to change the use of thecircuits at will.

The first phase of the project -a £30 million contract coveringthe installation of more than60,000km of fibre - is nearingcompletion and work startsshortly on installing terminalequipment (customer servicemodules) on the premises of the100 customers who will be thefirst to use CFN. The fibrealready installed has been linkedto a System X service accessswitch (s.a.s.) at the mainBaynard House exchange in theCity of London to provide thecross -connections between

different fibre links needed to setup private circuits between cus-tomers.

Under Phase two, the CFNwillbe extended to more than 600sites and additional switchingcapacity will be installed. Thecontrol centre will give en-gineers access to BT's computer-ized fault-finding system for pri-vate circuits known as RATES(remote access and test equip-ment system) which allows cir-cuit faults to be pinpointedquickly.

Nearly 100 Dealerinterlinkcustomers will be the first to usethe fibre network. Dealerinter-link was launched in December1985 as part of BT's preparationfor the "Big Bang". Customersrent groups of analogue (non -fibre) circuits which are all con-nected to a central point. Thisenables a private circuit to be setup with any other Dealerinter-link user within 24 hours. Otherservices will then be progressive-ly provided via the fibre connec-tions including telex, packetswitched data, digital private cir-cuits and analogue links.

STC, as prime contractor. willbe supplying the complete sys-tem and management of the pro-ject, while Plessey will providevarying combinations of tele-phone and data links on demandas user needs change - and evenvaried at different times of day.Allocation, reconfiguration andaddition of new channels up tothe 34Mbit/s capacity of eachoptical -fibre link is done via itscontrol computer keyboard.

Battle for shareof the cells

With the backing of a £5 millionadvertising campaign, Cellnethas announced that, until Au-gust, subscribers joining the sys-tem will only have to pay half thenormal subscription for a sixmonths period. Vodafone re-sponded by authorizing its deal-ers to give a £75 discount -effectively the same as Cellnet -to those who join within thisperiod.

Cellnet launched its attackwith the objective of gaining thelion's share of the around100,000 new subscribers ex-pected in the forthcoming year.At present, Vodafone is believed

to have 80,000 of the 150,000users and has been sufferingfrom network congestion in theLondon area. It has only beenpartly successful in its attempt toget further channels allocated inthe London area on a temporarybasis until further channels be-come available next year.

First System Xfor Guernsey

TelecomsGuernsey Telecoms has com-pleted the first stage in its mod-ernization plan with the recentofficial opening of its first Sys-tem X exchange.

Plessey Major Systems Ltd isthe prime contractor for thisturnkey project for the island,whose economy depends to alarge extent upon off -shore fi-nance. Scheduled to be phased inover a 15 year period, it includesseven digital exchanges and anoptical -fibre network to linkthem. The fibre network hasbeen completed, together withthe Castel exchange. This is thefirst "split -mode" combinedtrunk and local System X ex-change to be brought into ser-vice. It carries 5,400 subscriberlines, plus all initial trunk andinternational traffic. Its initialcapacity of 480 Erlangs will re-duce to 360 when the secondtrunk/local exchange is cut overearly next year. This will belocated just behind GT's head-quarters in St Peter Port. (Erlangis a unit of traffic intensity whereone permanently engaged circuithas a traffic flow of 1 Erlang).

The remaining exchanges willbe progressively replaced by Sys-tem X and this programme willbe supported by the provision ofa new optical -fibre cable to theUK that will initially operate at140Mbit/s. This will be backed -up by a duplicate digital route viathe French mainland, employingdigital microwaye links toFrance, via Jersey, and thenlandlines across France.

At the present time, transmul-tiplexers have been installed toprovide an interface between theexisting analogue submarinecables to the UK and GT's grow-ing digital network.

Guernsey Telecoms serves allthe islands of the Guernsey Baili-wick: Guernsey, Alderney, Sark,


iiiittEconvis3cima.Herm and other, smaller, inha-bited islands. Currently there areover 26,000 exchange lines andnearly 50,000 telephones con-nected to the system. Other ser-vices provided by GT includepacket switching, wide -area pag-ing and, at the beginning of May,it linked into the Cellnet cellularradio network.

Mercury/Australia

switching dealA direct international publicswitched telephone service be-tween the UK and Australia hasbeen initiated by Mercury Com-munications Ltd and the Over-seas Telecommunications Com-mission (Australia).

Mercury already has publicswitched service agreementswith the USA, Canada, HongKong, Bermuda, Bahrain andBelize. Its customers making in-ternational calls to these destina-tions can obtain savings of up to17 per cent. Mercury currentlyoperates voice and digital privateleased lines between the twocountries and plans to expandthis service during 1987.

Fibre aroundthe world

Pacific Telecom Cable Inc. hasbeen granted a licence by theFederal Communications Com-mission (FCC) to land a new,high -capacity transpacificoptical -fibre cable in Seattle andAnchorage. PTC is 80 per centowned by Pacific Telecom withUK company Cable and Wirelessowning the balance.

The proposed cable will be thefirst direct fibre link between theUSA and Japan and has an esti-mated cost of $500 million. Theventure is fully supported by theUS as serving national securityinterests through providing di-verse routing for critical com-munications and as furtheringcompetition in internationaltelecommunications.

Mr Joe Crouch, C & W directorfor North America and a directorof PTC, pointed out that "InJapan, the Anglo-United States -Japanese consortium (Interna-tional Digital Communications)

is proposed for a second interna-tional telecommunications car-rier licence. All the partners inIDC see the Pacific cable as cen-tral to its business and lookforward to the Japanese govern-ment expediting its endorsementof the US Government's approvalof the new cable and believe thatthe speedy issue of a licence toIDC as an international telecom-munications operator and co-owner of the cable is essential toenable the project to be im-plemented within the time framestipulated."

Leased line toJapan

Mercury Communications, inconjunction with the Japaneseinternational telecommunica-tions company KDD, is expectedto launch a leased line service toJapan even though its parentcompany, Cable & Wireless, isstill embroiled in the fierce battlebetween rival consortia fightingfor the second Japanese interna-tional telecommunications li-cence.

KDD has recently applied tothe Japanese Ministry of Posts &Telecommunications for permis-sion to operate, in conjunctionwith Mercury, a service to theUK. This is seen as being animportant breakthrough forMercury as it could be the firststep in obtaining permission tooperate a switched service toJapan.

Pagingupheaval

There is growing competition inthe wide area paging market.Mercury Paging, a companyowned jointly by Mercury Com-munications (51 per cent) andMotorola, has entered the mar-ket where British Telecom has an80 per cent share. Motorola isproviding the pagers and settingup the v.h.f. transmitters, whichwill be connected via the Mer-cury backbone optical -fibre net-work, while Mercury will operatethe service.

Initially, the service will oper-ate in the Greater London area,but during the year its coveragewill be extended with the targetof nationwide coverage by 1989.

In a noticeable departure fromusual industry practice the com-pany is offering pagers on leasein addition to rental. Managingdirector, Bob Ure, says that thecompany aims to be broadlycompetitive with British Tele-com, but believes that he hasbetter designed coverage plans.He went on to say that the UKlags behind the USA and othercountries in the use of paging.

In addition, when questioned,Ure said that the software iscurrently being written that willallow Mercury pagers to beaccessed via the Mercury Linkelectronic mail service.

Even greater competition canbe expected before the end of theyear with at least one of the othertwo licencees for national pagingservices commencing opera-tions. One is Racal and the othera consortium of existing pagingoperators; Aircall, Digital MobileCommunications, InterCity andPage Boy.

Dialcomextends toFinland

The Finnish PTT has signed aletter of intent to provide theDialcom electronic mail servicewhich is marketed under thename Telecom Gold in this coun-try. Dialcom will now negotiate alicensee contract to enable thePTT to use the company's soft-ware and services.

Finnish PTT manager, LeenaSave commented that "X.400 isfast emerging as the internation-al standard in Europe and Di-alcom has the lead in X.400software in the public messagehandling domain. No progres-sive telecommunications com-pany can overlook that fact. Asecond important reason for ourchoosing Dialcom is because it isthe choice of many other PTTs. Itis serving the largest communityof international licensees withsoftware and techniques thathave proven themselves over theyears."

Dialcom has 16 other licen-sees around the world and claimsto be the leading internationalsupplier of value-added mail andother services with more than250,000 mailboxes worldwideincluding 76,000 in the UK.

Plessey andRacal link on

mobilesPlessey and Racal haveannounced the formation of a50/50 jointly -owned company toundertake the design, develop-ment, manufacture and market-ing of civil mobile communica-tions infrastructure and subscri-ber equipment. This followsclosely on the heels of the Euro-pean Quadripartite Agreementon digital cellular.

The main objectives of the newcompany, Orbitel Mobile Com-munications Ltd, is to developequipment for the pan-Europeandigital cellular system which isto be introduced throughoutEurope in the early 1990s. Whilethe new company will be at arm'slength from its parent com-panies, it will complete existingVodafone orders for base stationsand terminals. This will result ina 1987/88 turnover of some£25million. According to DavidDey, managing director of Ples-sey Telecommunications andOffice Systems Ltd and chairmanof Orbitel - this latter post willalternate annually between thetwo companies - the companycannot afford to sit on just oneproduct and service to enable itto meet a 1990/91 target of£100m. In addition to cellular, itwill be looking to other areassuch as private mobile radio.Consequently, he expects thetwo parent companies to investaround £30m over three years.

Orbitel intends to be a mem-ber of a major European cellularconsortium. It is engaged intalks across Europe and aims tojoin the group which offers thebest scope. In addition, it wouldbe interested in a USA involve-ment - especially if it providedaccess to the US market.

Dey stressed that customerdemand for cellular is proven;digital technology is available;and the pan-European politicalwill is present. The total equip-ment market, estimated to bebetween £500m and £600m forWestern Europe alone, is ex-pected to grow significantly asthe European standard isadopted in other parts of theworld.

By Adrian Morant


APPLICATIONS SUMMARY

:3 ,

Logic

operationunit

Maskregister

256b data

register

1

io ?Soselector

D,

Memory array

125E x 256)

Serialcontrol

Serial/0

By 1 mode110261(1)

Other Serial i/o (2-41high -impedance state

MULTI -PORT VIDEO RAMComprehensive details for display -systemdesigners are included in the Multi portvideo ram application design guide. The110 -page guide gives a general backgroundto video rams and details on Hitachi

Function Operation Result

0

1

2

3

4

5

6

7

0

AND1

AND2

X4 -> X1

AND3

THRU'

FOR

OR1

ZERO

D . M

15 . M

1024 x 1

D .i.11

D

DC)M

D + M

Function Operation Result

8

9

10

11

12

13

14

15

NOR

ENOR

INV1

OR2

INV2

0R3

NAND

1

D + M

DC)M

D

5 + m

M

D + ii

D . M

ONE

M = MEMORY CELL DATAEXTERNAL DATA IN

HM5346112 v -rams including address map-ping and timing information.

Most of the applications section describesa 512-hy-512 pel hit -mapped display systemfor interlacing to a 16hit microprocessor.Sixteen 256Khit v -rams arranged in fourhanks of four provide 16bit/pel and the pelrate is 25MHz. Smooth scrolling, clearingpels at high speed and multi -processor inter-facing are also discussed in the guide.

Video rams in the 53461/2 series consist ofa main 64K-hy-4hit dynamic memory and a256-by-4bit serial -access memory. The53462 shown here includes a Boolean logicunit. During a data -transfer read cycle.addressed data is transferred from the mainmemory into the data register ready forserial access. Access time of the serial mem-ory is 40-60ns whereas access time of the

main memory is 100-150ns.

LVDT WITH PHASE -SENSITIVE DETECTOREssentially, the 4260 synchronous detectoris a unity -gain amplifier that can he switchedbetween inverting and non -inverting modesusing a t.t.l. level signal. It also includes anuncommitted general-purpose op -amp simi-lar in specification to a 4558.

In RC.Vs preliminary data sheet for theRC4260. the linear variable differential

transformer with phase -sensitive detectorshown here is accompanied by circuits forsuppressed -carrier modulation and preci-sion rectification. Output polarity of therectifier circuit is logic -selectable.

Distortion of the detector is typically0.01%. it switches in IONS and its gaindifferential is at most 1%.

R f

---lP-P--'\AAP-14°-1'

2

11111111-'4 2 6C

5nd

of

8

-15V

10

100n 100n

.15V

\ lout0

Centrezero

ELECTRONICS K WIRELESS WORLDs:29

APPLICATIONS SUMMARYOPTICAL FIBRE TESTINGWith pulsed diode lasers and fast detectors itis possible to measure optical fibre band-widths up to 10GHz and pulse dispersionsdown to 20ps at wavelengths from 800 to1550nm. An application note for the FPSIOFourier processing system outlines theequipment and test set-ups needed to makethese measurements.

The note, from Opto Electronics, de-scribes how measuring bandwidth and dis-persion involves comparing results from twodifferent lengths of the same fibre, onetypically greater than I km long and theother a few metres long. Pulse -responsemeasurements of long and short lengths offibre are made and Fourier transforms ofeach measurement are carried out.

Frequency response of the long fibre iscalculated by taking the ratio of long -to -short length Fourier transforms. Pulse re-sponse of the fibre is obtained by performingan inverse Fourier transform on thefrequency -response curve.

USING FFT

1300nmpulsed -diode

laser

Light out

a Lens

XYZ positioner

Mode mixer

Mandrel wrap

Mode stripper

Fibre under test (>1km)

irr)Long fibre ( a)

(a) or (3)

Short fibre (b)

Phc-

r Lens

XYZ positioner

FIR FILTERThree identical arithmetic cells each con-taining a 10 -by -10 bit two's complementmultiplier and 23 -bit adder form the

TMC2243 finite impulse response filter. Pre-liminary data on the TRW device includesspecifications, general information and briefdetails on how to configure the deviceregisters for adaptive filtering.

Applications of the 20MHz device includevideo and radar signal processing, one andtwo-dimensional convolution or video filter-ing and arithmetic elements in systolic arrayprocessors.

Latch

CK

Theft e

Latch

CK

Coefficie t

write enable

Data

10

CK

Sum in

Latch

CK

To allregisters

6

No of linesper bus

CK

CO

20

CK

-CK CK

Ip

Feed -through contrAmMIL--/

CK

23

Constant registers x3/

Secondary constant2° / registers x3

Overflowflag

OE

Sumout

161"


APPLICATIONS SUMMARYClock 'A 11.

Input Input -:

---.---*re-proc '.

Main bus

Sampler Eau

Micro I Microprogram program

sequencer I memory.,--.I

Dmocontroller

C lock 2 II

14 a

Memoryd- to -a cony

$

Address

sequencer

I/Onterface

4

Output

Complex

data file

Ifarithmetic

Complexa

unit

I

SIGNAL -PROCESSING BUILDING BLOCKS PROVIDE FASTER FFT

An article in Logic Devices' product cata-logue describes how high-speed c-mosbuilding blocks form an FFT system thatworks up to 100 times faster than somesingle -chip solutions. The article discussesFFT in general, the two main types ofsingle -chip digital signal processors, and abuilding-block approach that performs a1024 -point complex FFT in 500ns.

One group of single -chip digital signalprocessors, which includes the TMS32010/20 and I.LPD77230, has on -chip instructionand data memory. These processors workmost efficiently when the program and dataare in the chip's memory throughout thewhole process. But computational through-put is still slow since only two data operandscan be processed in any given cycle. If thenumber of sampled data points to be proces-sed exceeds the on -chip memory capacitythen data has to he stored in externalmemory which slows down processing andincreases software overhead.

The second group of processors, whichincludes the LM32900 and ADSP2100, relieson external program/data memory. Theyhave two independent buses to make acces-sing of external memory efficient but theycan still only process one set of operands inany given cycle. Running two devices inparallel increases performance but controland synchronization circuits are needed.

Building blocks can minimize the number

of machine cycles needed for each butterflycomputation (flow of each basic FFT cellforms a 'butterfly' pattern). With thearchitecture shown each butterfly takes twomachine cycles: pipelining is for 32hitcomplex -data ITT

The article. entitled 'New high-speed c-mos building blocks provide a clean imple-mentation of the FFT for applications wheresingle -chip DSP microprocessors cannotprovide the necessary throughput' is in theproduct catalogue supplement.

Performance figures for 1024 -point complex FFT

Signalprocessor

Memot y Instructioncycle time

1024 -pointcomplex FFT

Samplingrate

TMS32010 144, 16 data :am

pPD7723015361 16 ram1K 32 ram

75.9ms 13.1kHz

1K x 32 data rom2K 32 torn 150re 10 75ms 100kHz

LM32900 External lOOns 13.42ms 78kHz

ADSP2100 Exter nil 125ns 7 2ms 142kHz

Budding blocks External 5Orts 0.5ms 2MHz

ADDRESSESHitachi Electronic Components21 Upton RoadWatfordHertfordshire WD1 7TP0923 46488Telecom Gold 76:HECO11

RCASeminconductor Specialists159 High StreetYiewsleyWest DraytonMiddlesex UB7 7XB0895 445522

Opto ElectronicsOriel ScientificPO Box 311 Mole Business ParkLeatherheacSurrey KT22 7AU0372 378822

TRWAccent Electronic ComponentsJubilee HouseLetchworthHertfordshire SG6 1710462 686666

Logic DevicesAbacus ElectronicsAbacus HouseBone LaneNewburyBerkshire RG14 5SF0635 38670

DUAL -PORT STATIC RAMSThe telephone number with lastmonth's description of a dual -port ram application was forVLSI Technology's facsimilemachine; it should have read0908 667595.

ELECTRONICS & WIRELESS WORLD s:I

[SATELLITE SYSTEMSElectronic

tracking forinter -orbit links

Fast acquisition, simple micro-wave components and no needfor a separate tracking receiverare features of a British electro-nic beam -pointing system to beused for inter -orbit communica-tion. The idea is to keep theantenna beam of one satelliteaccurately pointed at the anten-na of another satellite, in a differ-ent orbit, to ensure goodradiocommunication betweenthe two spacecraft. This electro-nic tracking system, developedby ERA Technology Ltd, will beused by ESA for an inter -orbit,two-way, communication ex-periment at 20/30 GHz betweenthe low -orbiting European Re-coverable Carrier (Eureca) andthe geostationary Olympus satel-lite (see item in July issue).

Direct radio links betweensatellites - without passingthrough ground stations - arenot yet an established part of thetechnology, though the principleis quite old. They were used, forexample, in 1969 betweenspacecraft in the Apollo (moonlanding) project, several timesduring the 1970s, and in 1983between a data relay satellite(TDRSS-1) and the Spacelab andLandsat spacecraft. For satcomsystems, the possible advantageslie in the avoidance of doublehops (requiring earth stations). a

Antenna feed system forelectronic -scan tracking de-veloped by ERA Technology forthe European Space Agency's20/30-GHz inter -satellite com-unications experiment betweenthe Eureca and Olympusspacecraft.

832

better utilization of the availablefrequency spectrum and a lesscomplicated earth segment. Any-way, this is what ESA plans toinvestigate with the Eureca-Olympus experiment.

The WARC 79 inter -satellitefrequency allocations to be usedfor this particular experimentare 22.5 - 23.55 GHz and 32-33GHz, but there is another allo-cated pair of bands at 54.25-58.2GHz and 59-64 GHz - well intothe millimetre waves. In addi-tion, both ESA and INTELSATare considering the possible useof optical frequencies generatedby diode lasers with wavelengthsof 0.8 to 1.5 gm to achieve verynarrow beams over long dis-tances.

ERA's electronic tracking sys-tem is centred on a microwaveantenna feed device, as shown inthe photograph, to be installed inEureca. The feed horn, for bothtransmit and receive, illumin-ates a shaped, dual -reflectorantenna. In response to a beaconsignal from Olympus the anten-na beam is electronicallysquinted in different directionsand the resulting different re-ceived signal strengths provideinformation for antennasteering.

Beam squinting is achieved bygenerating proportions of higherorder waveguide modes withinthe feed horn to modify theaperture illumination. A higherorder mode of suitable amplitudeand phase, in combination withthe fundamental mode. producesan overall phase tilt. This feedphase tilt results in a displace-ment in the antenna's radiationpattern - hence a beam squint.

Referring to the photograph,the mode generator section com-prises a central circular wave -guide with four short-circuitedlengths of rectangular wave -guide coupled to its periphery in

selected positions. Each auxili-ary waveguide connected to thecircular guide is terminated by ap-i-n diode. The mode genera-tion is activated by reversing thebias of this p-i-n diode.

To locate the Olympus beacontransmitter the four higherorder mode generators gothrough a search pattern underthe control of the p-i-n diodes. Inthis the beam is sequentiallyswitched from true boresight toeach of four positions: elevationup and down; azimuth right andleft. The strength of the beaconsignal at each beam position isdetected by the normal com-munication receiver in Eureca,converted into digital data by ana -d converter, and eachmeasurement is passed to a mic-roprocessor. where it is stored inconjunction with its co-ordinatedirections.

Computation in the micro-processor then provide an errorsignal for a closed -loop controlsystem operating the antennasteering. Thus the Eureca anten-na is steered until its beam isaccurately pointing at the Olym-pus antenna.

In the electronic scanning sys-tem the rapid switch -and -measure sequence allows thewhole search pattern to be com-pleted in a fraction of a second.Although the two spacecraft arealways moving relative to eachother, no substantial change inposition occurs during this timeframe. So the four measure-ments of the search pattern canbe regarded as simultaneous.The ERA electronic tracking sys-tem has already been tested onearth stations working to satel-lites.

Third Europeancomsat

A third European communi-cations satellite for telecom-munications and tv signal dis-tribution is expected to be laun-ched by Arianespace in Augustthis year from Kourou. FrenchGuiana (Africa). This is ECS-4. socalled because it is part of theEuropean CommunicationsSatellite system. In the ECS sys-tem the comsats are specified,purchased and launched by theEuropean Space Agency (ESA).They are then operated and man-aged by Eutelsat, another inter-

national organization but set upby the various European tele-comms administrations (see De-cember 1978 issue, p.63).

Under its Eutelsat coloursECS-4 will be known as EutelsatI, Flight 4 (or I -F4). Here theRoman number indicates thatthe spacecraft belongs to a firstgeneration of comsats, which sofar amounts to two satellites inregular operation, Eutelsat 1-F1and Eutelsat 1-F2 (or ECS-1) andECS-2). Altogether, five comsatshave been planned for Eutelsat I.They are manufactured by aEuropean consortium led byBritish Aerospace as prime con-tractor. Later on a second gen-eration of comsats. Eutelsat II,will be launched. Here themanufacturing consortium is ledby Aerospatiale of France. ESAhas so far ordered four spacecraftfor Eutelsat II and there is anoption for a further four.

Readers may well ask why athird European comsat is calledECS-4. Those with long memor-ies will recall that an ECS-3 (tobecome Eutelsat 1-F3) wasactually built but was lost in alauncher failure in 1985. So thenew ECS-4 really replaces ECS-3. Assuming this year's launch issuccessful and the new comsatgoes into regular operation, ESAwill then continue with the plan-ned ECS-5. There is also a possi-bility of adding a sixth spacecraftto make up the total com-munications capacity originallyintended for Eutelsat I.

Like ECS-1 and -2, the latestgeostationary comsat is designedfor relaying European publictelecommunications traffic(telephony, data, facsimile etc.)and for distributing televisionsignals, both for EBU broadcast-ers and for commercial satellitetv operators. It will be stationedat 10°E. Like ECS-2, it also car-ries an additional payload forprivate business data com-munications, called the SpecialMultiservice System (SMS). Thelast-mentioned provides digitalchannels for data rates of 64kbit/s, intended for various kindsof information - speech, data.text, graphics, facsimile, slow -scan tv and video-conferencing.

The communications capacityfor all these services is providedby twelve 72 -MHz transponderchannels for the public telecom-munications and tv signal dis-tribution traffic, plus two 72 -MHz channels in a separate fre-


[SATELLITE SYSTEMSquency band for the private SMStraffic. Uplinks are at 14 GHz anddownlinks at 11 GHz, as is stan-dard in the Ku -band Fixed Satel-lite Service. The 12 main chan-nels are actually provided by sixchannel frequencies, which aredoubled -up by the use of hori-zontal and vertical polarizations.Thus in the 11-GHz downlinks(within 10.95-11.7 GHz). chan-nel 1 and channel 7 share thesame channel frequency, butwith different orthogonal polar-izations, and so on. up to chan-nels 6 and 12, which use a com-mon channel frequency on thesame principle. The SMS repea-ter has its separate downlinksub -band at 12.5 to 12.58 GHz.

Coverage of the European landmass is achieved by two broadantenna beams, calledEurobeams, and three spotbeams named respectively Atlan-tic, West and East. As can be seenfrom Fig. 1, these are producedby six antennas on thespacecraft. All the primary tele-comms and tv distribution up-link signals are received by aEurobeam receiving antenna.The public telecomms downlinksignals are sent back to earth bythree spot -beam transmittingantennas, depending on destina-tion, while the tv distributiondownlink signals are radiated bya Eurobeam antenna. All thespot -beam transmitters give ane.i.r.p. of 40 dBW, while theEurobeam transmitters provide34 dBW. As can be seen from thediagram, the sixth antenna isused for both reception andtransmission of the SMS payloaddata signals.

Uplink signals are all pre -

amplified in four wideband re-ceive chains, arranged to give aredundant configuration. Theythen undergo double frequencyconversion, with common down -conversion to the first i.f. The72 -MHz bandwidth is establishedin the i.f. section. In the primary12 transmitting chains there is aswitching system to the outputduplexers which allows the com-sat some flexibility to meetvarying traffic requirements.

The digital modulation systemadopted by Eutelsat for sendingthe p.c.m. public telecommstraffic through their comsats isfour -phase p.s.k., operating at120 Mbit/s. In this the transmit-ted carrier assumes any of fourphases according to the instan-taneous value of the modulating

EurobeamRx

146Hz Fourwidebancreceivechains

Twelve,ansmit

mains

Eurobeam Is

Spot east To

Spot west T

Spot Atlantic Tx

11GHz

- Primary payload -.-.-.-.-...-- $MS payload

1GH:Two

transmitchains

12 5 GHz

1.6Hz

Fig. 1. Simplified block diagramof transponders in the ECS-4comsat, showing the maintelecomms/television payload atthe top and the special businesscommunications payload at thebottom.

signal. Each phase state conveystwo bits of information. The tele-vision signals are, as usual, in theform of analogue f.m., requiringtransponder bandwidths of 27 or36 MHz. Audio is carried bysound-in-syncs for broadcast sig-nal distribution and by subcar-riers (typically 6.65 MHz) for thecommercial satellite tv program-mes. The separate SMS payloaduses p.s.k. for its digital trans-mission, in a single -channel -per -carrier (s.c.p.c.) frequency -division system. It will handle,for example, BTI's SatstreamEurope business service (and seephoto of German earth terminalin June issue, p.603).

Perhaps the most characteris-tic feature of the ECS system isthat from the very beginning itwas designed to provide multipleaccess to transponders fromdifferent earth stations on atime -division, rather than fre-quency division, principle. ECS-4, therefore, uses time -divisionmultiple access (t.d.m.a.) for itsmain telecomms traffic and doesso in conjunction with digitalspeech interpolation (d.s.i.).This combination is more effi-cient than earlier frequency -division multiple access systems,allowing a greater number oftelecomms circuits to be carriedby a single comsat.

In t.d.m.a. a relatively low -speed continuous bit stream is

Speciai servicesTx/Rx

processed in the earth terminalto be transmitted through thesatellite in short bursts at a

much higher speed. These burstsare interleaved with bursts fromthe other earth stations workingthrough the same satellite ttans-ponder on the same frequency. Acomplete set of bursts from allthe different earth stations usingthe satellite is called a frame.Because the comsat amplifierhandles only one burst at a timebefore re -transmission it can beoperated at high power withoutunacceptable distortion -hence the greater efficiency.

Digital speech interpolationexploits the fact that, on average,each participant in a telephoneconversation is silent for about60% of the time. A caller is onlyassigned a satellite channel whenhe/she is actually emittingsounds and so generating digits.In the intervening times thechannel is available to carry thedigital speech burst from otherconversations.

Addendum onAstra

Since the item on Luxembourg'scommercial television satellitewas written for the June issue(p.603) more details have be-come available on this system.

The 16 channels at frequen-cies from 11214.25 MHz to11435.5 MHz are, of course. slot-ted into what was once a 250 -MHz gap in the Fixed SatelliteService (FSS) Ku band alloca-tion. from 11.2 to 11.45 GHz.

This spectrum space was made

available for FSS by an allocationdecision at WARC 1979. Astra'stransponder bandwidth is26MHz. This may not seem to fitwith our statement that the 16channel frequencies are spacedat intervals of 14.75 MHz, but isexplained by the fact that fre-quency re -use made possible bydifferent polarizations allowschannels to overlap. Thus thevertically polarized channel 2partly overlaps both channel 1and channel 3. which are bothhorizontally polarized.

Astra's European coverage,determined by beam shaping inthe RCA 4000 satellite antennasystem, will be a footprint some-thing like a rectangle withrounded corners, wider least -to -west) than it is high. An innerfield -strength contour corres-ponding to a nominal transmit-ter e.i.r.p. of 50 dBW will extendabout 22° in longitude, encom-passing Dublin to Berlin, andabout 13° in latitude, encompas-sing Edinburgh to Marseilles.Outside of this, a 46-dBW con-tour bulges at the south-westcorner of the `rectangle' to coveralmost the whole of Spain. Thisouter contour will also enclosethe northern half of Italy, south-ern Norway and Sweden. and thewhole of Ireland and Scotland.

Societe Europeenne des Satel-lites is a company which ownsand operates the Astra system. Itdoes not produce televisionprogrammes. The commercialprogramme companies using thesystem will in general transmittheir programe feeds directly tothe satellite (in the 14.25-14.50GHz uplink band) from fixed ortransportable earth stations intheir own countries or fromwherever convenient. In the UK,for example, uplinks will operatethrough BT's London Teleport.

However, SES does provide anuplink earth terminal at its Betz-dorf control station in Luxem-bourg for any programme com-panies who may wish to use it.Also at Betzdorf is an 11 -metredish antenna for satellite track-ing, telemetry and control pur-poses. A transportable earth ter-minal is available there for back-up telemetry and command andfor occasional uplinking require-ments.

Satellite Systems vas written byTom Ivan.


A new look at gain-

bandwidth productDoes the bandwidth of a feedback amplifier necessarily

have to shrink as the gain is increased? Is it a fundamentallimitation or is it simply a result of the way we design

feedback amplifiers?

IIt has become generally accepted overthe years by most designers of electroniccircuits containing feedback voltage

amplifiers that the gain -bandwidth productavailable is constant. Observations of drama-tically reduced bandwidths at higher gainshave become so commonplace that it hasbeen elevated subconsciously to the status ofa physical law. But is this necessarily cor-rect? Is it in fact an expression of a fun-damental concept: or is it simply somethingthat occurs solely as a result of the way inwhich we design feedback voltage ampli-fiers?

THEORETICAL

Figure 1 shows the traditional form of anon -inverting feedback voltage amplifierconfigured from a voltage operational ampli-fier (v.o.a). The aim is to produce a stablevoltage gain by employing feedback around av.o.a. which displays a very high gain,perhaps in excess of 100,000 or 100 dB, thatis only imprecisely defined at the manufac-turing stage. After the application of feed-back the overall voltage gain is lower, butstabilized against changes in the v.o.a. gain.

If we restrict our attention to a v.o.a.exhibiting a single -pole response, the trade-offs can be seen graphically in Fig.2. It is selfevident that, because of the falling v.o.a.gain above some low -frequency breakpoint,the bandwidth available for low gains isgreater than that available for high gains.For the example illustrated, the gain -bandwidth product is always constant at1MHz. In other words, this particular ampli-fier would be advertised as having a unity -gain bandwidth of 1 MHz.

Analytically, one may represent the over-all voltage gain C after feedback by the usualexpression

G--

B. WILSON

Fig.1. A feedback voltage amplifier isusually configured from a voltage oper-ational amplifier (v.o.a).

120

100

-a3 80.33

60-

7-1 40-

20-

0

Af .10Hz

Aloop r A yo 1 Ay . Ao I f/fL 1

6v

10 100 1k 10k 100k 1M 10M

FREQUENCY (Hz)

Fig.2. A v.o.a. using traditional feedbackproduced lower bandwidth at higher gains.

120

100

33 80->

'D 60-zQ 40-

20-

Gv=601:13, f _3der 1 k Hz

Gain -bandwidth produc tr1MHz

Gv=20dB,f-3da =100kHzGain -bandwidth product

.1MHz

$ +

0 10 loo 1k 10k 100k 1M 10M

FREQUENCY (Hz)

Fig.3. Reduction of loop gain at higherfrequencies results in a constant gain -bandwidth product

where A, is the v.o.a. forward voltage gain asa function of frequency and 13 is the feedbackvoltage fraction. For conditions under whichA>> 1 this is usually simplified to thefamiliar result

Bandwidth restrictions come about be-cause is a falling function of frequency.intersecting a line representing C and den-ying the system the loop gain that it needs inorder to function correctly as a feedbacksystem, as shown in Fig.3. The v.o.a. gain Amay be related to its very low frequency gainA through the relationship

11 + j(f/fL)I

where k is the low -frequency breakpoint,taken here to be a typical value of 10 Hz. Atfrequencies well removed from f1. the ex-pression for the magnitude of A may besimplified to

To a high degree of precision it can be saidthat the system bandwidth is reached at thefrequency where the loop gain Al,, hasfallen to unity (or 0 dB). Figure 3 shows thatthis happens when

A(3= 1.

Substituting for A, then gives

A013= WO -

Changing the general symbol fat the band-width limit to BW, and recalling that Gy=(3 I, then yields

Re -arranging.

G BW.T.=

Gy.BW=A,,.11


In other words, the gain -bandwidth productmost certainly is a constant and is equal tothe product of the low -frequency gain Amultiplied by the low -frequency breakpoint

For amplifiers with a two -pole response,the bandwidth restriction effects are evenworse, of course. After the second pole hasbeen passed, A,. falls even faster by anadditional 20 dB per decade, reducing thebandwidth extension below that availiablefrom a single -pole response. This effect isnot always immediately apparent, however,since the majority of v.o.as. are used with adominant single response to avoid potentialinstability.

Can anything be done to improve thesituation, especially since both the theo-retical predictions and experimental resultsagree so closely? Is it indeed possible todesign a feedback voltage amplifier wherethe bandwidth is independent of gain? Theanswer, as usual, depends upon describingthe original problem in a different, andhopefully more illuminating. manner.

Looking again at Fig. 3 shows that theloop gain A,43 may be re written as A,JC,,open -loop voltage gain divided by closed -loop voltage gain. For a given v.o.a., there-fore, the loop gain available at any frequencyis determined solely by the desired closed -loop voltage gain of the system . It is this factwhich is responsible for producing a con-stant gain -bandwidth product. This restric-tion could be avoided quite simply by mak-ing the loop gain of a topology independentof its closed -loop gain, in which case thebandwidth would then be the same for allvalues of closed -loop gain C.

This may be arranged by designing avoltage amplifier using feedback around acurrent amplifier, as in Fig.4. In an analo-gous fashion to standard v.o.a. feedbackanalysis, one may assume that the voltagedifference between the input terminals ofthe current operational amplifier (c.o.a.) iszero, equivalent to zero input resistance.Simple algebraic manipulation then pro-duces an expression for the output voltageunder no-load conditions as

R RF A= ( --V 2). (1+A)

RI ,

where A, is the open -loop current gain of thec.o.a., assumed to display a similar frequen-cy trend to the open -loop voltage gain A,. fora v.o.a., Restricting the analysis to thesingle -ended case with II equal to zero forease of comparison gives

V RF A

V2 R2 1+A,

Breaking the feedback loop conceptually at Ldemonstrates that the loop gain is given byA, alone, with no contribution whatsoeverfrom the closed -loop voltage gain. Hence the-3 dB bandwidth will now be governed by thefrequency behaviour of A. Consequently,the bandwidth will be the same for all valuesof closed -loop voltage gain. set by RF and R1.

A parallel analysis may be developed whena v.o.a. is used to produce a defined closed -loop current gain, resulting also in a band-width that is independent of gain. It is

therefore apparent that the restrictive caseof constant gain -bandwidth product is adirect result of the feedback topology em-ployed in previous traditional designs.

For a current amplifier that displays ahigh value of gain A,. the closed -loop voltagegain of Fig. 4 will remain independent of A,until the magnitude of A, approaches unity.The frequency at which the magnitude of A,becomes unity then corresponds to the -3 dBfrequency of the closed -loop voltage gain.However, it is not necessary for the lowfrequency value of A, to be very high, or evengreater than unity, for the gain independ-ence of the bandwidth to remain valid. Aslong as A, remains well defined and stable theonly consequence of a reduction in A, is toreduce the factor by which RF/R2 is multi-plied. The two situations of most practicalinterest will be to employ a very high.perhaps imprecisely defined, value of cur-rent gain such that the multiplication factoris unity; or to use a precise current gain of 1,in which case the factor will be exactly 0.5.

A class II current conveyor is an idealcandidate for this task. especially since it hasa precisely defined unity current gain and ahigh -impedance input terminal in additionto the current input. Earlier articles inWireless World have previously looked atmirrors, amplifiers and conveyors. but it willbe useful to consider certain aspects of themhere. Figure 5 shows the symbol for acurrent conveyor, where a voltage applied atthe high impedance terminal Y is transferreddirectly to the low impedance terminal X. Acurrent may be injected directly at X, or mayoccur as a consequence of an impendancebeing connected there. A direct copy of thiscurrent is then provided at the isolatedcurrent output, terminal Z. A current con-veyor may he thought simply as a hybrid ofv.o.a./c.o.a.

The most versatile design of current con-veyor produced so far utilizes a v.o.a. andtwo current mirrors, connected to producean output current by sensing the v.o.a.supply current, as in Fig. 6. This arrange-ment is necessary since manufacturers donot provide direct access to the collectors ofa v.o.a's output transistors. To complete thepicture, Fig. 7 illustrates how the currentmirror may be constructed from four iden-tical transistors to produce current replica-tion with an accurately defined gain of unity.It is unusual to employ only transistorsformed on the same substrate and packagedtogether to obtain the best matching andthermal tracking.

How may the current conveyors be used inpractice to achieve voltage gain? Figures 8and 9 illustrate the feedback connectionsrequired to produce both inverting andnon -inverting voltage gains respectively.The topologies are extremely similar to thev.o.a. case, but it must be remembered thatthe inverting terminal is now a very lowimpedance current input. A little algebraicmanipulation will show that the closed -loopvoltage gain in each case is

G,.=- RLI inverting),2R1

Cy= 1 +-(non -inverting)2R1

Fig.4 Voltage ampli iers may also beform.dated by using feedback around acurrent operational amplifier (c.o.a).

lin

0Vin0

CCII

tor 110

Fig.5 A class II current conveyor has avoltage input terminal (Y). a current inputterminal (X) and a current output (Z).

Fig.6. A current conveyor mayplemented from a v.o.a. and twomirrors connected to give a

output.

A

be im-currentcurrent

Fig.7. Stable and accurate current mirrorscan be designed from four transistors.

Fig.8. An inverting voltage amplifier with acon >tant bandwidth can be designed usinga current conveyor.

Fig.g. A non -inverting version is also easilyava,lable.


EXPERIMENTAL

The performance of voltage amplifiers con-structed from current conveyors is extreme-ly impressive. An inverting voltage amplifiermay be designed by using CA3096AE arraytransistors for the mirrors and a 'bifet'LF351 as the v.o.a., formulated as in Fig. 10.Setting Ro at 3.3 kit and varying R1 producesgains ranging from unity to more than 100,with perfectly behaved square waves of up togreater than 10 V pk-pk output. In confirma-tion of the new analysis the output voltagerisetime was found to be constant, in thiscase at about 100ns with a small overshoot.irrespective of the value of gain. Sinewavetests indicate a constant -3 dB bandwidththroughout the gain range of 3.5 MHz,exactly in accordance with the risetimemeasurements. Figure 11 presents the ex-perimental gain -frequency plot, illustratingdramatically the freedom from gain limitedbandwidth available from the new approach.The LF351 is advertised as having a unity-

gain bandwidth of only 10 MHz: with thenew topology it displays a behaviour equiva-lent to a gain -bandwidth of up to 350 MHz at10Vpk-pk!

A constant bandwidth throughout thegain range is not the only advantage to begained from this topology. Output voltageslew rate is also remarkably improved, sincethe v.o.a. output is now connected as avirtual earth and so does not move throughany significant voltage whatsoever, leavingthat task to the current mirrors. A risetimeof 100ns for a 10Vpk-pk square wave whenusing an LF351 represents a slew rate of100V/µs,whereas the data sheet indicates ithas a maximum of 20V/p.s. Substitution ofother v.o.as, LM301A and 741 for example.with a much reduced stewing ability of lessthan 0.5V/µs. still results in almost identicalresults of nearly 100V/µs; a slew rate magni-fication of almost 200 in many cases.

In the non -inverting configuration thebehaviour is extremely similar. Again settingR0 at 3.3 kit, very sharp triangular andsquare waves can be reproduced with rise -times as low as 100ns. Figure 12 shows aphotograph of 400 kHz output waveforms.both taken at a voltage gain of 15. However,in the non -inverting configuration, thev.o.a. output terminal is no longer a virtual -earth connection and slew rate effects beginto be apparent again at lOy pk-pk. especiallyas the gain is reduced. The effect is mostnoticeable at unity gain, where the v.o.a.experiences the full voltage excursion.

DISCUSSION

The new amplifier formulation uses currentfeedback to achieve bandwidth independ-ence and cannot therefore supply a substan-tial output current without degrading theoverall gain accuracy. A high -impedancevoltage follower or high-speed, unity -gaincurrent buffer should be added at the outputfor circuits where a significant output cur-rent is required. Alternatively the conveyoritself may use the current mirrors in afeedback connection to release a low -impedance output terminal. It is also possi-

Fig.10. Actual circuit implementation ofthe new constant bandwidth amplifier withinverting gain.

40

z - 307.1

c3 - 1

0

f -3dB 3.5MHz

10 100 1k 10k 100k 1M 10M

FREQUENCY (Hz)

Fig.11. The frequency response from thecircuit of Fig.10, showing a constant band-width for all gains.

Fig.12. The non -inverting version also pro-duces good fast triangular and squarewaves with risetimes as short as 100ns.(Scales: 1V/cm and 500ns/cm).

ble, if required, for gain to be introducedinto the conveyor circuit by including acurrent attenuator in the conveyor localfeedback loop'.

The effect of capacitance at the output ofthe current mirrors combines with resist-ance between output and ground to intro-duce a high -frequency pole into the circuit.For no-load conditions this resistance isdetermined primarily by R2 alone. In mostcases it is this pole which limits the high -frequency performance of the system, re-sulting in a risetime proportional to thevalue of R2. Below a certain value of risetime,however, the current mirrors also becomebandwidth limited, producing a characteris-tic double -pole overshoot in the square waveresponse below 100ns. For the best possibleresults, therefore. R2 should be chosen toachieve the fastest risetime before the finitebandwidth of the mirrors becomes a limitingfactor. R, may then be chosen to give thedesired voltage gain.

To summarize, high gain and wide band-width can be obtained simultaneously if thetraditional topology of a feedback voltageamplifier is modified to use current con-veyors or current operational amplifiers. Asan additional benefit, the restricted slewingability of the v.o.a. can be overcome. even fora 741, opening up the area of high perform-ance at relatively modest cost and com-plexity.

References1. Wilson. B.. Using current conveyors, Electro-nics and Wireless World. April 1986. Vol.62.pp.28-32.2. Lidgey. F. J. and Toumazou. C.. 'Accuratecurrent follower' Electronics and Wireless World.April 1985. vol. 61. pp 17-19.

Brett Wilson lectures at UMIST in Manches-ter, where his two main interests are fibreoptic communications and high-speed ana-logue electronics. He and his research stu-dents are currently investigating new de-signs of optical -fibre receivers and the trans-mission of high -quality multiplexed videosignals by fibres.

BOOKSProficient C by Augie Hansen. MicrosoftPress. Penguin Books, 494 pages, papercovers, £19.95. Extensive handbook forthe intermediate or advanced program-mer working in a DOS environment.Sections deal with DOS and programdevelopment, using standard librariesand interfaces, file utilities, the ANSIdevice driver and screenj routines.Attractively presented. The many exam-ple programs are also available on twodiscs, as source code or as executableprograms.

Private Mobile Radio -a system plannerby John Davies; 48 A4 pages ring -bound.soft covers. Available direct from theauthor, price £8.50 inclusive (Europe£9.50, overseas £11), at 26 Walnut Close,Penwortham, Preston, Lancashire. Do-it-yourself kit for prospective users, to beused in conjunction with the author'sPrivate Mobiel Radio: a practical guide(Heinemann). Contents include a ques-tionnaire designed to establish basic sys-tem requirements, a preliminary speci-fication for discussions with equipmentsuppliers, draft form of tender, guidancenotes and so on.

Programmable Logic Handbook by GeoffBostock. Collins Professional Books, 243pages, hard covers. £25. Authoritativesurvey of this fast-growing device family.The final part of the book deals at lengthwith a wide diversity of logic elementsshowing how they may be fitted intoprogrammable logic devices. Appendiceslist p.l.d. manufacturers and companiesoffering c.a.e. support.

836 ELECTRONICS & WIRE LESS WORLD

MARCONI TF2371 200MHzSPECTRUM ANALYSER

RALFE - ELECTRONICS10 CHAPEL STREET, LONDON. NW1

TEL: 01-723 8753/724 8074

[6000

Parc. am

[2500

TF2006 AM FM Solna, generate' 10 1000MH, [950TF2011 FM Sera, generator 130 113044H, [32TF2015A 1 AM FM Sognal generator I 0.520MH, 600TF2016A VHF AM FM *nal geneator watt 2173 lock.boo I850TF99545 AM FM signal generator 1 5-220MH, [1952017 AM FM sxral generator 0 01 1024MH7 SuPerb spec £6.000TF2170 digital synchronizer Om 2002 series/ [250TF2300S modulabon mere, 35100081H: AMFM 495TF2303 Modulation meter 2 5.520MH,AM FM £325TF1313 1 4, RCL Unnersal Component Sedge 1250TF1245A 1 0 -Meter 1246 8 1247 [550 0250 0150T144520 Standard etcluctor set 1250TF893A Audso power meter I ntW-10W IsO C75

TF1020A 100W 2504414, RF power meter [75TF115241 75W 250MHz RF power meter [75TF2604 ElectronK multi -/unction meter C150

TF2430 80MH, F,equency counter [95TF2807A PCM mumptes tester [4502829 PCM digital analyser [7502833 PCM chgaal Imo rnoriltor [2752828,4 Dognal senutator [750TF2356 20M14, level osc,tlator (650TF2167 OF amnia., 0 05.80MHz 4706 gam [450TF1263S 1GH, agenuator 14200 1db steps t250TF2092C nose mover [350TF2910.4 non-linear dostortion analyse, t750TF2162 MI Attenuate/ 0 I 1 lab in 0 tab steps (100

e now w/amele lot all items or 1.4.,,.1` 1000 0' MU,.

Parts and labour warranty kr a /A AI Id WI' 1111

osest competitor. Certa/cate,star/Pau:Is trace ib/Iity are ayallabe upon u5lueS1 to Conant ItemsPease let us quote you ()tre,ent rates .v,aildtee on cheaper eporprneor

MARCONI TF29058 Sme squared pulse and tw, gene,WOr C450MAR(lINI TF2908 Olanstng and sync mixer [250MARCONI TF2502 PF Wattmeter lOW 1st] t GH, C250MARCONI 2435 2GH7 8 dosyt frequency counter £750GREEN 2601 Wattmeter 0 3.300W to 500MHz C250TEL ETYPE Tetnena model 43 RS232 , lace C450DATALAB 01905 Transient recorder C275TELEOUIPMENT C-71 tranststor curve tracer C800TEKTRONIX 575 Trsnststor curve Pacer C250TELEOUIPMENT 0113 OstIttoStopes SOMH, 0350TEL EQUIPMENT D4464 Storage scopes horn 0250OSCILLOSCOPE .toct, Phone tor best quoteBRVANS 22020 %VT Platter A3 C175

BRUEL 8 KJAER 1432 Random rase ger, £250DVMAR 785 modu tat., meters 30-480MHz AM FM £125[MAAR 781 S RF Dower crater 1-30W [75GPW 7620 Teiegrapt, 7. Data generator 0195AVO VCM163 Valve charactensttc mete, 0350RACAL 9301 OF Vcelmeter 1 5GH, £450RACAL 383 4- TracedespIay oscilloxope [175

VAL RADIO .'4V [IC or,a to 230V 50H./ output IPOW "ere WaveE185 (new). £125 )oo.d)

[45

ALL OUR EQUIPMENT IS SOLD IN EXCELLENT. FULLYFUNCTIONAL CONDITION AND GUARANTEED FOR 90 DAYSMAIL ORDERS AND EXPORT ENQUIRIES WELCOMED. PLEASETELEPHONE FOR CARRIAGE QUOTE. ALL INSTRUMENTS AREAVAILABLE EX.STOCK AS AT COPY DATE GOOD QUALITYTEST EQUIPMENT ALWAYS WANTED FOR STOCK. PRICESQUOTED ARE SUBJECT TO ADDITIONAL VAT

80070 Pulse generator6555A Spectrum anaiyser andto 40C/H4

400F Vorlmele, '00,0 300V 4MH,73$11R Cal.b,ato.4800A Vector impedance meter116020 T/ansistor 11.lure to 8745A11600E1 As above8004 pulse gene,ator7045A X Y Plotter 01,1001 Teme-basei355E Programmable anenualor 1GH, 0 12c16355F As above but 0 120db en 1006 steps3490A Deptai Munt.rneter

8733A tin ino0ule1o/8.55313 5oectr um analyser plug./n 110MH4334A A,,,,,reats deto/r/or mete/

[503

[2950C2250195

750[450

[225[750

[95[95

[450[4501500£600

HEWLETT-PACKARD Spectrum Analyser system t II Whiz ,_oruousing141T meinarerne 85528 I F plug-in and 85538 110MHz HE Plug In UnaExcelleua condition throughout [3250

Banco Model CTVM 337 12 stud. cos., Wiens..., rnonata Compositevideo fro inputs Excellent condition £850

MARC( 'NI 2430A 80MH4 7 digitMARCONI 2435 2GHz 8 chgetPHILIPS PM6672 1GH4 8 digit and doer

0100[750[550

HP8407A

With pes 84127, Phase -Magnitude and 84148 POlar display units 110MHa

Recent Celebration 02500


PINEAPPLE SOFTWAREPrograms for BBC computers with disc drive. FREE updating service on all software.

DIAGRAMStill the only drawing program available t ,, ;lb ,r) gives you Inn 'y to draw really large

diagrams and scroll them smoothly around tne screen stopping to edit them at any time if requiredPineapples unique method of storing the diagram information on disc means that the sae of diagrams is

limited only by the free space on disc and not Me amount of computer memory you nave evadable IA blank80 track disc will allow up to 39 mode 0 screens of diagram)

The superb prInt routines suppled with the program enable large areas of the diagram rc be panted In asingle print run in a number of different sizes and rotated through 90 deg if required Full rise can also bemade of printers which have a wider than normal carriage available

The program is fully compatible with the Marconi Tracker ball described below

PLEASE STATE 40 or 80 TRACK DISC A WHETHER STANDARD BBC or MASTER VERSION. ISREOUIRED

PRICE £25.00 VAT

DIAGRAM UTILITIESA suite of six utility programs which add additiona, features to the Diagram draw,ng program The utlitieS

include the saving and loading of areas of diagram to and from disc The ability to display the whole of yourlarge diagram on the screen at one time on either J*4 or 84r8 screen format) Tne addition of borders andscreen indents to diagrams and the ability to shift a whole diagram in any direction

PRICE £10.00 VAT

PCBThis new reiease from Pineapple is a printed C draughting aid which is aimed at producing

complex double sided PCBs very rapidly using a standard BBC micrc and any F X compatible dot-matrix

printerThe program is supplied on EPROM and will run with any 32k BBC micro (including Master series) Also

Supplied is a disc containing a sample PCB layout to demonstrate the programs featuresBy using an EPROM for the program cede the maximum amount of RAM is available for storing component

location arid ASCII identification files et, (Up to 500 components and 500 ASCII component descriptionsmay be stored for a given layouts These is no limit to the number of tracks for a given PCB. altnough the

maximum size of board is restricted to 8' ' 5 6'Using a mode 1 screen tracks on the top side of the board are shown in red. while those on the underside

are blue Each side of the board may be shown individually or superimposed A component placement screenallows component outlines to ore drawn for silk screen purposes and component numbers entered on thisscreen may be displayed during track routing to aid identification of roundels

The print routines allow separate printouts of each side of the PCB in a very accurate expanded definition

1 1 or 2 1 scale enabling Ow" contact printing to be used on resist covered copper clad boardThis program nas too many superb features to describe adequately here so please write or phone for more

information and sample prinouts

PRICE £85.00 - VAT

MARCONI TRACKER BALLThis high quality w r, 'si,wri II,n A r..71,1S...-r 11,16 ro enable it to be

used in place of keyboard keys loysticks or wen your own programsFor Model 13 & 13 (with Icon Artmaster) £6000 VAT Bare Trackerball ,no software' £4900 VATFor Master Series i with Pointer ROM) £6000 VAT Pointer ROM evadable separately C1250 VAT

P6P on Trackerballs V 75

TRACKERBALL ADAPTORSConverter leads

Inc DIAGRAM' Please stalk. 0' 0 . r 1.

PRICE £8.00 VAT

ack,ha software

ADFS Utilities ROMADU is a new product wnicn will prove mvaluaole lo all users of In, ADf S 7, Winchester owners)

The ADU Rom allows direct access to utilities which previously required the use o ADFS utilities disc

Many novel features are included in the commands allowing such facilities as the copying of all the files on a DFSdisc onto ADFS in a single pass and Backing up between discs of different size lincli.ding Winchesters'There is also a very powerful sector editor with automatic correction of CRC bytes and a menu routine with over30 possible operations on ADFS discs

Please write or phone for more details

PRICE £29.00 VAT

MITEYSPICEA new addition to our range of engine,' s_':.,r, s a +ery powerful DC and AC analogue circuit

simulator package with graphic display for any model BBC computer

As well as all the usual facilities available with this type of program, non-linear effects. small signal noise

measurements and sweeps may be performed Component values may be swept. allowing component tolerances tobe investigated as well as thermal performance etc Comprehensive transistor modelling is incorporated using a 20parameter Ebers Moll description The program is supplied on disc with a very comprehensive 49 page manual

Please write or prone for more information

PRICE £119 VAT P&P FREE

ALL ORDERS SENT BY RETURN OF POST

It1,-- 39 Brownlea Gardens, Seven Kings, Ilford, Essex 1G3 9NL. 'N Tel: 01-599 14761*ENTER 200N REPLY CARD


obfID TEST andMEASUREMENTEQUIPMENT

AUDIO TEST EQUIPMENT STG £ POWER SUPPLIES STGLEADER LFM 3615 Wow, Flutter Meter 650 BEHA 30v 3A Single PS 60LEADER LAV 191 Audio Tester 275 DELTA ELECTRONICS SM 6020 60v 20A VAR. 900

LAMBDA 5v 60A Fixed 700COMMUNICATIONS EQUIPMENT RACAL DANA K9232 DUAL PS 175SPECTRON D101 Datascope 3500 SORENSON 60v 10A Var. 525TEKELEX CHAMELEON TTX06 Protocol Analyser 15000 WEIR 423-460 60v 3A Variable 175

WEIR 413-430 30v lA Variable 175DIGITAL VOLTMETERSDANA 5900 61/2 Digit Multimeter 800 PROM PROGRAMMERSFLUKE 8025 61/2 Digit + IEEE 488 1250 DATA IO 29A c/w UnipackFLUKE 8050A 41/2 Digit Multimeter 250 Unipack 2HEWLETT PACKARD 3468 61/2 Digit with HPIL 400 Logic Pack

303A - 002DIGITAL THERMOMETERS 303A- 001 5000FLUKE 2176A 500 DATA IO 16 Gang Programmer 1000DIGITRON 100 GP INDUSTRIAL UV141 UV Eraser 75

RS 301 - 791 UV Eraser 100LOGIC ANALYSERSFLUKE 9010A + Z80 Pod 3500HP 1610A 32 Channel 1500 RFI TEST EQUIPMENTTEKTRONIX 1240 N16 + 3 D2 Cards 8000 SCHWARZBECK NSLK 8127 Artificial Mains Network 1250

SCHWARZBECK FMLK 1518 Low Frequency Receiver 7000FREQUENCY COUNTERS SCHWARZBECK VUME 1520A VHF UHF RF1 Receiver 8250FLUKE 1920A 91/2 Digit 5Hz - 500MHz 900 SCHWARZBECK FMZL 1514 Field Intensity Accessories 2500HP 5315A + Opt 3 DC 10CMHz 900

SEMI CONDUCTOR TEST EQUIPMENTMAINS ANALYSERS & TEST EQUIPMENT MASTECH SEMICONDUCTOR TESTER 2610R3 3000AVO RM215L/2 Insulation Tester 450 TEKTRONIX 576 Curve Tracer 7500ASSOCIATED RESEARCH Hipot Tester 900 3H CORP Linear IC Tester 3H202 1150DRANETZ 626 Mains Analyser 3250

626 PA 6001-1 Plug In 750 SPECTRUM ANALYSERS626 6002 - Al Plug In 750 TAKEDATIKEN TR4132 2500626 6003 - 1 Plug In 1500

ELGAR 1001B Variable Voltage & Frequency Source 1350GOULD GS 2552 Power Conditioner 500 IA PROCESSOR DEVELOPMENT SYSTEMSROD L (Inc) M100 AV -S-5 Hipot Tester 1500 TEKTRONIX 8002A + Z80 Emulator 1500

ROD L 25 AMP Ground Tester 1250 TEKTRONIX 4024 VDU 450SHAFFNER NSG 900 INTEL iSBC 661 Chassis + 8085 Eval. Kit 600

NEC 80C42 Evaluation kit 400OSCILLOSCOPESBWD 540 100MHz Delay T/B 500 MISCELLANEOUSHP 1741A 100MHz Analog Storage Delay T/B 2400 FEEDBACK EW6 Power Meter 250TEKTRONIX 468 Digital Storage Oscilloscope 4500 FLUKE Y8100 Current Gun 200A 160TEKTRONIX 475A DM44 250MHz Delay T/B 2500 FLUKE 80K40 HV Probe 50TEKTRONIX 2215 60MHz Delay T/B 700 RN BELL 615 Gaussmeter 1200TEKTRONIX 2236 100MHz DVM Frequency Control Delay T/B GENRAD 1864 Megohmeter 1200

1500 GOSSEN Light Meters 50TEKTRONIX 2337 100 MHz Ruggedised DVM + Delay T/B LEVELL TM14 Insulation Tester 175

2500 P.S. PH Meter 610-540 50TRIO 0013030 100KHz Oscilloscope 125 SAUNDERS 100HF Crystal Impedance Meter 650TRIO CS2100 100MHz Delay T/B 1500 SHACKMAN 700 Polaroid Camera 300

Prices are subject to availability and are exclusive of VAT

SECOND TEST LTD. 25 CHANCERY LANE DUBLIN 8Tel: Dublin 775166 Telex 91446 Fax 781136

ENTER 19 ON ItEPLY CARD


Acorn's fast riscmicrocomputerMajor manufacturers are developing reduced instruction set

processors for minicomputers, but Acorn's has gone in:o a micro.

corn has launched a rangef high-performanceicrocomputers, which

it says is the fastest in the world.The range is named Archimedes,after the greatest scientist ofclassical antiquity, and it is basedon the company's 32 -bit reducedinstruction set processor.

Two families make up theinitial range: the 300 series,which becomes the new BBCMicro and will carry the supportof the BBC's computer literacyproject: and the enhanced 400series, aimed at scientific. engin-eering and business users.Further series are promised.and both processor andsystem will be available too.e.ms and value-addedresellers.

A feature of the newmachines is their abilityto emulate or co -workwith other processors:in this way they will beable to run "legal"software written for earlier BBCMicros, and with 80x86 emula-tion they will have the power torun MS-DOS applications suchas those for the IBM PC.

The standard operating system is calledArthur, which Acorn says will offer a highdegree of familiarity to users with BBC Microexperience. But the family also supportsUnix: Acorn will supply the Berkeley 5.0version. A windows -icons -mouse -pointeruser interface comes with all Archimedesmachines, together with BBC Basic V. Otherlanguages supported include Ansi C. !so -Pascal. Fortran 77, Prolog, Lisp and Comal.

A wide range of applications software is onthe way from various suppliers, includingthree word -processors, an integrated data-base/spreadsheet/graphics package and aterminal emulation program. The BBC willlaunch a full commmunications package inOctober. Early products from independenthouses will be c.a.d. and electronic pub-lishing packages, a relational database,educational programs and several com-mercial applications.

Archimedes machines can be expanded byplugging modules into a backplane. The first

SPECIFICATIONS

ARCHIMEDES 300Memory: 0.5Mbyte (1Mbyte on 310).0.5M rom.Storage: built-in 3.5in floppy disc. An additionalfloppy or hard disc drive can be fitted internally.I/o: Parallel printer and RS423 serial ports:co -processor bus: Acorn Econet socket.Expansion: backplane for two modules.Price. ex-v.a.t: £799 for model 305: £1075 for440 with colour monitor, excluding educationaldiscountsAvailabil ty: now. Large volume from September.

ARCHIMEDES 400Memory: 1M byte (4M byte on 440). 0.5M rom.Storage: built-in 3.5in floppy disc (plus 20Mbyte3.5in. hard disc on 440). A second drive can beadded to the 410.

I/o: Parallel printer and RS423 serial ports:co -processor bus: Acorn Econet socket supportfor 1162 x 864 high -resolution monitor.Expansioa: four -module packplane.Price. ex -vat: £1399 for model 410: £2499 for440 with colour monitor, excluding discounts.Availability: November (440): early 1988 (410).

modules are for i/o (with userport. a -to -d and 1MHz bus), romexpansion, midi (musical in-strument digital interface) andMS-DOS. Early next year willcome an Ethernet card, a hard-ware floating point unit and aSCSI card. Third party supplierare preparing an IEEE -488(GPIB) module, a high-perform-ance a -to -d interface, a videoframe -grabber and a modem.

All members of the series havea three -box configuration whichincludes an IBM -style keyboard.They offer 18 standard screenmodes, with an additional threeif a multi -sync monitor is used;two support 132 -column text. Upto 256 colours can be used at

once, from a choice of 4096. The'"., sound system allowsusers to call on eight

voices simultan-eously across a

seven -point stereostage. Synthesized

speech is possible byredefining the library of built-insounds.

Acorn's risc microprocessor isa v.I.s.i. design in which many of

the more elaborate internal operationsprovided in other processors have beenpared away to streamline operation of thechip. It gives the new machines a typicalprocessing speed of around four millioninstructions per second.

According to Basic benchmark test resultsquoted by Acorn, Archimedes comfortablybeats rival micros with a speed advantagewhich is rarely less than three or four timesand in most cases is very much larger. Onlyon graphics processing is Archimedesoutstripped, and then by 80386 machines.

BBC Basic V is downwards -compatible.with the versions provided on earlier BBCMicros. but includes new structures such asthe 'Case...When and While...Endwhile'borrowed from Comal. The built-inassembler now generates Acorn riscmachine code rather than the 6502 variety:Acorn says that any programmer familiarwith the 6502 will take to Arm code like aduck to water.

ELECTRONICS A WIRELESS WORLD 839

Dirty designYou will still have some readerswho listen to a.m. radio but I fearthat if my experience is anythingto go by we are members of adying race.

Designers of much equipmentseem to concern themselves onlywith the functioning of thatequipment in isolation andappear completely indifferent tointerference occasioned to otherequipement.

Both switched -mode powersupplies and digital circuitry inmany devices appear to be all therage. My Dynatron (Philips)television set, whether in use orwhether on stand-by, renders itimpossible to listen to eithermedium -wave or long -wavebroadcasts. A neighbouringhotel has installed a Viceroy 6 +16 call connect system whichradiates particularly nastyinterference on long wave,medium wave and the low -frequency end of the short-waveband. Only after threateningBritish Telecom with courtproceedings was some work doneto reduce the interference fromthe hotel telephone installation,

[FEEDBACKbut even now it still remainstroublesome.

My overall impression of theBritish Telecom engineers wasthat they did not have a completeunderstanding of the equipmentin question and were therefore indifficulty in attempting to dealwith the problem.

It seems to me entirelyreasonable that we shouldrequire design engineers toconsider other equipment userswhen evolving their designs andfor manufacturers to go toconsiderable lengths to preventspurious unwanted radiationfrom their equipment. Thematter is serious and displays aslovenly contempt for others.When I was young, a friendrefused to show me how to builda super regenerative receiverbecause he thought that wouldbe unfair to my neighbours. Hewas certainly right and I think ita pity that others are not sopublic spirited.

I recently spent a short holidayin Scotland. Alas that hotel had aViceroy as well! Is there noescape?F. B. KyleWorkington, Cumbria

MathematicsNotwithstanding the commentsof Messrs MacHarg and Clementsin the May issue, my criticism ofMr MacHarg's letter (January)still holds.

I'm not sure what point MrMacHarg is trying to make in hisMay reply. In his first twoparagraphs he has confirmed mypoints that his 'proof "does nothold" mathematically, but in isfifth paragraph he uses the same'proof to show that fallacies canlie behind mathematics. If hewishes to show the possiblefallacies of mathematics, a'proof which breaks several rulesof mathematics really can't beused to argue his case, withoutfirst showing these rules to beincorrect.

As for his ramblings againstmathematics being an activedynamic logic, which came first,the chicken or the egg? MrClements' axiomatic demon-stration of the lack of value ofmathematics fails with his firstaxiom. Mathematics is not just aform of communication. toquote the Oxford Dictionary,mathematics is "The science of

numbers and space". By itsnature, any science is more thana form of communication. I

assume he is confusing thescience of mathematics withmathematical symbols - usedto communicate ideas. In myMarch letter I said "mathematicsis an active dynamic logic andmay be expressed in plainEnglish if so required" meaningthat mathematical symbols maybe expressed in English, not thatthe logic is just that of words.

Mr Clements also seems tohave the wrong idea on math-ematical division and integers.Yes, cells in a nutrient mediumdo divide, or do they multiply?Little matter, they reproduce inan exponential manner. Exponen-tiation, as it happens, can besimplified into multiplicationand further into addition,something I would gladly show,if you so wish.

Division of x by y is the same asasking how many times can y besubtracted from x before beingleft with zero, or an integer lessthan y if only considering in-tegers. As an example, consider abag of ten apples. I can give 10people one apple each only once

NEXT MONTHComputer interference. Using an Amstradcomputer as an example, Richard Marshallanalyses the problem of electromagnetic in-terference from computers and describes theworking of a commercial device designed toreduce the emissions.

Engine management. Pat Jordan looks at theuse of microprocessors in electronic enginemanagement systems. The sensing of para-meters, processing of data and emission andperformance control are discussed and possi-ble future developments in adaptive systemsand electronic driving controls are consi-dered.

Rocks and radio. Medium -frequencies(300kHz to 3MHz) can propagate in rock,albeit with difficulty. Dr Austin examines theprinciples upon which the process dependsand describes the work done over many yearsto solve the problem of reliable subterraneancommunication.

Tripos - life after the Amiga. The workingsand future of this real-time multitaskingoperating system for 68000 family processors.

ELECTRON ICS& WIRELESS WORLD

Computersand radiointerference:a case study

Microprocessorengine control- -nomHow tocommunicatethrough rock

Logic.intention andsexual equality

Handshakingvia statediagrams

711111STripos forindustrialbuses

Logic and sexual equality. Logic 1 and logic 0are either equally significant opposites, or 1 isassertive and 0 is simply the absence of 1.John Kennaugh explores the difference be-tween the two points of view. with referenceto intentional logic symbols.

Memory systems for small computers. Theadvent of 32 -bit processors into the personalcomputer allows the more efficient use ofmemory by the concept of virtual memory,brought about by using paged or segmentedmemory techniques. Rodney Stubbs explains.

Marconi - the Father of Radio. GuglielmoMarconi may. or may not, have been the first.but at least he could make transmissions overa reasonable distance when others were stilltrying to send fifty metres. He was animprover of others' work, but he also had theambition to communicate reliably by radio.

Handshaking state diagrams and field -programmable logic sequencers allow you toproduce your own m.s.i. chips with a PC andlow-cost logic programmer.


(i.e. 10 ÷ 10 = 1) or I can givefive people one apple each, twice(10 ± 5 = 2), else I can give fourpeople two apples each with twoleft over or else give them 21/2apples each (10 ÷ 4 = 21/2 or 2remainder 2). I hope myrambling has demonstratedexactly what mathematicaldivision is.

Interestingly, if reproducingcells are considered to divide (inthe mathematical sense) thenthat would suggest that divisionof the integer one by tworepeatedly (cells 'splitting' intotwo repeatedly) is equal toinfinity, when in fact slicing anapple in half repeatedly will showthat one divided by tworepeatedly tends towards zero.Therefore 'division' of cellscannot be considered as beingthe same as mathematicaldivision. Peno's postulate that'one is a positive integer' has notbeen violated, but shown to betrue.

Assuming that our culturestarted with one cell (it cannotstart with less) as the cellsreproduce we have more cells, sowe have not violated the pos-tulate that 'no positive integerhas one as its successor! I

conclude then that replication isno more than multiplication(and hence addition) and somultiplication, division.addition and subtraction can allbe found in nature.John R. RidleyMasirahSultanate of Oman

Qualityassurance

I find myself in broad agreementwith Mr Whitehead (JuneLetters). I have recently helpedto steer my company through asuccessful application for AQAP-1 and it is true that the focus ofthe system is on enhancingcompany quality performance,not on continuous knuckle -rapping. The only issue I wouldtake with your correspondent ison Secton 4, where he impliesthat QA requirements can berelaxed when ordering from asupplier who can supply the sub-assemblies to BS 5750,components to BS9000, etc. Thisis simply not true: none of thesereleases provide any express orimplied enhancement of quality.

FEEDBACKThey are objective indicatorsthat systems are in place tomonitor and, if necessary, tocorrect problems if they shouldoccur (or before they occur, ifsuccessful statistical methodsare involved).K. V. Castor -PerryBeckenhamKent

Toroidals andsurface mount

I was interested to read thearticle in the March issue on thebackground to toroidaltransformers. Alas (notsurprisingly) it was a bit one-sided in the listing of pros andcons. The following commentsbased on my experience as adesigner might perhaps helpothers.Points in favour of toroids

low radiated magnetic field low profile good rated -power to size ratio(but see below!) acoustically quietAgainst higher interwindingcapacitance vulnerable to "shorted turn"failure if the mounting bolt iselectrically connected both ends,even by accident! output power rating is alsoabsolute maximum rating. i.e. abrick -wall current limit not continously rated at fullpower, so all right for class Baudio, but for power suppliesmust be downrated to 60% typically twice the prite perrated watt on small units Varistor stronglyrecommended to protect theprimary against the "switch -offpulse" even on the smallest cores

I use toroids a lot, but for realpower with "that little extra inhand just in case", a standard E/Icore with its softer saturationcharacteristic is still by far thebest.

On the subject of surface -mount techniques, the recentarticles are highly relevant to themass manufacturing side of theindustry, but there is very littleattention being paid to themaintenance of the product andthe needs of the small quantityuser.

My work includes repair andmodification of modern smallJapanese equipment full of

s.m.ds. Only RS Components, sofar, out of half a dozen industrysuppliers I've checked, stock arange of R's and C's and gettinghold of i.cs and othersemiconductors is like drawinghens teeth from a storeman ofthe old school. S.m.t tools arealso hard to find (I canrecommend the DeSoutter pick-up pen) and very expensive forwhat is on offer.

If we are to be able to make useof the system then it should bemore widely available and lesslike joining the Masons.R. F. StevenslckenhamMiddlesex

RelativityDr Scott-Murray's June letterfurther confuses the issuesraised by the Hafele-Keatingexperiments, in which two sets ofatomic clocks were sent on aseries of airline flights whichtook them round the world inopposite directions, and werecompared before and after theirjourneys. In these experimentstwo distinct effects occurred,namely: -differential ageing,arising because the clockscarried round the world wereexposed to accelerations whichcaused each set to traverse .aclosed path within the inertialframe in which, for the few daysthe experiment lasted, the centreof the earth was approximately atrest- gravitational red shifts. As thealtitude of each aircraft changedthe corresponding change in theearth's gravitational potentialaffected the rates of the clocks itcarried.

The results Dr Murray refersto as the ' "accelerationpotential" term', and as the'Lorentz velocity term- arisefrom alternative ways ofcalculating the first effect. Thefirst way gives an approximateanswer, since it is based on the'Principle of Equivalence', whichis itself approximate. The secondway, which involves the implicitassumption that accelerations donot affect the rate of an idealclock, gives the exact answer.The second effect can becalculated approximately fromSpecial Relativity using theclassical expression for gravi-tational potential, which isusually more than adequate.

To my knowledge noexperiment involving the motionof finite objects, whether usingrotors, or aircraft, or satellites, isclaimed to verify terms in theSpecial Relativity formulaeinvolving powers of v2/c2 greaterthan the first. However theexperiments with his energymesons which Dr Murray socarefully ignores verify theformulae with great precision.

As for his comments on thedesign of the Stanford electronlinear accelerator, what SpecialRelativity actually says is that themeasured velocity of an isolatedelectromagnetic pulse travellingin a vacuum is the same in allframes. The electromagneticpulse travelling down a wave -guide is a composite of manypulses absorbed by and re -radiated from the atoms in thewall of the guide, and so has agroup velocity smaller than 'c'.Thus particles such as electronsmay either fall behind the pulsetravel in step with it, or runahead of it.

In any case, as I pointed out inthe March letters. to beaccelerated the electrons musttravel in step with the waveletswithin the pulse, i.e. not at thegroup but at the phase velocity.The design problems arose fromthe requirement to excite thewaveguide so as to keep thephase velocity smaller than thevelocity of light, while avoidingthe excitation of mixed modes.C. F. ColemanGroveOxfordshire

Re the misinterpretation ofexperiments. If we are given thatunlike charges attract and thatcharges of one kind do not existseparately, it follows that twolike charges will be attractedapart by their companioncharges. There is no logicalnecessity for a repulsive forceand no experiment can provethat such a force exists. The ideaof a repulsive force has survivedfor 700 years but the record forthe subversion of science byexperiment probably belongs tothe Sun God.

A similar situation exists withthe electromagneticexperiments used to justifyrelativity. That theory treats theconstancy of light velocity as akinematic phenomenon.Classical physics treats it


electromagnetically by requiringthat Maxwell's equations apply toall observers and then derivingthe field relations. Both methodsinvolve the Lorentzcorrespondence but, in theclassical case, not as a kinematicrelation between observers. Theend result is the same for bothmethods so the experiments areno more a proof of relativity thanof classical physics.

There appears to be a difficultyin distinguishing Newton'sscheme for the description ofphenomena from thatphenomena. Thus we have suchassertions as 'mass is constant inclassical physics' followed by thederogation of classical physics.This is akin to saying the validityof a computer program dependson the data fed to it. In fact,Newton's requirement thatspace-time be neutral merelyallows each branch of physics tohave its own independent lawsand imposes no restriction onany phenomena, so one wouldexpect his scheme to beapplicable whatever theexperimental data may be. Thisargument needs to be faultedbefore the scheme is rejected andit is not clear how experimentcan do it. Nor, of course, canrelativity's soothsayer logic do it.R. BerrimanPalmerston NorthNew Zealand

Tercentenaryof Newton's

PrincipiaThe tercentenary of Newton'sPrincipia. perhaps the greatestbook ever written,is sadly beingsullied by the limitations andalteratioins allegedly imposed onits contents by Einstein's re-latively (Tom 'vat!, Satellite Sys-tems, E&WW, vol. 93, Feb. 1987p.159. In fact all the republica-tions of the Principia in thetwentieth century are tarnishedby this disgraceful smear withintheir own pages.

Naturally, there are inaccur-acies and errors in the Principia,but Einstein did not correctthem; nor did Einstein general-ise Newton's laws. On the con-trary, Einstein repudiated expli-citly and unreservedly the mostbasic tenets of the Principia,though because of Orwellian

FEEDBACKdoublethink the contemporaryconventional physicists andhistorians seem incapable of rec-ognising this little anomaly.

The most flagrant example isEinstein's rejection of Coperni-can heliocentricity (i.e. thephysical superiority of the Sun'sreference system over the framesof, say, Mercury or Earth, orTitan or Apollo 9, or the OrientExpress) in favour of generalrelativistic polycentricity (i.e.the complete equivalence of allframes, both inertial and non -inertial). This point is significantbecause without heliocentrismthere would have been no Ke-pler's laws, no Newton's laws, norockets, and no artificial satel-lites.

Another elementary relativis-tic error is the extraordinarystatement "force is a mathema-tical fiction, not a physical real-ity" which reduces Newton'ssecond law (f=m a) to a meredefinition of force. the absurdityof this queer relativistic preceptbecomes manifest when statedthus: "a mathematical fiction(force) produces a real effect(acceleration)". As for the thirdlaw, that is rendered pure fictionin its entirety. Of course a law ofnature is a causal relationshipbetween at least two (defined byother means physical realities.

Evidently the only effectiveway to bring the relativists totheir senses is to blast a punch intheir faces, preferably by a cham-pion boxer. (But perhaps therelativists can Lorentz -transform to a frame in whichthe force of interaction betweenfist and head is exactly zero?)Newton of course treated 'force'as the physical reality that it is.

A correct appraisal of the con-tents of the Principia, and of theachievement of Newton (Bri-tain's, and possibly the world's,greatest ever scientist, will bepossible only when the relativis-tic strictures are recognised asthe blatant errors that they reallyare.

But these issues are not mere-ly matters of academic concern.For the relativistic errors impedeand prevent not only theoreticalbut also technological advances.We had the opportunity to pointout in the American Journal ofPhysics (vol. 54, Nov. 1986.p.969) how the confusion stem-ming partly from the relativisticerrors had hindered and delayed

the invention of the laser and ofthe intensity interferometer; andin Nature (vol. 321, 1986, p.734)and Electronics & Power (vol.32, 1986, p.789) how the confu-sion springing entirely from thesame errors has so far preventedthe development of the opticaltranslation sensor.T TheocharisM PsimopoulosBlockett LaboratoryImperial CollegeLondon SW7

Planck'sconstant

In reply to Mr Akil's comments inApril's Feedback and to MrBrindleu's of May on my noteabout Planck's constant and theatomic fine structure constant,February 1987, I give the follow-ing answer. The relationh=27trnecre (a-1) is got by thefollowing premise:

Start with Bohr's basic atomicmodel of orbiting electronsaround a proton nucleus. Theforces acting in such a system areelectrostatic forces and massinertial forces. Assume the massdensity is the same in both theproton and the electrons, andthat the proton behaves like acurrent wire loop in a magneticfield, hence effected by atorsional force when turning outfrom a neutral position in theelectrostatic field between theproton and electron. Thisoscillation creates smalldisturbances in the orbitingmovement of the electron,giving rise to a resonancecondition between thisoscillation and the orbit time.That is the quantum condition.Then assume the frequency ofradiating energy is in proportionto the medium value of thedifference between twosuccessive proton oscillatingfrequencies, then you have thesolution in a nut -shell.

The fine structure constantapproximately constitutes thequotient between the protonmass and the electron massraised to the exponent of 2/3,given by the constant massdensity of both proton andelectron. The model generatesSchrodinger's wave equations aswell as other well-knownrelations from quantummechanics. The model gives the

proton an extension (radius) ofabout 35 fermi as well as newknowledge of the properties,structures and behaviour ofelementary particles. Thus, I saythat the relation above concealsdeep secrets of the nature ofmatter.Ove TedenstigMarstaSweden

Inductivecoupling

Two minor points regardingTom Ivall's article on my range -insensitive links (June issue:The first is that the approximateexpression for the 'pulled' fre-quency is

and not as stated. The secondconcerns Fig. 2. The 74HC04N isa hex inverter; one inverter isused for the first stage of thetransmitter, and five stages inparallel for the second, outputstage.P.E.K. DonaldsonMRC NeurophysiologicalProsthesis UnitLondon

Hugh PocockThe obituary of Hugh S. Pocockin your May 1987 issue bringsback to me many memories overa very long period, for I knewhim well and can emphaticallyendorse your brief summary ofhis personal character andexpecially its gentlemanlynature.

My first contribution to whatwas then "The Wireless World &Radio Review"was in the issue of15th August 1923, and receivedthe distinction of 'top billing' onthe front cover. Even at thatearly date it was issue number209, and cost 4d net (11/2p), andwas registered as a WeeklyNewspaper. Unlike most of mysubsequent contributions, thisfirst one was entirely practical,being a detailed description ofmy invented machine made fromsecond-hand parts to provideabout 800 volts for my valvetransmitter from the then -usuald.c. mains supply.M. C. Scroggie (Cathode Ray)BexhillSussex

Feedback also appears on page786.


Q and oscillator stabilityRecent interest in the origins and use of Q is widened by

involving pendulums and timekeeping

Iwould like to join the recent debate onthe origins of Q, both by adding to thearticle by Dr Smith' with some more

equations involving Q and in the historicalsearch for the first use of Q. In the formercase it is how Q is related to the frequencystability of an oscillator.

I used the factor Q (and then derived theequations given later) by researching in anentirely different field - that of horology. Anumber of years ago I made a precisionpendulum clock (Fig.1 ) which is moderatelyconventional in having a "one second" pen-dulum (period 1/2Hz) with a lead bob andinvar rod. The pendulum, however, is main-tained by a photoelectric sensor and electro-magnetic drive. In addition, the sensor andlogic circuit detect when the amplitudeexceeds a defined value: when this occurs thenext drive impulse is omitted. The energy istherefore supplied on demand more or less

D.A. BATEMAN

as drive -one -period, miss two, drive one.miss one, drive two, etc. Another refinementon the pendulum, which is hidden behindthe dial plate, is an aneroid barometriccompensator to correct for small changes inrate following from changes in buoyancywhen the air pressure changes. Before fittingthe compensator the error was of the orderof - 0.0092 seconds per day per millibarincrease in pressure.

As may be deduced from the above, thetimekeeping is fairly good, having a standarddeviation of better than 70ms/d over a 30 daysample. In crude terms, the clock keeps timeto within about a second a month.

To check the accuracy of the amplitudecontrol (which is }1/s second of arc) I

measured the decrement, and incidentally,the Q. Swinging freely, the pendulum has aQ of 15,400, whereas the presence of the caseand sensor restricts the air circulation to

ERROR VERSUS UALITY FACTOR

PCNNT ON GRAPH CLOCK OR OSCILLATOR Q (Loaded) ERROR

a

b

d

e

gh

k

mn

p

w

Cs -1Cs -2

Rb

Fe'''

y

Balance wheelsCheap alarm clock, pin palletsCompensated, good quality Jewelled leverTorsional ring 'pendulum' (Atmos clock)Hamilton chronometer

Mercer chronometerRandall constant force escapement prototypes 1978.1979

Tuning forksJunghans Resonic (300Hz)Omega Megasonic (720Hz)Bulova 2181 (360Hz)NPL standard, 1934 (1000Hz)

40 15s/d100 4s/d330 2s/d330 1.7s/d

(marine specification)580 0.5- 1.7s/d700 0.29 -0.45s/d

1000250020005000

Laboratory standard, USA 1932 (480Hz.) 52000

PendulumsSynchronome 3400

Big Ben 6600Bateman regulator, photo -electric amplitude control 12700

Gravity survey, transportable doublependulums, low pressure (Gulf Oil)Gravity survey ('Cambridge pendulums')Shortt, low pressure. 30mm.HgFedchenko, low pressure, 4mm.Hg

Quartz crystalsWristwatch, flexure mode. NT cut

Square plate, GT cut (laboratory)

Essen ring

Marconi F3160 laboratory standard. lenticular AT cutHewlett Packard 10543A laboratory standard

AtomicCaesium beam. NPL2 1972Caesium beam. NPL3 1977Hydrogen maserRubidium cell, commercial (Efratom)Radio -active iron, Mossbauer resonance effect

ElectronicHigh stability tuned inductor and capacitor

Superconducting cavity resonator

3500090000

106000200000 -

500000

50000150000200000

10550000

60000002000000

1.8 x1085.7x 10'

2 x 1085x 10'

3 x 10"

200

3x10

2s/d'Is/d'

0 5s/d5 x 10 `(hourly)3 x 10 '(weekly)

10 I

2s/week0.5s/d

0.07s/d

to 'to '

0.008s/d

0.0003s/d

5s/month'10 10(1 minute)

10 17 hours)4 x 10 10(1 hour)

10 8(1 week)1.5x 10 11(1 sec.)

10 11(1 sec.)

2x 10 12

2x10 135x10 "2 x 10 "

10 Is

10 6(1 hour)10 5(1 day)

1.2 x 10 "(10 sec.)

Commercial specifications

give a loaded Q of 12.700.The traditional view in horology was that

control of the amplitude of a pendulumwould eliminate errors due to drift in ampli-tude and thus give immediate and signifi-cant improvements in timekeeping. It willbe recalled that the restoring force is notexactly proportional to amplitude and there-fore a pendulum does not execute perfectsimple harmonic motion. I did not entirelysubscribe to the view about major improve-ments in timekeeping by amplitude control(nor isochronism by other means), butsuspected that the accuracies of clocks as awhole are dominated by the Q of the oscillat-ing component, be it balance wheel, tuningfork, pendulum, quartz crystal or atomicresonance. There is nothing new under thesun, for I subsequently found that a numberof other technical horologists has alsoobserved the link between accuracy and Q,although I believe that I did take the subjecta little further, both theoretically and ex-perimentally.

Again in horology, George Airy (later SirGeorge Airy, the 7th Astronomer Royal). in1826 proved that the act of impulsing apendulum would change its period'. Hederived a relation for the fractional changein period in terms of the mass and length ofthe pendulum, driving force and the escape-ment geometry. Pursuing the link with Q, Iwas able to prove that Airy's equation couldbe expressed more simply in terms of Q as

Ai/=- 1

2Qtanl3

where ..112 is the shift in frequency due to theimpulse phase errorl3.

Ira practical clocks, the majority impulsenear the optimum of R = 90°. Even if this isnot achieved, it is of no great consequence asthe shift in frequency AM compared withthe "ideal" unimpulsed "free" pendulum.can be treated as a constant and corrected forby an almost trivial adjustment of the lengthof the pendulum.

Accuracy, on the other hand. is the vari-able behaviour of the oscillator, and the rateof change of frequency due to small changesin phase angle is

1=2Qsin-

A(3 -

For example, with a Q of 1000. 13 = 90°17r/2)and a phase instability of AB = 1°1-tr/180),then -1w=10-5.

This is a clear statement that accuracy, atleast in the short term, is directly related toQ. (Note the factor of 2 in the equations -quo:ing Lady Jeffreys "Other definitionsappear in the literature: one has to cultivatea robust attitude to the intrusion of 2 orTr." Y1


Short-term accuracy is sufficient formany cases, hut the major interest is almostcertainly in the longer term, particularly inclocks. Again there is the link with Q,assuming, of course, that drift errors due totemperature, etc, are negligible over therelevant time interval. After a good deal ofresearch I was able to collect together datafor Q and accuracy for a very wide range ofoscillators. Although the bias is towardshorology, the table gives the results of thesurvey; they are also plotted in Fig.2. Somewell known clocks are included: one suchexample is the clock at the Palace of West-minster, popularly known as Big Ben. Myown clock - in a domestic environment - fitsneatly on the nominal ,43 = 1° line. For theindividual source references see reference 4.

The overall correlation of accuracy with Qis strong, the main deviations being thecommercially quoted quartz watches andthe atomic clocks. For the watches theprincipal limitations are the productiontrimming tolerances, whereas atomic clocksare refined to suffer the minimum instabili-ties.

Given the mechanical definition of Q

Q=rR

(where R has dimensions of MT -11 theequation supports the more or less com-monsense view that a clock pendulumshould be massive, with low resistance, and alittle unexpectedly for tradition, a highfrequency. It is worth noting that before theadvent of the microelectronically drivengrowth of quartz clocks and watches, therewere the beginnings of trends in thesedirections. Mechanical wrist watches, too,were starting to incorporate "hi -beat" ba-lance wheels (5Hz) in increasing numbers.

It has been pointed out that the accuracy -versus -Q equation is based on the instabilityof the input energy'. This is true, butparadoxically. I have long held the opinionthat it is the output of the energy that isuncontrolled. Strong support is given by theperformance of the non -impulsed Cam-bridge and Gulf Oil gravity survey pendu-lums. These pendulums were swung in -linein anti -phase at low pressure and decayingfreely: measurements of the period withradio time signals gave the local value of g.Whatever the source of instability, an analy-sis of the errors of clocks using Allenvariance shows that for well made clocks (i.e.with temperature control and of soundconstruction) the principal source of error isdue to flicker noise.

The use of Q is certainly beneficial inunderstanding certain aspects of precisiontimekeeping. Conversely, it is a puzzle whytextbooks on vibration theory and oscillatorsmake so little of the benefits of high Q forfrequency stability. Perhaps this article is asmall step in that direction.

To return to the early origins of the term.several years ago I also scoured the litera-ture, with the Handbook of Wireless andTelegraphy, 1938. coming close for a UKpublication. A colleague' carried out a simi-lar search, finding that Terman in his bookRadio Engineering used Q extensively in hisfirst edition in 1932. A second find, with the

844

10

--1

104-.- 0 1

1001

o 001

g 9-0 0001

1015

Balance wheels

bc

xtrx=e:

+-4)(Yi

Tuning forks

Pendulums

Quartz crystals

Rbo

Cs -to

Cs -20

Phase instability AP

10°

1°

1°

Atomic

Z

toor 1 11 11 11 11 11 11 1

106

QUALITY FACTOR IQ)109

Fig.2. Correlation between accuracy and Qfor practical clocks and oscillators: ex-tremes in scatter are associated with com-mercial tolerances and laboratory stan-dards.

quite explicit definition of Q as wUR. wasBayly". 1931. A footnote to the title of thispaper states that the paper was first pre-sented at a meeting in Toronto on 8 May,1929. As a dedicated user of Q. I am delightedthat the hunt started by Lady Jeffreys hastraced the earliest published use of the

Fig.1. The author's precision pendulumclock with photoelectric control. Q of pen-dulum 12,700; accuracy (SD) 70ms/d.

term", that is, to K.S. Johnson"' in 1924. Mycolleague's references, together with thoseunearthed in the search, confirm that Q wasbecoming established in the early 1930s.

Finally, to add yet more to the collectionof formulae involving Q. there is the veryuseful factor of 4.53, where Q equals 4.53multiplied by the number of periods for theamplitude of a decaying oscillation to decayby half.

References1. Smith, K.L.."Q" Elect. & Wireless World. 92, pp51-53. July 1986.

2. Airy. G.B.. On the Disturhances of Pendulumsand Balances, and on the Theory of Escapements.Trans. Camb. Phil. Soc.. 3, pp 105-128. 1830.

3. Jeffreys, B.. A Q-rious Tale; the Origin of theParameter Q in Electromagnetism. Q. MI. R. Astr.Soc.. 26, pp 51-52.1985.

4. Bateman. D.A., Vibration Theory and Clocks(part 31. Horological Journal. 120. pp 48-55.September 1977.

5. Woodward, P.M.. private communication.

6. Woodward. P.M., Stability Analysis -A Branchof Modern Horology. Horological Journal. 129. pp11-13, January 1987, pp 9-11. February 1987. pp15-16, March 1987.

7. Wallman, H.. pfivate communication.8. Bayly. B de F., Selectivity. a simplified Mathe-matical Treatment. Proc. 19. pp 873-881.May 1931.

9. Jeffreys. B.. (lett.) Elect. & Wireless World. 93,37. April 1986.

10. Johnson. K.S., Transmission Circuits forTelephonic Communication. Van Nostrand. 1925.Western Electric 1924.

D.A. Bateman, B.Sc is at RAE. Farnborough.


NEW PRODUCTSTEST & MEASUREMENTS

: 11104,A CP 0COOL RE AOC. RC...3C t

Accurate timekeepingA range of new products. allconcerned with timekeeping, hasbeen produced by Radiocode ClocksLtd. Several of them are related totime and frequency standardsbroadcasts. RCC 8000 isa new mici o -controlled instrument to receive.decode and analyse time -codedtransmissions to provide a secureand automatic time/calendar system.The decoded signal is used to providea reference for the instrument's owncrystal -controlled clock. RMC 1000isa master clock synchronizer whichcan provide precise control of a timesource, such as an atomic clock. It isprogrammed through a front -panelkeypad. It offers a resolution of 100nsand can he pre-programmed to allow

for summer time. leap years, leapseconds and any other correctionneeded. There isa range of master/slave synchronized clock systems.controlled by one or two masterradiocode clocks with a number ofmonitoring, distribution and displayopt ions.

The internal serial time codes usedin broadcast transmissions. such asIRIG. XR3. SMPTE and the like, arecatered for both in generating.receiving and synchronizingequipment.

Similar options are provided forfrequency standards. with rubidiumatomic oscillators, or off -airreceivers with a wide choice of

monitoring and distributionequipment.

Throughout the product range.emphasis has been placed onversatility, compatibility andreliability. Modular systems alloweasy modification or upgrading.Standard inputs and outputs. such asRS232 and GPIB. are used.

Applications are as diverse asmonitoring seismic activity, phase -locking transmitters and receivers.synchronizing computer systems.time stamping, master -slave controlsystems, and controlling powerstation line frequency.

Radiocode Clocks Ltd.Jennings Road. Kernick Road.Penryn, Cornwall TRIO 9LY.Tel: 0326 76007.

New Avo metersA pseudo -analogue 'dynamic pointer'is included in the AVO M2000 rangeof digital multimeters. This is theliquid -crystal equivalent of a needleand. says Thorn -EMI. has a responsesimilar to that of a moving -coilmovement so that varying readingcan he closely monitored. The M2000is not one meter but a whole rangewith a number of different modelsand varying uses and prices from ameter for a trainee electricianthrough to one for laboratorystandards work. They come in bothhand-held and bench models. Thehand-held ones are designed to hedroppable and have features to makethem easy to use, such as a probeholder to allow single handedoperation, with 'range hold'. 'datahold' and 'peak hold' facilities. TheI.c.d. readout has 10.5mm highnumerals and the meters' accuraciesrange from ±0.7f6 on the M2004 to±0.1% on the M2008. The AVOM2008 also has true r.m.s.measurement. Thorn EMIInstruments Ltd, Archclife Road.Dover, Kent CT 17 9EN..Tel: 0304 202620.

Electron -beamintegrated -circuitinvestigatorA new system for diagnosing anddebugging v.I.s.i. circuits comesfrom Sentry Schlumberger. Itcombines a scanning electronmicroscope with cad/cae tools into acompact workstation. When lookingat faults in i.c. structures. designersneed to investigate the circuit underoperation. This usually involvesmicroprohing but there are inherentproblems with this method: theprobes offer loading on the circuit.they can damage it. the circuit needsto be prepared for probing and thereare buried circuits that cannot heprobed at all. With light geometry.probing is more difficult.

The new technique works byfocusing a beam of low energyelectrons onto the device under test.Secondary electrons are emittedfrom the surface of the device. andtheir energy depends on the voltagepresent. An energy filter and electrondetector recover these secondaries.and generate an analogue signal thatvaries with the chip voltage. If thebeam is scanned, the analogueoutput is a video image where thebright areas represent zero voltageregions and the dark areas positiveregions.

"By effectively linking a voltagecontrast image of a live circuit withcad/cae information" says projectdirector Neil Richardson. "Sentry'sIDS5000 achieves unprecedentedefficiency and accuracy in v.l.s.i.debugging, allowing chipmanufacturers to compressprototype development by months".

Using the voltage contrastphenomenon offered by E-heamtechnology. IDS5000 can effectivelyprobe submicron features. makevoltage measurements, displaywaveforms and make comparisonsbetween actual and simulatedwaveforms. Powerful software locksthe microscope display with thelayout display. allowing each displayto track the other, eliminating thegreat 'navigational' problemsinvolved in searching for particularnodes on a chip with hundreds ofthousands of elements.

Extensive system integration hasproduced a high-performance easy -to -use diagnostic system that makesE -beam probing accessible to allengineers. Secondly. advancedsoftware provides links between thesymbolic netlist, geometrical layout.and the physical device. removingthe probe positioning problem forthe user. The system will he ondisplay at exhibitions this year andwill be available for investigatingsub -micron i.cs in 1988. SentrySchlumberger. Ferndown IndustrialEstate. Wimhourne. Dorset 811217PP. Tel: 0202 871718.


[NEW PRODUCTSTEST & MEASUREMENTS

Untended datarecorderService and process engineers canuse this data recorder to capture adata stream and record it on anIBM-PC compatible disc for lateranalysis. The recorder has a standardRS -232 serial port which can acceptASCII or 8 -bit binary data which itrecords onto a 5.25in floppy disc. Theunit can be left unattended, apartfrom the neccessity of changingdiscs. Files may be opened or closedby push-button operation and thebattery hack -up continues storingthe data in a buffer memory while adisc is being changed, thus giving anuninterrupted record. A warninglight indicates that the disc is nearlyfull. Monocle Ltd. 12 Market Square.Leighton Buzzard. Beds LU7 7EY.Tel: 0525 383883.

Transducer monitorA 5 -input module will display thelevel of transducer signals of any oneof the inputs. selected from the frontpanel. PS -D5 is the latest addition tothe Entran range of tranducermeasurement systems. When usedwith the PS -30A transducer supplyand amplifier module when it can hecalibrated in engineering units, andcan display up to ± 19.99 of theselected units. The module runsfrom its internal 9V battery, andthere is a 'low -battery' signal. Theunit fits into a standard rail to allowmulti -channel systems to heconstructed. Entran Ltd, 5 AlbertRoad. Crowthorne, Berks RG11 7LT.Tel: 0344 778848.

Analogue data acquisition on a PCTwo analogue input cards have beenadded to the Blue Chip range ofindustrial i/o cards for the IBM PC.the AIP card plugs into any PCexpansion slot and allows voltages upto 10V to be read by the processor theAIP-24 has 24 differential channelswhile the AIP-48 has 48 open-endedchannels. They are designed forsampling fast analogue signals inindustrial, laboratory of scientificapplications. APC with six free slotscould read up to 288 inputs.

A high-speed multiplexer is usedto scan the input channels. Low-levelsignals are boosted by the

programmable amplifier. The signalis sampled and converted to digital. A12 -hit a -to -d converter with signdetection logic is used to give aresolution of 1:8000 or 0.0125%.

The hoard is port -mapped toallow it to be addressed by mostprogramming languages. Exampleprograms in GW-Basic and TurboPascal are supplied. The boards arecompatible with the XT and ATcomputers and most clones. BlueChip Technology Ltd. Main Avenue.Hawarden Industrial Estate. Deeside.Clwyd CH4 3PP.Tel: 0244 520222.

Automated audiomeasurementDeveloped specifically for makingautomated measurements for allareas of audio technology, thecompact. multi-funcition NFA-I .from Wandel & Goltermann. isclaimed to eliminate many of thelaborious procedures that arenormally needed for manual tests. Itsmenu -controlled operation allows itto be used as a test set for specialistapplications. Set-up times aredrastically reduced.

Depending on the options fitted.the NFA-1 has up to 30 measurementmodes. The automatic testprogramme permits measurementsto be taken simultaneously at asmany points along the line asrequired. If necessary, it can he left tomake these measurementsunattended. The instrumentcombines both receiver andtransmitter and allows threeoperating modes - transmit. receive.or both together for loopmeasurements. with a time delaybetween the two programs. Transmitand receive programs aresynchronised by means of a start/source/program identification signalsent over the link to he measured.This identification lasts for one

receivers connected. Results can bedisplayed or documented in threeways: v.d.u.. printout or memorystorage. for further processing.

Fitted with an automated linetester the NFA- I can quickly testcheck mono and stereo circuits.Wandel & Goltermann Ltd. ProgressI louse. 412 Greenford. Middlesex.Tel: 01575 3020.

Monitor testerThe MT700 tv monitor tester fromGrundig has a range of signal outputsincluding composite colour, RGB.and RGB-TTL. making it suitablefor testing a wide variety ofmonochrome and colour tvmonitors. The tester provides thecommon standards of 625/875/1249lines at 50Hz. 525/625/735/1023lines at 60Hz, and three additionalline patterns between 512 and 1535lines with a maximum line frequencyof 38kHz at 50Hz.

Each tester is programmed to suitindividual customer requirementsand the V and I I pulses may be varied

through the t.t.l. output to allowtesting of monitors with pulseconfigurations deviating from thestandards.

The instrument offers a range oftest patterns including a circle andstandard colour bars. Other featuresinclude 16 different charactersignals to evaluate sharpness andreproduction of vertical characters: a

multiburst signal for frequencyresponse measurements withexpansion up to 18MHz maximum;and an integrated 'monitor test'signal of +2 and 0% to -2 and 100%for adjusting the brightness andcontrast settings respectively.

Positive/negative video inversionis possible with all 16 displayedcharacters and the MT700 also offersa character test pattern whichestimates the quality of the display ofcritical characters on the screen. Theinterlace function can be turned offfor jitter -free monitor test images.Available through ElectronicBrokers Ltd. 140 Camden Street.London NW1 9PB.. Tel: 01-267 7070.


ANEW PRODUCTSAnaloguetraining moduleA bench -top module from FlightElectronics provides a self-containedbase for prototyping andexperimentation of analogue circuitdesigns. The FIT -101 includes abuilt-in power supply. two meters.function generating oscillatorvariable up to 10011z. potentiometer.switches and a loudspeaker. Theworking area is a matrix of 1680tie -points which will accept mostcomponents and all standard -pitchintegrated circuits.

Circuits can he built, powered andtested and modified quickly andaccurately, it is claimed. FlightElectronics Ltd. Ascupart Street.Southampton SO1 IQL. Tel: 0703229041.

rremer..:

Free design servicefor thick -filmhybridsA company that specializes in thick -film resistors is offering a free designservice to companies wanting to usesuch components for p.c.h. andsurface mounted hybrid circuits.tolsworthy Electronics. with its

sister company CGS Resistance.claims to he able to help customers inmany ways. They are particularlyaware of the unwritten codes ofpractice evolved during thedevelopment of such circuits. Theyare willing to work from a completecircuit diagram or "a doodle on theback of an envelope.

"In collaboration with thecustomer.- says Alan Creenwell.technical director. "we can helpdefine component operatingparameters. active or passive, andensure that the circuit design is rightand that it can he manufacturedeasily and competitively.-lolsworthy can he commissioned to

produce protypes and undertake alltesting in manufacture and qualityassurance.

The company can also advise onwhether a conventional p.c.h. circuitcould he better produced as asurface -mount hoard. HolsworthyElectronics (Sales Md. I lacche Mill.South Molton. I EX36 311.Tel: 07695 3151.

New generation teletext decoderThe advent of the subscribe groupbroadcast teletext service hasprompted the development of newdecoders. The information isencrypted so the decoders need toreceive, decode and decrypt theteletext pages. The Data Text fromICC Industries will receive the'pseudo page' format broadcast bythe commerical tv networks and'Packet 31' pages transmittedthrough Ceefax and Oracle.Individual decoders are addressableand will only process informationdesignated for their use. The decoded

data can be displayed by a computeror dumb terminal, printed or shownon a dot-matrix message display.

Data Text decoders have automaticdigital tuning, analogue filtering.full 'hamming' processing and alarge input buffer memory. IGCIndustries are also informationproviders and can transmitinformation to the Air Call computerto be linked to Oracle or the ITNnetwork.IGC Industries Ltd. 202 LondonRoad. Waterlooville. Hants P07 7QU.Tel: 0705 264477.

Clog -free desolder pumpsLle,Ider

pumps feature a totally sealed springmechanism with self-cleaning actionto avoid clogging by solder deposits.There are three models: standard.

anti -static and heavy duty and theyare available from !IRS ElectronicsLtd. Electron House. Great BarrStreet. Birmingham B9 4BB.Tel: 771 2525

Lan linker through Kilos treamLinking Ethernet local area networkson separate sites is possible with theNetlink 5 bridge from MBSCommunications. It uses a dedicateddigital hi -directional line sucli asBT's 641Thit/s KiloStream service or2Mbil/s MegaStream service as thelink between the separate local areanetworks. It will also operate with ashared dedicated line that is accessedby multiplexer. The system isprotocol independent and can link '

similar networks. It satisfies therequirements of ISO level 2recommendations for datacommunications.

The system operates in two halves.one at each end of the line. It cantransmit up to 100 Ethernet packetsin a second at 641thitis and up to2000 packets at 2MhiUs. It has a

buffer memory for up to 128minimally -spaced packets. Internalprocessing enables the system tolearn which packets are local andwhich are to transmitted down theline. Bridges on each side of the linkcan he paired so that only the stationsintended to receive the data can doso. thus offering data security.

The system is part of a range ofequipment for linking networks.Netlink 6 consists of a pair ofsubsystems housed together anduses a dedicated two-way link at4Mbit/s to join two differentnetworks on the same site. Comingare a fibre optic link and a laser orinfra -red link.MRS Communications Ltd.119I I igh Street. Windsor.Berks SL4 6A.

Microsteppingmotor controllerA dual d -to -a converter in a 22 -pin ind.i.l. package can simplify the designof control circuits for microsteppingmotors. The PBM3960. from Rifa. is amicroprocessor compatible dual 7 -hit converter with sign. Analoguevoltage outputs can be set to arequired level and the circuitgenerates two current decay signals.one for each channel. When currentdecay, sign and analogue outputs areconnected to Rifa's P131,3771 amicrostepping motor control systemcan he produced with minimumcomponent count and reducedassembly costs. The PBM3960 iseasily interfaced to Motorola 6800.6808. 6809 and Intel 8085 8 -hitmicroprocessors. All signals are t.t.l.compatible. Two address pins controlwhich of four D -type flip-flopregisters are addressed and these areautomatically reset on power up. Thedevice runs from a nominal + 5Vsupply. Details from Rifa AB. MarketChambers. Shelton Square.Conventry CV1 101. Tel: 020355:1647.

Trainingsystems for tele-communicationsA range of modules has beenproduced by l.1 Electronics has beendesigned to offer hands-on trainingfor electronics engineers andtechnicians. Planned to he part of amuch wider range, the first set ofmodules. Modicom I to 3. covers acurriculum that looks at thetechniques of: sampling andfiltering: pulse amplitudemodulation and demodulation: timedivision multiplexing: transmitter/receiver synchronization. includingp.1.1. pulse code modulation anddemodulation: and error -correctionsystems including the use ofHamming code. A set of course notesICT021 offer a complete curriculumfor a programme of work based onthese initial modules.LI Electronics Ltd. Francis Way.Bowthorpe Industrial Estate.Norwich NR5 9.1A. Tel: 0603 748001.


DATABANK

We sell a wide range of 'second user'computer equipment, microcomputers.computer terminals, peripherals anddevelopment systems.

We also sell test equuipment - all fullyguaranteed.

We buy good quality under utilisedequipment.

CARSTONELECTRONICS

CIRCLE 42

FOR FURTHER DETAILS.

THEELECTRONIC

SOURCEPre Elsarealle !ma/Wm ON/Oreim WSW Mr

41141W11111sr. 11CIRCLE 44

FOR FURTHER DETAILS.

Has access to in excess of £100m. worthof Second User General PurposeTest Equipment. P.C's. InstrumentControls. Dataloggers. MDS Systemsand Audio Visual Equipment. Outlatest catalogue indicates price andspecification of our more readilyavailable items.

Without reference to this Catalogue.can you really he sure that youhave secured the hest possible deal?

SEMICONINDEXES

LTD

CIRCLE 43

I OR FURTHER DETAILS.

This index quotes the major character-istics and maximum ratings of over85.000 international types of both discrete and integrated semiconductor devices.

The index is in three volumes in aloose-leaf format to enable updating froma service offered by the publishers.

Volume I is on TRANSISTORS. Volume2 on DIODES and SCRs and Volume 3 ison digital and analogue INTEGRATEDCIRCUITS.

In addition, each volume contains pin -out diagrams and alternative manufac-turers and distributors.


DATA BANK

let your brochure do the selling

If trio ssaiit III get dour product information into the right hands fast.call Nlartin Perri on III -661 3130.

FIELD ELECTRIC LTD. VISAtaw3 Shenley Road. Borehamwood. Herts WD6 1AA. 01-953 6009

Official Orders,Overseas Enquiries Welcome Order by phone or post. Open 6 days, hallday Thursday. Please ring for C P details 001 shown. Postal rates apply U.K. mainland

only. All test equipment carries warranty All prices incl 15% VAT & c/p unless stated

LIB 00L25 00

,1 /24 95

re,. 0 or o 5. I uA 240 44 12550M7 A 5V IA rer 12775 Fare ookoI s node 55 104 5V IA 17V 4A132.00 - 14,1 '45 1 SA s mode n boned data 160 .00 Man,

L14.95 All P561:41

pm!, heµ liana any.,17.00 ...TO.. 1100 15 watt

'AI hi, 19 98. pal 15

L8.50. p. AC 110ron

Philip, 'A 561..per Molar stet/ angle I H m, 1 T ...".holding !toque 450NMN all data available 111 75 177 00Syn 55150 1010 150 0/ in steppe. mut. 700L3450 p 1 50

Owehty keyboard with Pull n PSII 55 75 key In Oh aNe .

/10.50 3 I() 205,50

liet set IF C right angle 6A 75rrun veeredy ..5 p75c

16 7716 Futon, new mounted or holders or PCB 11195.

NR7 data, assatte doves cti L1600

L220L200

I r 'OW tr. eao h end

fh owl Mostets IRF9531 75W po9rve hen on heats., i 3 00for 4 L115 rah VolIage Regulators I M337T *Iv, 1.'50tor 4 or L1.30 ea4h NFVV ISO 1 701.315

Mato on. Sanders 6604 Al M11, ocre Swirl ii,iintrarne .0176 5 4011147 RI ph* ,1' L2.300 00 , radrual, p l,kase rig

Mar, Sander. V,WP gni/ Mlii11113000

uh5 01. ere, ,..e at 5.:: 0.1340 00 : .

SF l et, Tr 11,, lLIIA,1,10 1275 00 II V I

Tr+h,,r.. 7404N Osi nk,tt ..1. 4155 00 /41araiThe C a plea, rov.

1,n 110000

Telithino 7111 29.1H;4 4. rn ;IS x1132500

TekillIAl 4051 11,4, Topand alpha F1411110.111 88 Lyle' 4anaiiArybond C w .lard L27500 r rr.

Hewlett Pa, kn.l illA fAstartion anah,ei a, hen frequency Slit I. 123000,N(11,14: p tease

1200 00

L46 00

150 00

Hewlett Pal And 9875A Desk?.please ring

Ganpute 1250.00,

1 1, 00, ..

11.11 Angle V,APiletel L40000 plea, ry, ler /V, el Pr gta hatter' , rim and boned 1.114

Hewlett Pa 1510 26454 FAsplay Station with 4. .1,.1W.Mo.& Weal,. up In 96(01 bad High res dwagey hiiediting L115.00 r p please ring

Cod h, I 1 1/50A DMA I 7804 CPU 1 MNth, 1 5MII. P1 Al H M88164 15 54vamps, Pips NFW e, equip /15 95

b-wlett Pa, AMA 673A Sweepop Gever,11. 1 1010. toAU, n PP. rj 10. 0 I rA6 ti. 1460 00

Pant No .' I WIT19348 01 in holder 16 vat'on, L15.00

Al

Card N.. I I 80146 1 024741: S 575546 5 .` 7 1 8201 1 ir aann. 78.94 h 4464 Dvivinii. Hanna,NI W 49 equip 119 95

, . A 16.11 A85 ylmr.lrk yg. Model III /75001 250 00 , Please ring Vihueteli M. glel 1 175.00 161

Ihravelek sg, M. x1..I 14,111 £90.00

eel,, ,onplitil,ror,g

Miles KO Vol, 4,1.6,w Te.anr IT'16,1.0 rig

prnh 1200 Ot

141. Monitor, 7 green 14,4,00 high re. 1.`5/ 111

boxed with all data / in int ilia eh Connect., I.

Ana ho 7511, .imp ulna supplred 119 95 .

ALL OUR PRICES INCLUDE 15% VAT & CARR'PACKING UNLESS STATED. SAVE TIME PHONEYOUR ORDER FOR QUICK DELIVERY WITH ACCESS OR VISA CARDS

Happy MemoriesPart type 1 off 25-99 100 up4164 150ns Not Texas 1.05 .95 .9041256 150ns 2.45 2.20 2.1041256 120ns 2.69 2.50 2.352114 200ns Low Power 1.75 1.60 1.556116 15Ons Low Power 1.40 1.25 1.206264 150ns Low Power 2.40 2.15 2.0562256 120ns Low Power 8.75 8.50 8.252716 450ns 5 volt 3.15 Call Call2532 450ns 5.402732 450ns 3.152732A 250ns 3.952764 250ns Suit BBC 3.1527128 250ns Suit BBC 3.9527256 250ns 4.5527512 250ns 8.95

Low profile IC sockets: Pins 8 14 16 18 20 24 28 40Pence 5 9 10 11 12 15 17 24

Please ask for quote on higher quantities or items not shown.

Data free on memories purchased, enquire cost for other.

Write or 'phone for list of other items including our 74LS seriesand a DISCOUNT ORDER FORM.

Please add 50p post & packing to orders under £15 andVAT to total. Access orders by 'phone or mail welcome.

Non -Military Government & Educational orderswelcome for minimum invoice value of £15 net.

HAPPY MEMORIES (WW),FREEPOST, Kington,

Herefordshire HR5 3BR.Tel: (054 422) 618

(No stamp required)

ENTER to ON REPLY CARD848

ENTER 13 ON 1{1:1'1.Y CAM/ELECTRONICS & WIRELESS WORLD

8051 Project...2

CA 7030 CPU

USER MANUAL

CA 7031,2 CPU

USER MANUAL

FORTH 'It

CA 7031/2 CPU

USER MANUAL

Copyright Bryti,Computers Inc

Programming on our cards using high levellanguage options is so simple it's like cheating!The Cavendish Automation 7030/1 CPU versions support the Intel BASIC -52package. This unique implementation allows direct manipulation of SFRs,internal data RAM, including bit addressing, together with full floating point. Manytasks which would normally require an assembler can now be written interactivelyusing this fast tokenized interpreter.

Of course, some time -critical routines may require an assembler. Neededroutines can be called from BASIC with interrupts being handled at either level,and what's more, an entire function library of over sixty routines may be calledfrom BASIC by the assembler, enabling simple negotiation of floating point, trigfunctions, multiple precision logical operations and string/character handling.

Single -ended 5V operation.

Serial ports up to 19.2K.

Cards support entire range o' 51 Family devices.

72 I/O lines on card.

BASIC version includes EPROM programming hardware on -card.

User selection of data/code memory boundaries.

11 MHz operation as standard. Low cost. Customised options forlow quantity.

We will quote on turnkey systems based on 7000 Series hardware.

CPU cards supported by large range of digital or analogue I/O.including 12 -bit ADC. DAC. Signal conditioning. memoryextension options. power supplies. backplaines and equipmentcases.

The Cavendish Automation CA 7032 CPU versions areavailable for programming using FORTH', and provideinteractive programme development and theadvantages of speed using a compiled language. Aswith the BASIC package, numerous unique primitivesallow full access to the 51 Family architecture.

Assembly language programming is supported togetherwith interrupt handling in FORTH or assembler andtimekeeping is included from years down to machinecycles with practically zilch overhead.

Our relay ladder imp ementation just has to be thesimplest and easiest way to manipulate I/O lines fromthe CPU. The CA 7033 version CPU cards supportRelay Ladder implementations which can stand aloneor operate under BASIC. A unique feature is the abilityto escape at any time from any ladder string into a UserFunction Module (UFM) which is fully defined by theUser. Virtually no restriction exists in calling a UFM,which may be written using assembler or in high-levellanguage, for controlling displays or scanningkeyboards. etc.

Short programmes run at loop speeds measured inmicroseconds, not mil iseconds, and standard featuressuch as RUN on hardware reset/power-on reset arealso simple to invoke. The user has complete controlover how I/O channels function. All outputs may beloaded as inputs by reading the output state.

Cavendish AutomationCavendish Automation Limited45. High Street. St Neots. Huntingdon. Cambridgeshire. PE19 1BNTelephone: 0480 219457 Telex: 32681 CAVCOM G

ENTElt 7 ON 1{EPIA (Alt!)

NEW PRODUCTSCOMPUTING

Five functions onone processorperipheral chipTo use the Fujitsu 8086 processor itwas necessary to use five peripheraldevices. These have now been tackedtogether into a single integrated 100 -pin flat package. the MB89391.

The chip can he connected directlyto the processor and includes a clockgenerator which also synchronizesthe external ready and reset signalswith the system clock. The buscontroller decodes the status signalsfrom the processor and generatescommand and control signals to theperipheral devices. The interruptcontroller handles priority levels forup to eight interrupt signals and canhe expanded by an additionalcontroller up to 64 levels. The directmemory access controller providesfour d.m.a. channels and can accessup to I Mbyte addresses. The fifthsection is a timer and also includedon the chip is a bus arbiter for dealingwith conflicts with other chipsrequiring access to the data oraddress buses.

The combination of thesefunctions into a single packageallows very compact p.c.h. assemblyfor a computer. FujitsuMicroelectronics I.td. I largraveI louse. Belmont Road. Maidenhead.Berks SL6 6NE. Tel: 0628 76100.

A sideways view ofmulti -taskingMulti -tasking usually consists of theability of a computer to run differentprograms at the same time bymultiplexing them together so theyappear to he running concurrently. Anew low-cost system for the 1313Ccomputer takes a different angle; itfreezes the current task, dumps thewhole computer status and memoryto disc and then frees the computerfor another task. A keypress restoresthe computer to where it was, andthe original task may he continued.The Sideset rom allows several tasksto he nested and also providesadditional utilities such as acalculator, a calendar. address hook.notepad and the like. The nestedwindows display allows the output ofseveral programs to he displayed atthe same time. £39.95 mail orderfrom Maze Technology. 11 BraemarAvenus. London NW10 0DY.Tel: 01 452 8372.

mittemittoiolv4\400-\

%rot:mice

wT C'hitx;

/4,4.,

Digital signal processing design kitThe "Designer Chips" tool kit is anew idea for developing digital signalprocessing I d.s.p.) prototype designsand is claimed to offer improvedperformance and greater architect-ural flexibility than other solutions.

The kit combines selected logicc.mos d.s.p. parts into a singlepackage. maximized for ultra -highperformance. Featuring 11 high-performance devices, plus sockets.the kit is optimized for maths -intensive, real-time processing withpipelined architecture.

Logic Devices' kit providesarchitectural flexibility whileminimizing device count by allowingthe designer to tailor the precise mixof logic circuits for each application.In addition. each building block. ordevice. in the kit is optimized forspeed. The kit allows prototyping ofsuch common d.s.p. functions asFast Fourier Transform OTT). andachieves speeds of up to 100 timesfaster than single -chip alternatives.

Applications include video

-1

processing. high speed data analysis.data compression, and imagerecognition.

Designer Chips included in the kitare two LSH32.1C 32 -hit cascadablebarrel shifters, two LRF081C five -Port 8 by 8 register files, twoL4C381.1C 16 -bit cascadable adder/subtractors, two LMA1010.1C 16by 16 multiplier accumulators.two LMU217.IC 16 by 16 micro -programmable multipliers, and oneLPR520.1C 4 by 16 multi -levelpipeline register.

All devices are packaged in surfacemount plastic leaded chip carriers. Inaddition. 11 p.l.c.c. sockets areprovided for easy prototyping. alongwith an up-to-date application noteand a complete product cataloguethat contains technical articles.Abacus Electronics is the soledistributor for Logic Devices'products in the UK. AbacusElectronics Ltd. Bone Lane.Newbury. Berks RG14 55F. Tel:0635 36222.

PC prototypingboardA prototype card is suitable forIBM PCs and compatibles. Known asOption One, the card combines aprototyping area with a built-inextender. It includes power buses.identification references etched inthe copper for all pad positions, anoice reduction grid and two surface -mounted component positions for68 -way chip carriers. The edgeconnectors have nickel/gold contactsand all pad positions are soldermarked to remove the risk of soldershorts. The card is fitted with anAT -type extender connector so that acircuit can he continued on anadditional card. without having toaddress an additional slot positionfrom the computer. The extensioncan be easily snapped off if not usedon an AT -compatible computer.Market Options Ltd. 75 MiddleGordon Road. Camberley. SurreyGU15 2.1A. Tel:0276 29005.

16 -bit processor from ZilogA 280 -compatible. 16 -hit processor.is the Zilog 2280. Fully compatiblewith existing 7,80 applicationssoftware, the device interfacesreadily with all 7.80 familyperipherals in its 8 -hit data busmode, and to Zilog's family of 7. -Busperipherals in the 16 hit data busmode.

Enhancements on the device aresimilar to those found on Zilog's 32hit processor, the 280000, andinclude a three -stage pipelined c.p.u.architecture. 256 bytes of fullyassociative cache memory. capable ofusing the burst mode access featureof nibble mode d.rams. and a

memory management unit capableof addressing up to 16Mhyte ofmemory. Complementing thesefeatures are on -chip peripherals suchas three 16 hit counter/timers. fourhigh speed d.m.a. channels and a fullduplex uart.

New instructions include signedand unsigned multiply and divide aswell as 16 hit variants of the Z80instruction set. The 7.280 also hasseveral new addressing modes whichextend the existing 7,80 addressingmodes. Other on -chip peripheralsinclude a bus interface unit whichperforms automatic wait stategeneration. bus scaling and global

memory handshaking. a 120 hitrefresher/controller and d.rams anda crystal oscillator. Other on -chiplogic can he initialized at reset to linkone of the 7.280's d.m.a. channels tothe uart which then bootstrap loads a256 byte program into programmemory. allowing the possibility of atotally romless system. The full Zilogpart number is Z8028010VSC. whichis a 10MI lz 7.280 in a 68 pin plasticleaded chip carrier. AvailablethroughGothic Crellon Ltd. 3 The BusinessCentre. Molly Millars Lane.Wokingham. Berks RG11 2EY. Tel:0734788878.


DEALER

ENQUIRES

WELCOME

A no-slot clock.The SMT No -Slot Clock from Parasitic. Buyit. Install it in any 28 -Pin ROM socket. Andyou have a real-time clock/calendar. It's thatsimple. Seconds (hundredths of them). Minutes.Hours. Date. Months. Years. All totally transparentto ROM operation, and with two built-in lithiumbatteries to leap you

and beyond. No empty A no-nonsense riceand1988, 1992, 1996 - ce.

socket? Remove thePROM from a full socket, install the No -Slot Clock,and place the ROM on the back of the No -SlotClock. The SMT No -Slot Clock from Parasitic.An absolute bargain. At a knockout price.

ENTER 56O<,4 REPLY CARD

Name.

\ Company.Address.

Telephone

Please send me No -Slot Clocks"

Remittance of £ enclosed

\ For use on computer

\ Software required

Parasitic ResearchInternational,

\ :Marwick. Chia; 5.iildciq 4Steel

EWWWith a no-nonsensethree year

guarantee. IBM -DOS,MS-DOS, Apple ProDOS,

Apple DOS 3.3, Appleworks.You name it, we have the

software to run it on your computer.

Parasitic Research International,Shamrock Quay, Buiding 4,William Street, Southampton SOI 101,Telephone: (0703) 330863

HART - The Firm for QUALITYLINSLEY-HOOD 300 SERIES AMPLIFIER KITS

Superb integrated MosFet amplifier kits with an unbeatable pedigreecircuit design by John Linsley -Hood and engineering by HART.Ultra easy assembly and set-up with sound quality to please the mostdiscerning listener Ideal basis for any domestic sound systems it

quality matters to you Buy the complete kit and save pounds off theindividual component priceK300-35. 35 Watt. Discount price for Complete KitK300-45. 45 Watt. Discount price for Complete KitRLH485 Reprints of Original Articles from HIFI News

LINSLEY-HOOD SYNCHRODYNE AM RECEIVERVery high quality kit for this recent design featured in Wireless World Advanced constructionsystem approved by the Author. uses 3 double sided PCBs in a stacked layout for total stability easeof construction and minimal wiring This module will form the AM section of an ultra high qualityAM/FM switched bandwidth tuner to match our 300 series amplifiers Power supply and tuning gangwill be included with the FM sectionK450 JLH Synchrodyne Kit Special Price £59.95

LINSLEY-HOOD SUPER QUALITY FM TUNERThe long awaited ultra high quality FM companion to the Synchrodyne AM receiver Novel circuitfeatures ready built pre -aligned front end phase locked loop demodulator and advanced sample andhold stereo decoder Circuits featured in Electronics Today International magazine Feoruary andMarch 1987 Complete kits for FM only or combined with the Synchrodyne are cased to match our 300Series amplifiers Send for detailsSpecial Introductory Prices. FM only version f III 69 Version to add Synchrodyne f11813 Bothprices Inc VAT 8 post Send for details

HIGH DUALITY REPLACEMENT CASSETTE HEADSDo your tapes lack treble' A worn head could be the problem Tapeheads are constantly improving and fitting one of our latestreplacement heads could restore performance to better than new'Standard mountings fit most decks and our TC1 Test Cassette willmake it easy to set the azimuth spot on As we are the actualimporters you get prime parts al lowest prices All our heads aresuitable for Dolby machinesHC20 Permalloy Stereo Head Good quality standard head fitted as original equipment on manydecks E7.66HS16 Sendust Alloy Super Head Ouite simply the best Longer life than permalloy higher output thanferrite. fantastic frequency response 0436H0551 4 -Track head for auto -reverse or quadrophonic use Full specification record and play head £14.60Special Offer Stereo RP Heads £2.494 -Track Auto -Reverse Play Head £3.50HS9510 2'4 Stereo DC Erase Head £6.70H0751 E 4'4 Erase compatible with H0551 £39.70Full data on these and other heads in our range are contained in our free list

HART TRIPLE -PURPOSE TEST CASSETTE TC1One inexpensive test cassette enables you to set up VU (Dolbyt level head azimuth andtape speed without test equipment Vital when fining new heads Complete with instructions C4.66

JLH QUALITY AM RADIOSets of Parts available now including Short Wave coils Full kits ready later Send for list

Send for your FREE copy of our lists with lull details of our complete range of Kits Components.PCBs. Cassette Heads and Decks - Overseas please send 5 IRCs for Airmail Post

Please add VAT to all prices Postage on orders up to CIO 50p f 10 to f49 f 1 Over f50 f 150

f98 79[102 36

f 1.05 no VAT

x;40

11111finELECTRONIC KITS LTD1, Penylan Mill. Oswestry, Shropshire Sri 0 9AF24 hr SALES LINE (0691) 652894 Please add VAT

TELESCOPIC MASTS

Pneumatic

Hydraulic RamOperated

Winch Operated

Hilomast Ltd.THE STREET - HEYBRIDGE - MALDON

ESSEX - CM9 7NB - ENGLANDTel: (0621) 56480 Te ex: 995855


ELECTRONICS & WIRELESS WORLD 85 I

LOW COST ELECTRONICS C.A.D.monism

TV If AMPLIFIER

IBM PC (and compatibles). RM NIMBUS. BBC MODEL B.,B+ and MASTER, AMSTRAD CPC and SPECTRUM 48K"ANALYSER" I and II compute the A.C. FREQUENCY RESPONSE oflinear (analogue) circuits. GAIN and PHASE INPUT IMPEDANCE.OUTPUT IMPEDANCE and GROUP RELAY (except Spectrum version) arecalculated over any frequency range required The programs are in useregularly for frequencies between 0.1Hz to 1.2GHz. The effects onperformance of MODIFICATIONS to both circuit and component valuescan be speedily evaluated.

Circuits containing any combination of RESISTORS. CAPACITORS.INDUCTORS. TRANSFORMERS. BIPOLAR and FIELD EFFECTTRANSISTORS and OPERATIONAL AMPLIFIERS can be simulated - upto 60 nodes and 180 components (IBM version).

LAI

PI lot

P? IOC

b.-

Ideal for the analysis of ACTIVE and PASSIVE FILTER CIRCUITS. AUDIOAMPLIFIERS. LOUDSPEAKER CROSS -OVER NETWORKS. WIDE -BANDAMPLIFIERS. TUNED R.F. AMPLIFIERS. AERIAL MATCHING NETWORKS. TV I.F.and CHROMA FILTER CIRCUITS. LINEAR INTEGRATED CIRCUITS etc.

STABILITY CRITERIA AND OSCILLATOR CIRCUITS can be evaluated by'breaking the loop-.

Tabular output on Analyser I. Full graphical output. increased circuit size andactive component library facilities on Analyser II

Check out your new designs in minutes rather than days.

ANALYSER can greatly reduce or even eliminate the need to breadboard newdesigns.

Used by INDUSTRIAL. GOVERNMENT and UNIVERSITY R & D DEPARTMENTSworldwide. IDEAL FOR TRAINING COURSES. VERY EASY TO USE. Prices fromf20 to f195

Full AFTER SALES SERVICE with TELEPHONE QUERY HOT LINE and FREE update service.

OSaa l 0For illustrated leaflets and ordering information please contact Ref: WW

HunmIsff FedHarding Way. Somersham Road. St Ives, Huntingdon. Cambs. PE17 4WR. Telephone St Ives (0480) 61778


QuarthandDEC Bought and Sold

OSCILLOSCOPESHP180A L235TEK7603 Mainframe 100MH: E1200

PLUG -INSTEK7A22 Dill Amplifier £850TEK7A26 Dual Trace Amp 200MHz E1200TEK7B53A Dual Time Base £950TEK7B80 Delayed Timebase 400MHz £850T EK 7B85 Delaying Timebase 400MHz C950

DVPA'sDatron 1051 51/2 Digit Autoranging £650Solartron 1765 E350Solomon 7065 617 digit C750

Solomon 7075 £500Solartron 7050 £450

VARIOUS TEST EQUIPMENTHP3400A RMS Voltmeter £425HP5326A Counter Timer £175HP5370A Time Interval Counter 100MHz £400Fluke 895A Differential Voltmeter £200Marconi TF2162 Attenuator 0-111dB £100Marconi TF2430 Freq Counter 80MHz C200

Racal Dana 9500 Counter Timer f250HP5940I A IEEE Analyser £550HP5004A Signature Analyser C395

COMPUTER EQUIPMENTTektronix 4051 E400

UNISYS11/346, DI. 11W, KK 1 I A. FP -11 £750DL-I1A Mux 'X375RL-11 RLO Controller £475DR -11C Gen Purpose I/1 £75DN- 11 Telex Uf [450LP -II Printer I/f C200

8X01 Floppy drives £300RK II + 2 RK05.1 Drives vac £250RX -211 8502 Controller £210

0 -WSBA-11NE Box, psu (new), H9273 B/place C175

PDP 11/03, Dual 8", 64KB £450DIV 1114 -port serial C230

II,34A £57511;03 DLVII I. RXO1 £600

ENVIROMENTAL CHAMBER for component testing, etc

MISCELLANEOUSNorrhSiur horizon + Tape backupTektronic 4051(r.- Ink + PlO Module NEW

shi 7Ht 96tpi DSDD Floppy

MODEM

WINCHESTER DRIVESlui. Height ST506 (New)

CORVUS 10MB4M for Apple

45MB full Height Voice Coil Atosi 40367 Heads 40 mSec Average Access

DAISY PRINTERSDiablo 63080 Serial INewlDiablo 63080 Sheet feederOlympia ESW3000 ParallelTEC Storwriter 110/40 (New) 12Bit f

Olivetti PR430 SerialDiablo 1345A 12 Bit I/F, + TractorNEC Spinwriter 3510, as new, RS 232

MATRIX PRINTERSTexas Silent 700 Serial 80 ColDec LA36 with Datasouth BoardA1650 Ink Jet

PLOTTERSHP7221(, HPIB, 8Pen A3HP98725 with Feed & CutterCalcomp 1012 Serial 4 Pen

VDU'sTelevideo 925Televideo 950Pericom 6803HP2624 with Thermal PrinterHP2621HP2647A Graphics Terminal

PSU'sGould MGI2 10 Switch ModeGould MG24 5 Switch ModeGould Triple Output 15V1 IA, 5V40A 12V4A

-55 to 200 C

£395£400C403

£90

£270E350E375

£150£175

E300

£650£175£450C550£375E175£375

E125

£175E120

£875£720£120

£160£220£125£275£l50£350

£25E25

£100

£750

ALL PRICES PLUS VAT AND CARRIAGF

Ring Colin, 01-885 5522Unit 4, Hotspur Estate, West Road, Tottenham, London N17 OXJ.

Vabtactio50 YEARS OF EXPERIENCE

IN POWER CONVERSION1937-1987

[ j DC -AC Inverters (Transvertors)

SINE/SQUARE WAVE INVERTERS 30 to 2000watts, 50 (or 60) Hz, 115/230V, single phaseAC. From 12, 24, 50, 110 or 220 Volts DC.

0 DC -DC Converters0 AC -DC Power SuppliesE AC -AC Frequency Changers

For details:VALRADIO POWER LTDLAWRENCE ESTATE, GREEN LANE, HOUNSLOW TW4 6DN. UKTel: 01-570 5622

i - - c

AC POWER FROMDC SOURCE

E Battery ChargersE Standby/UPS Systemsr] Line Conditioners

ENTIAt tts ON I{1:1'1.1 \ It I)

IN VIEW OF THE EXTREMELY RAPID CHANGE TAKING

PLACE IN THE ELECTRONICS INDUSTRY. LARGE

QUANTITIES OF COMPONENTS BECOME REDUNDANT.

WE ARE CASH PURCHASERS OF SUCH MATERIALS AND

WOULD APPRECIATE A TELEPHONE CALL OR A LIST IF

AVAILABLE. WE PAY TOP PRICES AND COLLECT.

R. Henson Ltd.21 Lodge Lane, N. Finchley, London, N12 8JG.

5 mins. from Tally Ho Corner

Telephone: 01-445 2713/0749

ENTElt 32 ON CARD

.-,

ENTE It :Ili ON IIF:PLY (AIN)

ELECTRONICS & WIRELESS WI

ISTMAT PRODUCTSCOMPONENTS

Flash analogue -

to -digitalconverter for video\ low cost's -hit parallel flash a -ho dconverter uses 256 comparatorsacross a precision input referencenetwork. The comparators' inputsare fed to latches and then loanencoder network to produce an 8 -hitdata byte. The data is then converkdto tri-statel.s.-t.t.l. outputs.

The Motorola MC103191. runsfrom a +5V supply and a widetolerance negative supply of -3 to- 6\': the power consumption isabout 6211mW. The full scaledeflection should he between IV and2V. and the analogue input hetweel- 2V and +2V.

The maximum sampling rate is25MI Iz. An over -range hit is providedto allow either over -range sensing orfacilitate the connection of twoMC10319s in parallel to form a 9 -hitconverter. Applications include videodisplay and radar processing, highspeed instrumentation and tvbroadcast encoding. The device isavailable in a 24 pin d.i.l. ceramicpackage. Motorola Ltd. 88 TannersDrive. Blakelands. Milton KeynesMK14 5BI'.

Membrane switch panels

membranes and p.c.hs can make anattractive and functional front panelfor instruments. The p.c.hs from NF1are double -sided will, plated -through holes. Switch circuits areprinted on Inc side. the switch beingmade by touching the membraneoverlay. On the other side of thehoard can he surface-mi iur.ted all

associated components. such asindicators and displays. i.cs. buzzersand relays. This removes the need fora secondary p.c.h. to carry suchcomponents.. There is also nu needfor additional support substrates. Allbutton functions can he printed on the

membrane. NFI Electronics Ltd. I kid -nor Lane Industrial Estate. Newport.Isle of Wight. Tel: i)983 526535.

Hot tweezers forsurface mountingA system designed for the manualassembly of circuits with surface -mounted devices. such as capacitorsand resistors have the additionaladvantage of being safe for use withstatic -sensitive components. TheSMI) solder tweezers from OKIndustries have soldering tips whichallow the simultaneous soldering ofboth sides of a component. avoidingthe problems of cold spots. Thetweezers are earthed through theircontrol unit which can he adjusted tocontrol the temperature between 100and 325' C. OK Industries (1K Ltd.Barton Farm Industrial Estate.Chickenhall Lane. Eastleigh. I 'antsSO5 51tIt. Tel: 070:1619811.

REDUCEDSUBSCRIPTION

Make life easier Dm Dry r iv.and save money t NM I 'at the same timeby accepting our

offers ofreduced -pricesubscriptions.

A one-yearsubscription of £19.50brings 12 issues toyour door Ix the priceof 10. saving nearly179.0 But fcr E53, youcan have a three-yearsubscription andsave nearly 25% -eight free issues.

And you won't have tokeep on ringing up to

as< forphotocopiesof articlesyou'vemissed.

REDUCEDSUBSCRIPTIORDER FORM

Please send me ElectronicsLJ & Wireless World for one yearat the special price of 09.50

Please send me ElectronicsL.J & Wireless World for threeyears at the special price of £53.00'

Inclusive of postage and packing.

I enclose a cheque/PO to the valueofmade payable to Reed BusinessPublishing Limited.

Please debit my credit account.

Expiry date:0 Access 0 BarclaycardNisa.0 Diners Club DAmerican Express

Signed

Name

Job Title

Address

Please return to Electronics &Wireless World, Room 316,Quadrant House, The Quadrant,Sutton, Surrey SM2 5AS.


Hitachi Oscilloscopes leading the way from£320in performance and pricing! +VAT

I he hiv,husi qualit)I he Hitachi name is synonymous with quality and reliability andis hacked by a 2 year or 3 year warranty on every oscilloscope.I he keenest pricing

ith prices starting at only £320 for a 20MHz dual -trace modelHitachi's price -performance ratio can not be bettered.The largest rangeNoss totalling 18 models the Hitachi range covers bandwidthsfrom 20M I 1/ to 150MHz and digital storage models to 60MHz.The fastest serviceWe can supply any Hitachi 'scope immediately from stock and weback it with full calibration and after -sales service.

For colour brochure giving specifications and prices ring 104801 63570ThurlIn Electronics Ltd., New Road, St. Ives, Cambs. PE17 4BG


IEEE -488 controlled laboratory Power Supplies - at low costthe new Thurlby PL- GP series

30V/2A and 15V/4A single and twin units Constant voltage or constant current operation Programmable to 10mV and 10mA resolution Readback of current demand via the bus Twin units have fully independent outputs Remote sensing terminals provided Bench mounting or 19" rack mounting Singles £395 + vat, Twin units £598 + vatThurlby Electronics LtdNew Road, St. Ives, Cambs. PE17 4BGTel: 10480) 63570

1/1 Thu rlbyl+designed and built in Britain

DOD °Ca 4°.

ENTER 4S ON REPLY CA It D

illicrokit --- 0536 744664hr 7 days/weekLtd. 24

741"213 40% orr SAVE OVER 400,0 PRICES SMAS111 D ON ouirLS

.20BC.25

LINEAR11112 .30 ON OUR MEMORIES

DYNAMIC RAH 5V NMOS20 .25 78MI2 .30 EX -EQUIPMENT ; M IS MK.. .ait

4 .20 .25 2` 154 .90 MEMORIES AND memory upgrade AMSrl ['Ai -11 .20 Al ,n808 9.90

EPROMS i ..:56 15 256K x 1 2.3908 .20 .25 . I .40 11.6 15 16Kx4 3.30In .20 .25 4 .40 1000's of delighted 1.464-12 646x4 3.30II .2014 .20

.25

.291.20

, -. .55 customers I ,:511000 12 IM-RIT CMOS 32.00IC511000 10 IM BIT CMOS 35.00

20 .2021

.29 ' el .45'

GUARANTEED UV STATIC RAM 5V RHOS.N27

.29

.27c() .40..,:.'P .65

ERASED AND TENTED .,114Lp 30 16 .4 1.50

30 . - ISO I I4Lp 20 16 x 4 1.80.2032 .20 35

.18I .90

1. 50STATIC RAM SV CMOS

42 49'50 I, 1.75 6116Lp 12 26 xi, 140

51 .2073

.25

.29.45

.45

..:i ,K. 7All hiken from workingI Ely You

6116LP 15 26x8 1.70

62646P 15 81001 2.4074 .20 .39 .70 ',M., Mme will

I, iUNDS ELLE 62256Lp IS 32K .8 14.5085

86 10.59.49 4000 SERIES

..-.-

12M MS SV NMOS93 .45070913

. 49

. 47

.49

i .1818

33.18

FLOPPY DSCCONTROLLERS

19.00

19.00

.!116 35 26 .- 2.70:;32A 30 46 x8 2.65,/6,4 25 8608 2.45.:?128A 2021V 16Kx 8 2.70

23 .60 .55 .3314.26 ';25625 320x 8 3.75

32

38 .4049.

.49

2121

PROCESSORS2.50

., 12-25 646 x 8 8.2020 64K x 8 8A0

313 .40 .49 .18 1.50 EPROM SY CMOS53 .45 .59 18 3.75 ..,C32A 25 46 . , 6.5054 15 33 06430 810,8 4.6057 .40 .55 .22 IC SOCKETS :33'128 25 16K ,38 2.70581 .40 .55 1 .32 .10 326 - 5.2061 .55 .55 .32 .12 641., 9.7063 .55 .55 .26 .16

64 .55 .55 .34

65 .6066 .60

.7560

.28

.18MICROPROCESSORS

1.60 FREE7 .40 .55 .85 5.00

7!-.4.35 2.60 POSTAGE.55

93 .60

.55

.75.28 6 . 90

94 .50 uttl..i.h S 1..L::* * * * HIGH TECH * * * *241 .65 .89 ORDERS U ND ER L25 ADD

244 .65245 .80

.89

.95NEC V20 UPD70108 zi P&P ADD 15% VAT

257 .50273 .65

.55

.89

resdaces 8088 to speed up IBM PC 10-40% High speed address calc-ulanon in hardware EXPORT WELCOME

*5 .40 ,59 Pui compatible with 8088 £3 P&P367 .40 Superset of 8088 instruction set373 .65 .95 Low power CMOS FREE374 .65 .95 8MB. V20 UPD70108-8 613.50

540 .65 8M311. V30 UPD701154 614.50 CATALOGUE643 1.25 * * * *SPOTLIGHT* * * * WITH FIRST ORDER

ORDERS DESPATCHED SAME DAY BEFORE 3 30pmDEPT (W7). 18 COLDORMEADOW AVENUE. CORBY, NORTHANTS NNI89AJ

°

PADS -PCBPRINTED

CIRCUIT LAYOUTSOFTWARE

PADS -PCB is a low cost, yet powerfulsoftware tool for the design of PrintedCircuit Boards. It is unique in offenng acomplete set of sophisticated CAD designfeatures at a price within the reach of anyPCB designer. From Human Engineeringthrough an extensive set of automaticdesign aids, PADS -PCB is distinguished byits flexibility, completeness and power.Human Engineering Horarchical Menu Structure Simultaneous Display of Routing Layers Fast System Response Zoom and Pan Part Names and Signal Names Automatic Database Security Back-up User Control of Colour Osprey (16 colours) True Track. Text. and Pad Size Display Single Connection Net or Track HighlightingSopti. mittlGridcated Layout Database

32". 32" design area 30 Routing Layers Top and Bottom Component Placement ASCII Database Input and Output Unlimited Pad and Track Sizes Fine Line Capability Complete Surface Mounted Device (SMDi

CapabilityBuriedand Thru Via Capebility

Getting StartedPADS -PCB accepts net list input from severalCAE systems Futurenet Dash -3 and Dash.4 SchemeTM Netlist OrcadTM Netlist ASCII

schematics((for manual prepared

Layout without Netlist (Drsign.On The -Fly)Automatic and Interactive ComponentPlacement Dynamic Rubberbanding of Connections Global Move Rotate Command Component Swapping Connection Length Minimization Group Move Main, Placement Rats Nesting

Interactive Track Routing Automatic Via Placement Display of Track end Point while Routing Interlace with PADS -Route Autoroute, Control of Track Widths for each Track

Segment Powerful Track Editing Commands Copper Ground Plane Creation Ability to Partially Route Tracks

Auto Design Rule CheckingDifference Compare Check against Initial NetListComplete Space Check to 1 MilPad -to -Pad Space CheckTrack -to -Track Space CheckTrack -to -Pad Space CheckErrors Displayed by Layer

OutputsArtwork (on Wet Ink Pen Plotter or MatrixPrinter (at I x or 2x)Assembly DrawingPower and Ground PlaneArtwork Duality Pen PlotDesign StatisticsDrill DrawingSilk ScreenSolder MaskNet and Pin ListsEngineering Change Report

CALL FOR FREEDEMONSTRATION

ON 0322 331984

BRIGDEN TECHNOLOGYUnit 24, Erith Business Centre, Erith High Street, Erith, Kent DA8 1QY.Tel: 0322 331984. Tlx: 897958 DARTEX G


854



RADIO BROADCASTIn -car

entertainmentFew would disagree with theview that there have been fewreally significant changes or im-provement in the design or per-formance of domestic radio re-ceivers in recent years, exceptperhaps for a few of the better"all -band" models designed forh.f. performance. Indeed, foroptimum interference -free re-ception of medium- and long -wave stations there are few cur-rent sets that can match some ofthe vintage valve sets of the1930s and 1940s. Similarly, theacoustic performance of, for ex-ample, some of the original Mur-phy designs in the late 1930s,with their baffle -mounted louds-peakers, has too often given wayto the ubiquitous "trannies" soldat prices low enought to attractthe impulse buyer.

But in one area at least, radioand associated audio systemshave been steadily developing,with seemingly little concernwith price, to become the pre-sent sophisticated "in -car enter-tainment" (i.c.e.) systems. Theseembrace equipments selling atprices ranging up to over thethousand -pound mark, includ-ing radio receivers with synthe-sized tuning systems, press -button or automatic tuning tostrong signals, high-fidelitystereo and multiple loudspeak-ers, CB and cellular two-wayradio; electronic navigation andthe various automatic tuningand station identification facili-ties soon to be extended to in-clude those offered by the EBU"Radio Data System" (RDS).

Specialist retailers/installersrather than the manufacturersappear to be reaping the mainbenefit. Manufacturers aregrumbling that this section ofthe market has become heavilyover -subscribed. Mitsubishi re-cently pulled out of i.c.e. andother major firms are reported aslikely to follow.

Buzz words in this area in-clude security coding; micro-computer control to selectstrongest signal, diversity -reception complete with two ae-rials and two front -ends, self -programming, etc. Recently, in -car Compact Disc (CD) units arebeing promoted in the UK, de-spite fears by some that they maylater be squeezed out by in -car

d.a.t. digital audio tape players.Security systems have become

a major selling feature. It is

claimed that in the UK alone onei.c.e. unit is stolen on averageevery minute of the day, oftenaccompanied by damage to thevehicle: no longer is it primarilyjust a question of external caraerials being vandalized. Secur-ity coding systems are now oftendesigned so that any withdrawalof power when removing the unitfrom the vehicle causes the set torefuse to function until an indi-vidual "secret code" is fed in viapush buttons, some with theadded refinement that punchingin an incorrect combination putsthe coding system out of actionfor several hours in order tofrustrate any attempt to find thecorrect combination by repe-atedly trying. Window stickersare often supplied intended todiscourage thieves from attemp-ting to rip out security -codedunits. Clarion. for instance. in-tends also to market the keyunits in a form such that theycan be removed and taken awayby the driver when he leaves hisvehicle unattended.

Some i.c.e. units provide veryhigh -power audio (up to morethan 100 watts per channel)claimed as needed for use withrelatively small loudspeaker en-closures, even though the safetyaspects of having excessivevolume of audio in vehicles andso isolating drivers from externalnoises, including horns ormechanical malfunctions, havebeen questioned by such bodiesas RoSPA.

A survey of "What's new in carsound systems?" in ElectronicsAustralia shows many of thesame trends as are evident in theUK and the USA, althoughmedium -wave a.m.-stereo hasyet to reach Europe and possiblynever will do. Considerableemphasis is placed on the use ofCD players in cars, praised by theAustralian journal as providing,when used with good amplifiersand speakers, "spectacularly bet-ter than f.m. reception or casset-te reproduction in most cars".Another reason is given as "theease of handling and the reducedconcern about dust and grime.Nor do you have to worry aboutrewinding tapes and you canrepeat any musical selection asoften as you want. at will."

It is claimed that "car CDplayers have proven to be reliable

and not at all prone to mistrackbecause of the high vibration andshocks in a moving car. There isstill an element of caution aboutleaving compact discs in directsunlight because of the danger ofwarping, but this applies more tothe CD cases than the discsthemselves."

The journal believes that thebiggest drawback of having CD ina car is the risk of theft, oftenaccompanied by damage to body-work and interior when equip-ment is ripped out by thieves.Sony has a remotely controlledCD unit which is sited in the bootof the vehicle out of sight ofthieves. Other firms incorporatevarious forms of security coding.

Clarion has recently an-nounced that it is to launch ad.a.t. car player in Japan thisautumn but that release of themodel in Europe will hinge uponwhen d.a.t. software becomesavailable, following the long dis-pute over copyright, with therecording industry still seekingto compel manufacturers to in-corporate anti -copying devicesto prevent direct copying ofdigital CD records.

In -shotstereo only?

Stereo radio has taken a longtime to establish itself in the UKand remains a minority listeningmode. There are many engineerswho remain unhappy with thepilot -tone system, not only be-cause of its cross -talk limita-tions, its proneness to interfer-ence and multipath distortionbut also because of the severereduction of coverage comparedwith the monophonic transmis-sions. None of these restrictionsapplies to the digital stereo sys-tems proposed for both terrest-rial and d.b.s. television systems.

Yet there must remain doubt,not on the stereo technology, buton the combination of a largesound stage with a relativelysmall tv picture. Nobody wantsviewers to switch off the picturein order to enjoy the stereosound.

During May, both the BBC andIBA mounted public or privatedemonstrations of compilationtapes featuring a varied selectionof clips from different tv prog-rammes that have been recordedwith stereo sound, including

orchestral, drama, sports, lightentertainment, etc. My impress-ion of both demonstrations wasthe degree to which they empha-sized the need for restraint onthe part of producers and theneed to avoid returning to thatnow distant era when everystereo demonstration featuredtable -tennis and passing trains.Undoubtedly the production ofgood tv-stereo is not easy, whatwith the problem of changingcamera angles, zoom lenses andthe like. These can give rise tosome most odd effects that havenothing to do with the basicexcellence of the stereo sound.Voices or instruments seen in -shot with the sound apparentlycoming from well left or right ofthe picture are decidedly off-putting.

Student radioThe long drawn-out struggleto establish a true, socially -committed "community radio"in the UK has rather lost its wayin view of the stop-go policy ofthe Home Office and the confu-sion between "community radio"and low-cost, deregulated com-mercial radio. It is, however,often forgotten that two types ofcommunity station have beenrunning for many years: hospitalradio on closed-circuit and stu-dent radio using low -powermedium -wave transmitters,often with inductive loops orother techniques to restrictcoverage mainly to the universi-ty campus.

It is interesting to find that along-time advocate of commun-ity radio, Norman McCloud ofthe "Wireless Workshop", aBrighton firm that caters specifi-cally for low-cost and special -event radio for which licenceshave also been available forseveral years, has launched, onbehalf of the National Associa-tion of Student Broadcasters, anew quarterly technical newslet-ter, oddly named Eric. The firstissue includes a discussion onthe use of v.h.f./f.m. stereo forstudent radio, based on a leakyco -axial cable system, and a noteon telephone balancing techni-ques for separating incomingand outgoing speech for phone-ins etc.

Radio Broadcast is written by PatHawker.


(COMPUTER WAREHOUSETHE ALLADINS CAVE OF COMPUTER AND ELECTRONIC EQUIPMENT

BARAINSFOR CALLER1000's

OF

NOT LINE DATA VASE

*DISTIL**The ORIGINAL FREE OF CHARGE dial up database. Buy, browse or place YOUR OWN AD forgoods or services to sell. 1000's of stock items,

spares and one off bargains Updated daily.ON LINE NOW. CCITT, 8 bit word, no parity.

For 300 baud modems call 01-679 1888For 1200-75 baud modems call 01-679 6183

FREEYoon4r, rrttgitor95itfrom its csomapustuepr!El For

HIGH QUALITY COLOUR TV SETThe fabulous TELEBOX an INVALUABLE MUST for the owner ofANY video monitor with a composite input Colour or monochromeMade by a major UK Co as a TOP QUALITY, stand alone UHFtuner and costing OVER £75 to manufacture this opportunity togive your monitor a DUAL FUNCTION must not be missed' TheTELEBOX consists of a compact stylish two tone charcoalmoulded case containing ALL electronics tuner, power supply etcto simply plug in and convert your previously dedicated computermonitor into a HIGH DUALITY COLOUR' TV SET. giving a realbenefit to ALL the family! Don't worry if your monitor doesn't havesound- THE TELEBOX even has an integral 4 watt audio amplifierfor driving an external speaker. PLUS an auxiliary output for superbquality television sound via your headphones or HI Fl system etcOther features include Compact dimensions of only 15 75 w x7.5 d x 3.5 ft latest technology. BRITISH manufacture fully -tuneable 7 channel push button tuner. Auto AGC circuit SAW filter.LED status indicator, fully isolated 240v AC power supply for totalsafety, Mains ON -OFF switch etc. Many other uses

LIMITED QUANTITY- DON'T MISS THIS OFFER!!!

DON'T MISS THE CPM DealOF the CENT RY

The FABUL U CPM TATUN P 2000Professional Business System

A cancelled export order and months of negotiation enables us to otter this professionalPC.CPM system, recently on sale at OVER £1400. at a SCOOP price lust over the cost of

Vs' the two internal disk drives,' Or less than the price of a dumb terminal,'Not a toy. the BIG BROTHER of the EINSTIEN computer. the DUAL PROCESSORPC2000 comprises a modern stylish three piece system with ALL the necessities for the

ISMALL BUSINESS. INDUSTRIAL. EDUCATIONAL or HOBBYIST USER Used withthe THOUSANDS of proven, tested and available CPM software packages such asWORDSTAR, FAST. OBASE2 etc. the PC2000 specification. a' our prices. CANNOTBE BEATEN.'The central processor plinth contains the 64K, Z80A processor. DUAL TEAC 55F 5'/a

Double sided 40/80 track disk drives Mb per drivel. PSU, 4K of memory mapped screen RAM disk controller. RS232,CENTRONICS and system expansion ports and if that's not enough a ready to plug into STANDARDS' DRIVE port for up to FOUR8 disk drives either in double density or IBM format The ultra slim 92 key, detachable keyboard features 32 user definable keysnumeric keypad and text editing keys even its own integral microprocessor which allows the main ZOCIA to devote ALL its time toUSER programs eliminating "lost character" problems found on other machines The attractive detachable 12" monitor combinesa green anti -glare etched screen with full swivel and tilt movement for maximum user comfort Supplied BRAND NEW with CPM2.2. user manuals and full 90 day guarantee. Full data sheet and info on request

PC2000 System PC2000 Business System with CPMwith CPM Etc and 'Ready to Run' FAST Sales and

COST OVER £1,100 Purchase ledger, supports up to9000 Accounts. VAT etc.

COST OVER £1700NOW only E399

ONLY £29.95 OR £24.95 if purchased with ANY of ourvideo monitors Supplied BRAND NEW with full instructions and 2 YEARwarranty Post and pecking £3 SO 'When used with colour crt

COLOUR & MONOCHROMEMONITO SPECIALS

'SYSTEM ALPHA' 14 COLOUR MULTI INPUT MONITORMade by the famous REDIFFUSION Co. for their own professional computersystem this monitor has all the features to suit your immediate and futurerequirements Two video inputs RGB and PAL Composite Video. allow directconnection to BBC/IBM and most other makes of micro computers or VCR'sincluding our very own TELEBOX An internal speaker and audio amp may beconnected to computer or VCR for superior sound quality Many other featuresPIL tube, Matching BBC case colour. Major controls on front panel SeparateContrast and Brightness - even in RGB mode Separate Colour and audiocontrols for Composite Video input BNC plug for composite input 15 way 'D'plug for RGB input modular construction etcetc

This Must Be ONE OF THE YEAR'S BEST BUYS. PC USERSupplied BRAND NEW and BOXED. complete with DATA and 90 dayguarantee ONLY f159.00as above OR IBM PC Version E165.0015 Day 'Cr skt £1.00. BNC skt 75p BBC interface cable £5.50DECCA 80 16 COLOUR monitor. RGB input.Little or hardly used manufacturer's surplus enables us to offer this specialconverted DECCA RGB Colour Video TV Monitor at a super low price of only£99.00. a price for a colour monitor as yet unheard of!' Our own interfacesafety modification and special 16 high definition PIL tube coupled with theDECCA 80 series TV chassis give 80 column definition and quality found onlyOn monitors costing 3 TIMES OUR PRICE. The quality for the price has to beseen to be believed' Supplied complete and ready to plug direct to a BBCMICRO computer or any other system with a TTL RGB output Other featuresare, in'erna! speaker, modular construction, auto degaussing circuit attractiveTEAK CASE. compact dimensions only 52cm W x 34 H x 24 D. 90 dayguarantee Although used units are supplied in EXCELLENT condition.ONLY £99.00 + Carriage.DECCA 80 16 COLOUR monitor. Composite video input. Same as abovemodel but fitted with Composite Video input and audio amp for COMPUTER,VCR or AUDIO VISUAL use ONLY £99.00 + Carr.REDIFFUSION MARK 3, 20 COLOUR monitor. Fitted with standard 75 ohmcomposite video input and sound amp This large screen colour display is idealfor SCHOOLS, SHOPDS, DISCOS CLUBS and other AUDIO VISUAL awl, -cations Supplied in AS NEW or little used condition ONLY £145.00 + Carr

BUDGET RANGE EX EQUIPMENT MONOCHROME video monitors.All units are fully cased and set for 240v standard working with composite videoinputs Units are pre tested and set up for up to 80 column use. Even whenMINOR screen burns exist - normal data displays are unaffected 30 dayguarantee12 KGM 320-1 B/W bandwidth input will display up to 132 x 25 lines £32.9512 GREEN SCREEN version of KGM 320-1 Only £39.959 KGM 324 GREEN SCREEN fully cased very compact unit Only £49.00

Carriage and insurance on all monitors £10 00

1141,1:14.711i1147i1t411GOULD OF443 enclosed. compact switch mode supply with DC regulatedoutputs of +5v @ 5.5a. +12v di 05a, -12v Cu 0 1a and -23v @ 002a Dim 18 x11 x 6 cm 110 or 240v input. BRAND NEW only £16.95GOULD G6 -40A 5v 40 amp switch mode supply NEW £130.00AC -DC Linear PSU for DISK drive and SYSTEM applications. Constructed on arugged ALLOY chassis to continuously supply fully regulated DC outputs of +5v

3 amps. -5v @ 06 amps and +24v @ 5 amps Short circuit and overvoltageprotected 100 or 240v AC input Dim 28 x 12.5 x 7 cm NEW £49.94

Car' aup on all PSU's £300

1:44/ MUM Manufacturer's BRAND NEW surplusDEC LA34 Uncoded keyboard with 67 quality gold plated switches on X -Ymatrix - ideal micro conversions etc. £24.95AMKEY MPNK-114 Superb word processor chassis keyboard on single PCBwith 116 keys Many features such as On board Micro Single 5v rail full ASCII

Coded character set with 31 function keys numeric keypad, cursor padand 9600 baud SERIAL TTL ASCII OUTPUT, Less than half price

Only £69.00 with data Carriage on Keyboards £3.50

E) 1..e; Ft L, :if'

NOW only £499Carriage & Insurance £ 12 00

SURPLUS SPECIALS ONPRESTEL - VIEWDATA - TELEXPLESSEY VUTEL. ultra compact unit. slightly largerthan a telephone features A STANDARD DTMFTELEPHONE 'tone dial) with 5 CPT monitor andintegral modem etc for direct connection to PRESTELVIEWDATA etc Designed to sell to the EXECUTIVE atover £600" Our price BRAND NEW AND BOXED atonly £99.00DECCAFAX VP1 complete Professional PRESTELsystem in siimline desk top unit containing Modem.Numeric keypad. CPU. PSU etc Connects direct tostandard RGB colour monitor. Many other featuresinclude. Printer output Full keyboard input. Cassetteport etc BRAND NEW with DATA A FRACTION OFCOST Only £55.00ALPHATANTEL. Very compact unit with integral FULLALPHA NUMERIC keyboard. Just add a domestic TVreceiver and you have a superb PRESTEL system andvia PRESTEL the cheapest TELEX service to befound,' Many features CENTRONICS Printer output.Memory dialling etc. Supplied complete With data andDIY mod for RGB or Composite video outputs. ASNEW only £125.00

Post and packing on all PRESTEL units 8 50

EX -STOCK INTEGRATED CIRCUITS4164 200 ns D RAMS 9 for £11 4116 itsE1.50 2112 £10.00 2114 £2.50 2102 £2006116 £2.50 EPROMS 2716 £450 2732 £3002764 04.95 27128 £5 50 6800 E2.50 6821 £168A09 £8 6E1E109 10 8085A E5.50 8086C158088 £8 NEC765 £8 WD2793 £28 8202A£22 8251 £7 8748 £15 Z80A DART £650Z80A CPU £2 00 Thousands of IC'S EX STOCKsend SAE for list

DISK DRIVESJapanese 5'. half height 80 track double sided diskdrives by TEAC. CANON. TOSHIBA etcSold as NEW with 90 day guarantee ONLY £85.00TEC F13,503 Double sided HH 40 TRK NEW £75.00SUGART SA400 SS FH 35 TRK £55.00SIEMENS FDD100 SS FH 40 TRK £65.00carriage on 51/4" drives £5.50Brand NEW metal 51/4" DISK CASES with infernal PSU.OSKC1 for 2 HH or 1 FH drive £29.95 +pp 04.00DSKC 2 for 1 NH drive £22.95 +pp E3.50DKSC 3 As DSK1 LESS PSU £12.95 +pp £2 50DSKC 4 As DSK2 LESS PSU £10.95 +pp 02008" IBM format TESTED EX EQUIPMENT.SHUGART 800/801 SS £175.00 +pp E8 50SHUGART 851 DS £250.00 +pp 0850TWIN SHUGART851's2 Mb total capacity in smart casicomplete with PSU etcMITSUBISHI M289463 8 DS 1 Mb equiv. to SHUGARTSA850R. BRAND NEW at £275.00DYSAN 8" Alignment disk £29.00 + pp El aVarious disk drive PSUs Es Stock SEE PSU section.HARD DISK DRIVESDRE/DIABLO Series 30 2 5 Mb front load £525.00Exchangeable version £295.00. ME3029 PSU £95.00DIABLO 44/ DRE4000A B 5+5 Mb from £750.00CDC HAWK 5+5 Mb £795.00. CDC 9762 80 Mb RM03etc £2500.00.PERTEC 03422 5+5 Mb £495.00RODIME 51/4" Winchesters ex -stock from £150 CALLClearance Items- Sold as seen - No guaranteeICL 2314 BRAND NEW 14' Mb Removable pack harddisk drive, cost over £2000 with data ONLY £99.00BASF 6172 8 23Mb Winchesters £199.00Unless stated all drives are refurbished with 90 dayguarantee Many other drives and spares in stock - call

sales office for details

PC2000 Wordprocessor Systemwith CPM and TEC FP25 daisywheel

printer

NOW only £799

MODEMSJoin the communications revolution with our superrange of DATA MODEMS prices and specificationsto suit all applications and budgetsBRAND NEW State of the art productsDACOM DSL2123 Multi standard 300-300 1200.75Auto answer etc £268.00DACOM DEL2123A0 Auto dial. smart modem withmulti standard AUTO SPEED detect and data bufferwith flow control etc £365.00DACOM DSL2123GT The CREAM of the intelligentmodems auto dial, auto call index, bufferetc etc £498.00Steisbeck SB1212 V22 1200 baud FULL DUPLEXsync or async optional auto dial £465.00TRANSDATA 307A Acoustic coupler 300 baud fullduplex, originate only, RS232 interface £49.00

Ex BRITISH TELECOM full spec CCITT. ruggedisedbargain offers Sold TESTED with data Will work onany MICRO or system with RS232 interfaceMODEM 13A 300 baud unit only 2' high fits underphone CALL mode only £45.00_MODEM 20-1. 75.1200 baud. Compact unit for useas subscriber end !o PRESTEL, TELECOM GOLD,MICRONET etc £39.95 +pp £6.50MODEM 20,2 1200-75 baud Same as 20.1 but forcomputer end £65.00 +pp E6.50DATEL 2412. Made by SE Labs for BT this two partunit is for synchronous data links at 1200 or 2400baud using 2780/3780 protocol etc Many featuresinclude 2 or 4 wire working self test auto answer etcCOST OVER £800. Our price ONLY £199 +pp E8.00DATEL 4800, RACAL MPS4800 baud modern EXBT good working order, ONLY £295.00 +pp MOO

SPECIAL OFFERMODEM TG2393. Ex BT, up to 1200 baud fullduplex 4 wire or half duplex over 2 wire line. ONLY£85.00 PER PAIR +ppE1000

F r more information contact our Sales Office.

1:114:nSPECIAL BULK PURCHA E of these compact. highspeed matrix printers. Built in Japan for the HazeltineCorporation this unit features quality constructiongiving 100cps bidirectional. full pin addressablegraphics. 6 type fonts. up to 9 5 single sheet ortractor paper handling. RS232 and CENTRONICSParallel interface Many other features BRAND NEWand BOXED COST £420 Our price Only £199.00

Dry Fit MAINTENANCE FREE by Sonnenschein& Yuasa.A300 07191 315 12v 3Ah NEW £1 3.95A300 07191312 6v 3Ah NEW £9.95A300 071 q ' 2 6-0-6v 1 8Ah TESTED Ex

Equip £5.99

fillEitriLTLffjpStandard VDU data entry terminals

at give away prices!!OUME OVT108. Current product, state of the artterminal with detachable keyboard, 12 Green screen.2 page RAM. TVI 925. Hazeltine. ADMSA emulations.software setup. 25 a 80. Clock. Swivel and tilt base.Printer port, Function keys etc BRAND NEW andBOXED AT ALMOST HALF PRICE Only £425.00AJ5I0 - EX RENTAL, Z80 controlled. 15 greenscreen 24 a 80 display. graphics, cursor addressing.printer port etc. Very good condition TESTED completewith manual only £225.00ADDS 520 - Dumb terminal, used, 12 b/w screenRS232 interface and printer port. TESTED.ONLY £125.00. Carriage on terminals £10.001DO's of other terminals in stock. CALL for more details.

All prices quoted are for U.K. Mainland, paid cash with order in Pounds Sterling PLUS VAT. Minimum order value E2.00.Minimum Credit Card orderE/0.00. Minimum BONA FIDE account orders from Government Depts., Schools. Universitiesandestablished companiesE20.00. Where post and packing not indicated please ADD £1.00 + VAT. Warehouse open Mon -Fri9.30-5.30. Sat 10.30-5.30 We reserve the right to change prices and specifications without notice. Trade, Bulk and Exportr ri rs 14! r 32 Biggin Way, Upper Norwood, London SE19 3XF

-iiiL I I\ LH II! -I- Telephone 01-679 441 4 Telex 894502 Data 01-679 1888 I* Qi


ERS PRINTERS PRINTERS PRINTERS

SUPER DEAL? NO - SUPER STEALTHE FABULOUS 25 CPS "TEC STARWRITER"

,/ /Made to the veryhighest spec the TECSTARWRITERFP1500-25 features avery heavy duty die castchassis and DIABLOtype print mechanismgiving superb

quality Micro -processorregistration and print 4 fraction of its original cost.electronics offer fullDIABLO/OUME command comparability and tuii con ro via CPM WORDSTARETC. Many other features include bi-cluectional printing. switchable 10 or 12 pitchfull width 381mm paper handling with up to 163 characters per line. friction feedrollers for single sheet or continuous paper. internal buffer, standard RS232 serialinterface with handshake Supplied absolutely BRAND NEW with 90 day guaranteeand FREE daisy wheel and dust cover Order NOW or contact sales office for moreinformation Optional extras RS232 data cable £10.00 Tech manual £7.50 TractorFeed £140.00. Spare daisy wheel £3.50 Carriage & Ins (UK Mainlandi EIO 00

NOW ONLY £499 + VAT

DIY PRINTER MEMBrand New surplus of this professional printer chassis gives anoutstanding opportunity for the Student. Hobbyist or Roboticsconstructor to build a printer - plotter - digitiser etc. entirely to theirown specification. The printer mechanism is supplied ready built.aligned and pre tested but WITHOUT electronics. Many featuresinclude all metal chassis, phosphor bronze bearings. 132 characteroptical shaft position encoder. NINE needle head. 2 x two phase 12Vstepper motors for carriage and paper control. 9.5 Paper platten etcetc Even a manufacturer's print sample to show the units capabilities"Overall dimensions 40 cm x 12 cm x 21 cmSold BRAND NEW at a FRACTION of cost ONLY £49.50 4 pp £4 50.

TELETYPE ASR33DATA I/O TERMINALS

Industry standard. combined ASCII110 baud printer, keyboard and 8hole paper tape punch and reader,Standard RS232 serial interface.Ideal as cheap hard copy ./nitor tape pep for CNC and NCmachines TESTED and in goodcondition. Only £250.00 VOWstand £10 00 C;ir, c'15.00.

EX NEWSSERVICE PRINTERS

Compact ultra reliable quality built unitmade by the USA EXTEL CorporationOften seen in major Hotels printing up tothe minute News and Financial inform-ation. the unit operates on 5 UNITBAUDOT CODE from a Current loop.RS232 or TTL serial interface May beconnected to your micro as a low costprinter or via a simple interface and 'ilterto any communications receive' 10enable printing of worldwide NEWS.TELEX and RTTY servicesSupplied TESTED in second handcondition complete with DATA, 50 and75 baud xtals and large paper ro'I.TYPE AE11

50 Column ONLY £49.95.50Spare paper roil for AE I 1

TYPE AF11R 72 ColRibbon £65.00

TYPE AH11R 80 ColASCII/BAJDOT £185.00

Carriage and Insurance £7.50

20,000 FEET OF ELECTRONICAND COMPUTER GOODIES

ENGLAND'S LARGEST SURPLUS STORE- SEEING IS BELIEVING',

DEC CORNER MAO TAPE DRIVESPDP 1140 System comprising of CPU. 124kmemory & MMU 15 line RS232 interfaceRPO2 40 MB hard disk drive.TU10 9 track 800 BPI Mag tape drive. dualtrack system. VT52 VDU, etc etc Tested andrunning £3,750.00BAll-MB 3 5 Box, PSU. LTC £395.00DH11-AD 16 x RS232 DMAinterface £1.900.00DLV11-J4 x EIA interface £350.00DLV11-E Serial Modem support £190.00DUP11 Synch. Serial data £650.0000200 Diiog - multi RK controller £495.00DZ11-B 8 line RS232 mux board £650.00KDF11-13 M8189 POP 1123PLUS £1,100.00LA30 Printer and Keyboard £80.00LA36 Decwriter E1A or20 mA loop £270.00MS11-JP Unibus 32kb Ram £80.00MS11-LB Unibus 128kb Ram £450.00MS1 1 -LD Unibus 256kb Ram £850.00PDP11/05 Cpu Ram i/o etc £450.00PDP11/40 Cpu 124k MMU £1,850.00RT11 ver 3B documentation kit £70.00RK05-J 2.5 Mb disk drives £650.00KL8 JA PDP 8 async i/o £175.00M18E PDP 8 Bootstrap option £75.00VT50 VDU and Keyboard- 20 mA £175.00VT52 VDU and RS232 interface £250.00

Give your VT100 a Birthday!"Brand New VT100 Keyboards

only C85.001000's OF EX STOCK spares for PDP8. PDP8APDP11. PD P1134 etc. SAE. for list or CALLsales office for details.ALL TYPES OF COMPUTER EQUIPMENT ANDSPARES WANTED FOR PROMPT CASHPAYMENT.

Many EX STOCK computer tape drives and spares byPERTEC, CIPHER, WANGO, DIGIDATA, KENNEDYetc. Special offer this month on DEI Cartridge tapedrives ONLY £450.00 each.

CALL FOR DETAILS

COMPUTER/SYSTEM CABINET 6 PSUAll in one quality computer cabinet with integralswitched mode PSU. mains filtering, and twin fancooling. Originally made for the famous DEC PDP8computer system ccsting thousands of pounds.Made to run 24 hours per day the psu is fullyscreened and will deliver a massive +5v DC at 17 amps. +15v DC at1 amp and -15v DC at 5 amps. The complete unit is fully encltDsedwith removable top lid, filtering, trip switch. power and run ledsmounted on all front panel. rear cable ent'ies. etc. etc. Units are ingood but used condition - supplied for 240v operation complete withfull circuit and tech. man. Give your system that professional finishfor only £49.95 + carr. 19 wide 16 deep 10.5 high. Useable area16 w10.5 h 11 5 d.Also available less psu. with fails etc. Internal dim. 19 w. 16 d. 10 5 h.£19.95. Carriage £8.75

66% DISCOUNT ON ELECTRONICCOPONENTEQUIPMENT

S

Due to our massive bulk purchasing programme which enables us to bring you thebest possible bargains. we have thousands of ICs. Transistors. Relays. Caps. PCBs.Sub -assemblies. Switches etc etc surplus to OUR requirements Because we don'thave sufficient stocks of any one item to include in our ads we are packing all theseitems into the BARGAIN OF A LIFETIME Thousands of components at giveawayprices Guaranteed to be worth at least 3 times what you pay Unbeatable value andperhaps one of the most consistently useful items you will every buy" Sold byweight

2.5kIs C5.25 + pp £1.2510kIs £11.25 + pp £2.25

5 kls f6.90 + £1.8020kts £19.50 + pp £4.75

GE TERMIPRINTER

A massive purchase of these desk topprinter terminals enables us to offer youthese quality 30 or 120 cps printersat a SUPER LOW PRICE against theiroriginal cost of over £1000 Unitcomprises of full OWERTY electronickeyboard and printer meth with printface similar to correspondence qualitytypewriter Variable forms tractor unitenables full width - up to 135 120column paper. upper - lower casestandard RS232 serial interface internalvertical and horizontal tab settingsstandard ribbon. adjustable baudrates. quiet operation plus many otherfeatures Supplied complete with manualGuaranteed working GE30 £130.00.GE1200120 cps £175.00Untested GE30 £65.00 : Drstand £12.50 Carr 8 1-

SEMICONDUCTOR'GRAB BAGS'

Mixed Semis amazing value contentsinclude transistors digital. linear ICs.triacs. diodes bridge recs. etc etcAll devices guaranteed brand new fullspec with manufacturers marking& fullyguaranteed50+ £2.95 100+ £5.15TTL 74 Series A gigantic purchase of anacross the board range of 74 TTL series

IC s enables us to offer 1004- mixed-mostly TIC grab bags at a price whichtwo or three chips in the bag wouldnormally cost to buy Fully guaranteed allIC s full spec 100+ £6.90,200+ £12.30, 300+ £19.50

MANS FILTERSCURE those unnerving hang ups and dataglitches caused by mains interference withprofessional quality filters SD5A match-box size up to 1000 watt 240 VLoad ONLY £5.95. L12127 compactcompletely cased unit with 3 pin fittedsocket up to 750 watts ONLY £9.99.

EPROM COPIERSThe amazing SOFTY 2 The CompleteToolkit for copying. writing. modifying andlisting EPROMS of the 2516. 2716.2532. 2732 range Many other functionsinclude integral keyboard. cassette inter-face. serial and parallel i. o UHF modulatorZIF socket etcONLY £195.00 + pp 2 50"GANG OF EIGHT" intelligent Z80controlled 8 gang programmer for ALLsingle 5v rail EPROMS up to 27128 Willcopy 8 27128 in ONLY 3 MINUTES InternalLCD display and checking routines forIDIOT PROOF operation Only £395.00 +pp 3 00"GANG OF EIGHT PLUS" Same spec asabove but with additional RS232 serialinterface for down line loading data fromcomputer etc ONLY £445.00 + pp £3.00

Data sheets on request

COOLING FANSK eep you, not parrs COOL an, MtGOING FANS ETRI 1261521 2,I.,tn 80 80 36mm (9 95 ETRI9.Dm 92 92 25rmtn ectu.orne. 95GOULD JI1.3AR D.nt 3 3 2 woe,,onnmg 2405 open awn NEW [6 95 Mut t .15ENTAUR908ER standard 120 120 38mm 'arts 5,de 1105 OR2405 3 or 5 blade NEW al t 10 50 or tested EX EQUIPMENT

5 50 low Voltage DC Fans RUINER 69.11 22 8 16v DCmcro mmarature reers.ble uses brusn.53 Ser,o TO,Ofalmost Vent running guaranteed 1 0 000 Mr .11e Measureso kir 62 62 22mm Current cost E35 00 OUR MICEONLY £13.95 complete 55, data 120 120 38mrnDC tans. PANSON1C FE 12C1211 125 DC 5 oiade [1800PRISE 4124% 24 - 285 DC 5 wade CIS 00

1004 of oiler fans Es StockCan for Derail Post & Pac,ng on all fans C2 00

1000's of other EX STOCK items including POWER SUPPLIES, RACKS, RELAYS, TRANSFORMERS, TESTEQUIPMENT, CABLE, CONNECTORS, HARDWARE, MODEMS, TELEPHONES, VARIACS, VDU'S,PRINTERS. POWER SUPPLIES, OPTICS, KEYBOARDS etc. etc. Give us a call for your spare part Wit

requirements. Stock changes almost daily.Don't forget, ALL TYPES and QUANTITIES of electronic surplus purchased for CASH


TELEVISION BROADCASTAircraft flutter

One of the long-standing, butlittle discussed, problems of tele-vision reception on both v.h.f.and u.h.f. is the occasional dis-turbance of pictures resultingfrom the scattering of signals bypassing aircraft. In effect, thereceiver becomes part of anelementary form of bistatic radarthough fortunately for broadcas-ters and most viewers the dis-turbances are not frequentenough or severe enought toprovoke complaints. Many view-ers remain in ignorance of thecause of these occasional picturedisturbances unless they live inthe immediate vicinity of a busyairport.

About 1970, the Rev. PaulSollom, G3BGL, of DouaiSchool, in collaboration withJ.A. Lane of the Radio and SpaceResearch Station (now part ofthe Rutherford Appleton Labora-tory). as part of an investigationinto tropospheric propagation,showed clearly how aircraft scat-ter is responsible for the "steam -train effect" (waa-waa-chuff-chuff-chuff) on v.h.f./f.m. radio,but relatively little research everseems to have been carried out inthe UK into the extent of theproblem on v.h.f. or u.h.f. televi-sion signals.

Certainly in South London,near an occasionally used flightpath into Heathrow, picture dis-turbances can be seen at timeswhen aircraft noise is audibleand sometimes before the air-craft is heard. On the strong localtv signals from Crystal Palace,the disturbance are not undulyannoying but nevertheless areclearly visible. Many viewersnear Heathrow must be affectedto a much greater extent.

Aircraft flutter and the equiva-lent "steam -train effect" aremanifestations of multipath re-ception, accompanied by Dop-pler effect and changing pathlengths. A fresh study of thisphenomenon is described in areport "Prediction of flutteringdisturbance in tv signals causedby aircraft scattering" by ShiroIto published recently (NHKLaboratories Note No 344, March1987) and occasioned by the in-creasing amount of air traffic,the larger size of aircraft and thegreater number of airports.

Initially, to determine thescattering characteristic of a

modern large aircraft, a 1/100thscale model of a Boeing 747,coated with conductive paint,was rotated in a radio anechoicchamber and illuminated with12 GHz signals. It was found thatthe scattering characteristics ofthe model could be modelled forcomputer calculations by a com-bination of 46 rectangular reflec-tors. This finding was subse-quently verified on real aircraftin the Tokyo area in a locationwhere Band 1 signals fluctuatedfor about 15 seconds each time a747 flew into Tokyo airport. Themaximum fluctuating amplitudeof the signals was about 13dB.The report shows that computerstudies can now be used to calcu-late and predict the amplitudeand period of the fluctuation offield intensity, the delay and theratio of desired to undesired (D/UI signal.

The method gives the v.h.f.fluttering zone for a single flight,leaving the problem of multipleflights for the future, and alsothe extent to which this modelcould be applied to u.h.f. trans-missions. But whether there isany practicable way of reducingthe effect on tv screens, otherthan by careful locations of nullsin the vertical radiation patternof the receiving antenna, re-mains open to doubt, and is notconsidered in the NHK report.

Satellitebroadcasting

poses problemsThe time may or may not becoming when we shall see wide-spread viewing in the home ofprogrammes coming direct fromgeostationary satellites. Few nowdispute the value and economicadvantages of satellite distribu-tion links that provide program-me feeds at relatively low cost tomultiple terrestrial transmittersor cable networks. But high -power satellite transponders atthe hundred -watt level are stillonly a promising if risky technol-ogy, held back by the difficultyand launch cost of providingsufficient electrical power on asatellite of reasonable weight toprovide more than about three offour transponders on the satel-lite and the lack of flight -provenhigh -power travelling -wave -tubeamplifiers.

The pulling out of AmstradConsumer Electronics, one ofthe five original founders of theBSB consortium, must make onewonder whether the suggested£200 for the cost of satellitereceiving equipment for use withan existing tv set will be met,particularly since it now appearsthat BSB's channels will be en-crypted and (subject to Govern-ment approval) use the D -MACtransmission standard. Thenumber of viewers who haveproved willing to pay £1000 or sofor equipment capable of receiv-ing cable programmes from thelow -power distribution satellitescan. in the UK, be numbered inhundreds rather than millions;in the USA the sale of C -bandtyros fell away sharply when thepremium film channels beganencrypting last year after thetotal of direct viewers hadreached about 1.6 million. Anumber of the UK firms thatentered this market have alreadyfallen by the wayside.

DBS and, even more so, low -power distribution satellites stilltransgress the broadcasting ad-age that capital costs should fallprimarily on the broadcasterrather than on millions of view-ers. There is also the problemthat DBS cannot cater for "re-gional programmes", so that theexisting terrestrial network willneed to continue for many yearsyet, if we are to avoid tv broad-casting becoming limited tonational programming.

But one wonders to what ex-tent the public is already beingmisled by the enthusiasm ofpopular science writers and themedia in general. Many writersstill seem to think that viewerswill remain fascinated at themere idea of watching program-mes from outer space, whetherin the English language or fore-ign languages. Surely, after a fewdays it is the standard of theprogrammes that will matter.DBS, like cable, is just a means ofextending programme choice,though eventually it may offerhigher definition and widerscreens.

I have been reading the recentPenguin paperback "The begin-ner's guide to satellite tv" byRichard Maybury (121 pages,£3.95) which insists dogmatical-ly that satellite tv is "the tv of thefuture" and enthuses over the 20or so channels already being

beamed "directly at every homein western Europe from orbitingsatellites 36,000 kilometresabove the equator", continuing"In less than five years therecould be as many as one hundredtelevision channels to choosefrom, and the equipment willcost less than £200 at today'sprices. Stick around - we are atthe very beginning of a revolu-tion in communications andbroadcasting that is potentiallyevery bit as important as Marconi'sfirst tentative clicks and buzzes!"

While such books do contain alot of useful background mat-erial on space technology theyalso, unfortunately, often in-clude material that is false ormisleading. What on earth forinstance is one to make of theconfident assertion (page 118 inthe glossary of terms) that B -MAC is a system developed by theBBC (sic) and D -MAC is a high -definition system! Similarly theearlier section on MAC seems tosuppose that the prime purposeof this family of systems is toprovide encryption. It gives Oak/Orion as one of the most com-monly used scrambling systems.with no mentioin of VideoCipherII, which has become virtuallythe de facto satellite -link encryp-tion standard in the USA. Norany warning that many of thetyros currently being sold willnot be suitable for reception ofD -MAC links or DBS systems.

One has no wish to nit-pickerrors in a book that does havesome merits, but it seems a pitythat a firm with the reputation ofPenguin, in addressing a read-ership that needs sound guid-ance in an area of considerablecomplexity and rapid changingof plans and technology, did nottake the precaution of having thetext read by those directly con-cerned with the technology andeconomics of tv broadcasting.

As John Gau put it recently inThe Daily Telegraph: "Forgetsatellites, forget cable, forget thelatest fashionable piece oftechnology that its supportersclaim will revolutionize televi-sion as we know it. The Tories'manifesto commitment to com-pel the BBC and ITV to take 25per cent of their programmesfrom independent producers willhave a far bigger impact on Brit-ish broadcasting."

Television Broadcast is written byPat Hawker


RADIO COMMUNICATIONSHandheldhazards

One of the objectives listed for aservice such as Donald Cox advo-cates is user safety from strongelectromagnetic fields. Thiswould clearly be achieved withthe envisaged power levels of theportable equipment. Severalyears ago, Motorola engineersinvestigated the potential hazardof using hand-held radio equip-ment with short, normal -mode,helical ("rubber duck") antennasonly a few inches away from theusers' eyes. The conclusion thenreached was that a safe limit forsuch a handheld transceiver,based on measured rates of tissueabsorption and heating, was ab-out 7 watts of r.f. output. Sincethen there has been, at lest insome countries, a tendency toimpose lower safety limits in thev.h.f. spectrum than were stan-dard at the time of the Motorolastudy.

In these circumstances it is

surprising that there is still atrend towards providing highertransmitter powers in handheldequipments, limited primarilynot by safety considerations butby battery life. For example, therecently introduced Yaesu FT73handheld transceiver foramateur radio with a very short(2in) "rubber duck" normal -mode helical antenna can pro-vide up to 5 watts of r.f., with theantenna likely to be very close tothe heat -sensitive eyes of theuser. While its power is below thesafe maximum suggested by theMotorola study it may well seema little too close for comfort ofmind. It is perhaps time formanufacturers and users tothink in terms of lower powerlevels or alternatively the use andpositioning of antennas thatwould be less inclined to concen-trate the r.f. field at eye level.

Handheld units of more thanmilliwatt power have also beenreadily available in the UK forseveral years for use as cordlesstelephones, sometimes offeringranges up to 20 or 30km. Thetighter DTI regulations relatingto the importation, sale and useof unauthorized cordless tele-phones finally came into force onMay, although not without ameasure of confusion broughtabout by the DTI press statementseeming to imply that no units

operating on frequencies below853MHz are legal. In fact, all ofthe authorized units use fre-quencies below 853MHz and re-main legal, operating on the DTIassigned channels at low -powerintended to restrict the range tounder 200 metres, although re-ceivable on sensitive com-munications receivers at greaterdistances. BT have been develop-ing cordless phones using digitalmodulation techniques ataround 900MHz for severalyears, and would be well -placedfor providing such services asapparently now envisaged byBell.

Universalportable radio

The influential Proc. IEEE jour-nal, in its April issue, accorded toportable radio communicationsthe accolade of an invited, book -length (42 pages) paper "Univer-sal digital portable radio com-munications" by Donald C. Coxof Bell Communications Re-search. In doing so, it notes that"providing voice and data com-munication to people away fromtheir wireline telephones has be-come a major comunicationsfrontier. This frontier is beingpenetrated by evolving newapproaches to portable com-munications, e.g. cordless tele-phones, mobile radiotelephoneand radio paging. However,these approaches have manylimitations: none can provideuniversal portable communica-tions services."

The paper examines thestrengths and limitations notonly of these already boomingservices but also radio data sys-tems, automatic call forwardingin electronic switching and evenCitizens' Band radio with itsadvantages of providing simpletwo-way radio but restricted byits "chaotic, undisciplined, unre-liable, simplex mode and noprivacy" nature.

Donald Cox seeks to define theobjectives and possibilities for auniversal portable radio com-munications service which, heconcludes, could be provided byusing demand -assigned digital -radio links for the final thousandfeet or so or telephone loops,using radio frequencies of theorder of 900MHz. He suggeststhat portable units with a trans-

mit power of less than 10 milli -watts would suffice both for out-door and indoor operation. Hebelieves that universaltelephone -linked digital servicescould provide special serviceswith enormous social benefits.

He shows that there are sever-al radio -link techniques thatcould be used to ensure reliableradio performance in difficultenvironments and appeals for asuitable block of the radio spec-trum around 900MHz or 1GHzto be assigned for such a service.

More amateurspectrum

The release of a substantialamount of additional v.h.f. spec-trum to both Class A and Class Bradio amateurs from June 1 andthe reclassification of the 70MHz (extended to 70.0-70.5MHz "secondary status" repre-sents a major gain for Britishamateurs, despite the continuedpower and other limitations onthe use of the 50 MHz band.

50.0 to 51.0MHz becomes a"primary allocation subject tonot causing interference to otheradministrations" (i.e. the Band 1television services in Europeancountries. Antenna height is

still limited to 20 metres aboveground level with horizontalpolarization only and maximumpower of 14dBW e.r.p. (carrier).20dBW e.r.p. peak envelope pow-er on s.s.b. No mobile operationis permitted. 51.0 to 52.0 MHznow becomes available on simi-lar terms, but is accorded a UK"secondary" status.

The "four -metre" band, whichhas never been an internationalallocation, is accorded "secon-dary" status, with maximumpower of 16dBW (carrier) and 22dBW p.e.p. but this is the poweroutput fed to the antenna ele-ments and is not reduced by theuse of high -gain transmittingantennas as is the case on 50MHz. The UK, Gibraltar andEire, however, are about the onlyother countries permitting op-eration in this part of the spec-trum so that most internationaloperation has to be "cross -band".

The new regulations will beespecially welcome to Class Blicensees who until June 1 wereentirely restricted to the use offrequency bands above 144.0MHz. It remains to be seen

whether Sunspot Cycle 22 willreach the level of activity at

which 50 MHz is liable to openregularly for long-distance F -layer ionospheric propagationbut it wil be surprising if thereare not at least some openings.Users of both bands are also wellplaced to take advantage ofmedium -distance propagationby Sporadic E, meteor -scatterand auroral modes.

H. f. resurgentEver since the 1960s, in the yearsthat have seen the apparent un-stoppable growth of satellitecommunications and broadbandcable, more recently in the formof optical fibres that some believewill eventually supplant satel-lites for point-to-point services, Ihave consistently argued thatthere is still a continuing role forh.f. communications, in spite ofthe limited capacity and all theproblems inherent in ionospher-ic propagation. In support onecould quote the recent revival ofinterest on the part of the USNavy with its current major re-quirement for jam -resistant h.f.systems for its ships.

Dr Nicholas Machin in his newbook H. F. Communications -asystems approach (Pitman) addsweight to this argument. Hestresses the re -awakening of in-terest in h.f, attributing thislargely to the growing realiza-tion of the vulnerability of satel-lite communications as a resultof the development of anti -satellite techniques.

He suggest this new lease oflife for h.f. has been furtherenhanced by its marriage tocomputer technology which heconsiders offers tremendouspotential to improve its perform-ance and reliability and has themajor aim of minimizing re-liance on the once essentialhighly skilled communicationsspecialists and operators who, headmits, had considerable under-standing of the transmissionmedium.

While Dr Machin clearly wel-comes this de-skilling of the craftof radio operating and thesubstitution of machine -intelligence, I feel we shouldspare a thought for the phasingout of the once all-importantradio operator.Radio Communications is writ-ten by Pat Hawker.


PHONE P. M. COMPONENTS LTD TELEX0474 60521 SELECTRON HOUSE, SPRINGHEAD ENTERPRISE PARK 966371

4 LINES SPRINGHEAD RD, GRAVESEND, KENT DA11 8HD TOS-PM.1040, 8.95 1085500 1 95 TDA2540 195

INTEGRATED CIRCUITS .70433 5.95 936560C 1.45 TDA.. 1 A 215.40435 7.95 TBA460C0 145 215.70437 7.95 186570 1.00 704753 2 75

45324 2 SO 170 .10439 7.95 113665111 2.50 1062571A 4 SuAN7140 250 1 20 .11467 11.50 T8A7204 245 7082581 295AN740P 280 1 50 ,11463 11.50 11147500 265 1082593 2 95AN612 2 15 1 ¶ 00 '47061AP 3.95 1134800 0 69 1E8%2600 6 50AN7116 I50 51 2 35 '47108P 1.50 934810AS 165 TDA2610 3 50697140 3.50 1 se '477201 1.65 TBA810P 165 70426114 1 95A97145 350FiA521 3 35CA13571 1 75CA3086 0 46

3 001 25

7 95

'07130P 1.50147176AP 2.95*47203 2.95,7204P 2.15

TH487061 0759346700 1.45914890 250934920 I 65

T0A2640TDA-1,854'WA

3 502.15245

083123E 150 0 SO '4720501 1.15 18195020 2.35 295HAI 366W 350 2 IS '4722261 1.80 TBA990 149 Tr " 5 soHA1377 3.50 1 15 '47227P 4.25 TCA270 1 50 250HA1156W ISO 6 75 T*73101 1.80 TC4270S0 1 50 29544473398 295 5 IS 70731347 215 TCA650 3.50 275H4,398 2.75 350 107321P 2.25 TC4940 1.65 1.95101551 295 7 25 :47146P 2.50 TCEP100 250 1.95LA1230 195 1 75 '1176097 3.95 TDA440 150 1.50A4031P 196 1 15 TA7611AP 295 1061001 2.95 2.75

LA4102 295 2 85 14A3104 150 10610024 2 95 0.75LA4140 2 95 7 50 7883206 3.50 10610064 2.50 1.95IA4400 4 15 1 80 TAA350A 3.50 9341035 250 1.25LA4420 195 110 788570 115 7061037 1.95 .118TH 2.95LA4422 2.50 1 10 8866113 115 TDA10744 2.50 PC.1185H 3.951.44430 2.50 395 1414700 1.70 7061170 1.95 3PC1191V 1.50LA446I 195LC7120 375LC7130 3.50

5 3951. 0 89'. I 75

191120AS/3,C--,/vS8111 1.001138395 1.50

TDA1190 2.1510412700 5.501061327 1.70

UPC1350CUPCI 353C

2.952.45

1C7131 5.50 N 130 188396 0.75 TDA2002 1.95 UPCI 365C 3.951.443249 0.45 5.59.1319 295 THA44011 2.55 TDA2003 115 UPC20021-1 1 95

19380914 1.75 105 7084800 1.25 TDA20150 215 555 0361.93837 2.95 . 1,53 iN 1 65 TBASIO 2.50 TDA2005 2.95 556 060LM39009 3.50 .5,44 265 1885100 2.50 TDA2006 115 723 05084515131. 2.30 065111,1 1.15 RA520 1.10 TDA2020 2.95 741 0351.1515151. 2.95 1085200 1.10 TDA2030 2.00 747 0501.4515211. 150 7.95 716530 1.10 7082523 2.95 748 0 351.4113712 2.00 7 95 T8A5300 1.10 TDA2524 1.95 7805 0 651.4C1307P 1.00 195 TBA540 125 TDA2530 1.95 7808 060MC1310P 1.50 7085400 1.35 7082532 1.95 7815 065

SEMICONDUCTORS0350 35

BFT42 03581143 035

R2322R2323

058066

0 350 49

BFW92 08581029 055

112540HCAI6334

2 48090

A6917 0 25 009 040 81084 026 806I6335 0.80ACI25 020 009 040 131085 012 SKES1 1.45ACI26 045 0 10 065 131)(86 020 T1P29 assACI27 020 0 13 0.75 131088 0.25 TIP29C 0.12AC128 028 0 13 0.32 81150 0.21 TIP30C 0.43AC1281 0 32AC 141 028

009009

0.65o st 0.651[14 075

013151 0.2187952 0.2581190 0.77

TIP3101IP320TIP330

0550.420.95AC1421 0 79 009 1,4 1/ 0 75 131. Y48 1.75 TIP34B assAC176 0 22 009 0.65 8113139 0.45 T11414 0.45AC1769 031 009 065 811100 026 TIP4IC 0.45AC187 025 009 095 BRIOI 0.49 T1P47C 0.47

AC1871 020 009 0.45 88103 055 T1P47 assAC188 025 009 1.25 BRC4443 1.15 TIPI20 0.60AC1881 037 009 I 25 13T100A 0 20.85 117125 0.65ACYI7 115 0 12 090 01106 I 49 vIP142 I 75AD149 070 0 12 150 BT116 120 TIP161 29540161 039 0 15 1 )75 035 8E119 315 TIP2955 08540162 039 0 25 Ft, 035 BT120 65 TIP3055 0 5560161 2 090 039 BF I a 0.65 BUIOS 95 11591 02041106 0 SO 030 0.39 BU108 69 TVI062 1.5041114 1 95 100 030 0.20 430124 258308 1.35AFII5 295 101 030 131158 0.22 BU126 60 292219 0.28AF116 295 103 026 81 160 0-27 B0204 55 2812222 029AF117 295AFI21 060AF 124 065

1078 009327 0 10328 0 10

81161 027BF 173 0221817' 038183 '7 026

139205BU208BU7084BU20130

30395285

2142905293053293054793055

0.400.400.590.52AF 125 0 65 137 0 10 fir 15. 034 803776 20 0.1261126 065

61127 0 65138 0091474 013

31173' 029PI 181 0.29

BU407130500

.242.25

5 0.120.1241139 040 461 035 I11 III.' 0.29 1305084 1.95 0.2061150 060 478 020 411714.1 0.29 80576 110 01261178 195 527 020 81704 128 8U800 225 .4,708 012

AF239 042 547 010 111131', 0.28 6096911 1,70 '11713 980A1112 3 75 548 0 10 81194 0.11 MJJ000 1.98 'W., 2.75AS927 1 25 5494 0 10 131195 0.I1 MJE 340 0.40 N 179.. 1.35AUI06 95 50 014 01196 011 WE 520 0.48 N:427 1.95Alin] 95 0 00 81191 011 mPSA13 0.29 1.15601076 11 080 81198 016 MPSA92 0.30 012001070 II 0 10 131199 014 MR1237 415 048BC 108801088

1012

I 600.30

131200 0 40431241 015

841114504 13.958481153 17.50

0.60ass

BC 1 09 10 ,,124P 0.59 B1745 0 30 8401454 26.50 0.9513010913BC1179C00114801166

12

12

IIIS

11131 0421'132 423

033

0.40'135 0.30

81257 028131258 0 2881259 0 28131271 02681273 0 18

8481155 17.50MR1475 2.95PARF477 135004111838 11950C16W 4.50

0.60asoassaso

BC1.7 19 .'736 0.30 01336 034 0023 950 117,0 2.2080119 24 (337 832 81331 029 0025 150 .C1173 1.15BC125 25 r(138 030 81338 032 0026 1.50 5C1306 1.40BC 139 20 13139 032 81355 037 0028 5.50 :SC 1307 2.95BC140 31 ('140 030 BF362 038 0029 4.50 2SC1364 0.50130141 25 i,144 110 81363 015 0032 5.50 2SC1449 0.0013C142 27 131500 029 131371 0.25 0042 1.50 ?SC I678 1.25BC143 24 3 59 065 81394 0.19 0044 1.25 2SC1909 1.25001478 12 360 150 13142., 0.32 0C45 1.00 2601945 3.75BC1484 09 095 8145- 032 0070 IvoBC1480 09 3179 0 72 81458 036 0071 0.75 010BC149BC t57

0912

.3181 04513182 0 70

181459 0.36188.1 0.68

0072 250OC 75 1.50

1 as1.15

BC158 09 11183 0 70 11/ 4:33 045 0081 1.00 1.95BC 159 09 13101 0.63 111 .1, 0.45 0084 1.50 2078 14500171,6 09 1,202 065 0.23 00139 12.50 085BCI 77PC178BC182

Is15

10

0 78O4 0 70

046

13159' 0.258111115 0.231311307 0.25BFRalt 010

0C171 4.5000200 4.5000201 5.5000705 1000

2.951.951801368018218 10 :'723 059 811390 1.50 8200813 1.45 0.50Re Al to . 0.'75 048 811191 1.75 II:, -1, 145 0 95

DIODES13 HO 0 33

r9S. 0.331440111 004744003 0.01 LINE OUTPUT TRANSFORMERS94004 0.0594005 005,98400 0.22

AA, 060 .`99-800 022 144007 0.06 795BA, 013 1,1310 an 1141413 0.02

k 17,:.0.1 .5i 9.9586145 016 139436. 1508 N4448 010 1/431. 4119, 8.9586148 0 II 020 N5401 0 12 DECCA 7730 8.2586154 006 BY 338-60011 95402 0.14 GEC 7040 8.95BA156 015 060 N5403 012 GRUNDIG 1500 15.4588157 030 00 055.600 oxi 95406 0.13 GRUNDIG 5186010 1011 00186013 004 BY/171 600 1.75 5407 0.16

13 45611016 0 12881058 030131151 0 79131126 01089127 o

82095030 0.35C5413 8.00CS 907 16 500,647 0150690 010

95408 0.161144 0.0017923 0.15TT2002 0.20

ITT CVC20ITT CVC30PHIL IPS GBPHIL IPS G9

820825850699

80133 0 15131164 0 45

0691 0150695 0 10 ZENER

PHILIPS G11PST 725

13.991095

81176 I 20131179 063

06202 020192108 500

DIODES R841204TANDBERGE 90*

124011.15

117182 0 55 182313 500 87061 Series ITLEFUNKEN 7114 11.1589184 0 35 19230 500 0.15 TH0119 8000 215081199 040 1842318 5.00 132Y68 Series THORN 9000 9.9589206 °IA 15123WE 100 0.10 1HORN 9800 2210

CATHODE RAY TUBES Please add £3 additional carriage per tube._ME412..1Z: 19 00CME822GH 25.00CME1428GH 45.00CME1428W 39 00CME1523W 39.00CME143113H 39.00CME143IW 39.00CME202GH 45.00CME2024W 45.00CME2325W 45.00CME3218W 45.00CME3132GH 45.00CME3155W 45.00CRE too 25.00CVI450 35.00CV1526 19.00CV21 85 15.00CV219I 19.00CV2193 15.00CV5119 85.00CV5320 85 00CVX389 55.0013110084 39.50010-2100H 45.00010-2100116813 65.00D10 -230tH 35.00Dl 0-230GM 35.00010293-01,90 5500013-300H 49.50D115101226 85.00013-510M 26 85 00013-45084 01 55 00013.610011 59.00D13.511011 59 00D13.611GM 59.00013-630011 59.00014.150084 75.00014.1500M 75.00D14-162011 84 59.00014.1720R 55 00014.1720V 55 001)14-113G84 55.00014.1730M 53.00014.1730/1 55.00014-181011/98 65.00D14 -1810J 55.00D14 -1810M 53.00014-181GM50 59.00D14-182GH 59.00014.200E1E 89.00D 1 a-rovGA/50 85 00D14.200054 75.00014.2100H 75.00014.2700H 50 75.00014.310W 110 00D14.320G11 8. 85.00014-340084 KM 45.00014.3400A 45.00D16 -1000H 65.00016-1000H/65 69.00D16-10001+7 65 00016-1000H/769 69 00016.100084/97 65.00018.1600H 69.00021 -10tH 65.00

DD G76DG7-6 5

3555.0000

DE17 36 55 00D07 32 45.00DGI3 2 45.00011091 55 00DID 91 45.00DP7 5 35.00

'35.00

091378 35.00F16.101GM 75.00F16.101LD 75.00F21.130013 75.00F21.130LC 75.00F31-10GM 75.00F31-10G1F 75.00F31.10LC 75.00F31-1010 7510731-12L0 7500F31 -13G8 75.00F31-13LD 75.00F31.131.0 75.00F41 -123LC 185.00F41.141LG 185.00F41.142LC 18100M7 -120W 19.005017.120084 19.008414.100GM 45.008414-10010 4500M17.151GVR 175.005117.15108 17510M19 -100W 45 005119.103W 5510M23.11008 55.008423.111LD 55.001.'121 I I2GM 55.005123-1120V 55.00M23-112GW 55101.123.11213A 5510M24.120GM 59.005124-12033C 59.001.124- I 20WAR 59.001.124.121084 5500'.128-12G1-1 55.001.128-1310 49.00M28-1310 4910M28.13G8 49.001128-131013 55105128.1330H 55.005431.101GH 5510'A31-18208 55.00'A31 -1820V 53.00..^31-184W 65.00''31.184084 85.001131.184731 65.001131-185W 69.00M31.190GH 55101.1111-190GR 55.001.131.190LA 55001.91.1910V 55.001.131-19100 45.001.131-220W 59.001131-27001 65.001.131271731 65.00,.111 771GW 65.00

6500

M3C1 ...F...,13 65.00M38120W 6100M38.120VVA 65.005138-1210840 6500M38.121LA 65.008438.1220W 65.00543814033A 65.00M38.142LA 6500M38-341731 65.005138-344739 65101440.120W 59.00M43-12LG. 01 68005144.120LC 65005450-120084 65.00M50.12008 65.00M50.1200V 6510M50-120LC 65.00SE3AP31 40.00SE4D P7 45.00SE 42873IAL 55.00SE42BP31 55.00SE5F731 55.001948H 65.00V5004L0 59.00V6048CAL 59.00V6048J 49.00V60648731 55.00V6069GH 55.00V6070731 65.00V7030 59.00V703101-1 59.00V703187A 5900V7035A 49.00V7037GH 45.00V8004GR 65.00V8006GH 65.00V8010A 65.0038P1 11.503071 11.503H OBM 55.003WPI 18.504E71 30.00581171 30.00588471FF 30.005884737 30.005CPI 10.005TOIA 15.006E77 S 39.0012CSP4 451013871 13.501381764 17.501 7DWP4 2500323'1085 69.001273 39.001564 45101844 80.009442E1 750095447GM 75009544901.' 75.00770963' 78.50

SPECIAL OFFER

WIREWOUNDRESISTORS

H1911 kusututiuri / 'InSps 125131-325GH C35.00 each

FTKIS - trans available

4 watt 284'' -

7 wait 1447-22, , 2011 watt 1R -15K 02517 watt 113-150 030

M. 1410/ 75.00M36. 1 7016 75.00M38 101611 65.00

VIDEO SPARES & HEADS

VIDEO HEADS3HSS Suitable for Mosl NC andFerguson rnobels 29,50361SSIHI Suitable for Heath.VT5000 VT8000 076030 VT8500VT7000 33954HSS Suitade to, cnost NationalPanasonic Models 33154HSS0J160 &stab* tor .anasonic14130e1s 370 and 380 33.95

RETAMIX VIDEO READSPS3B1 i P1111 Sultabie is Sony andToshiba 5000 Series and NECP02400 39,50RSV -3.8 Sotable for Sony 518000S18000 SLD7ME 39.50OSR-10-11154414blelo, Sony SL CS5106 SLC7 39.50

Sanyo Head forVTC930091500 49.50Sanyo Head to, VTC5300 5000 49.50

VIDEO BELT KITSAsa' VS9300 95,A, ,01,0 VTC 00313 3.759800 3.75 53nyn VTC 9300 3.7SE31,01,0n 3016 4.50JVCH HR 3330 3600 480JVC HR 3360 3660 4.50Panasonic NV 300 4.00 3 75Panasonic NV 20008 3.75 3 75Panasonic 300043 3.75 5,05 51 11310B 175Panasonic NV7000 3.50Panasonic 90860313 4008610E1008 3.75 4 50

SANYO ORIGINAL VIDEO PARTSSANYO ORIGINALPART NUMBER M0011 DESCRIPTION4.529.108008 VT Reel 140101 3 6W 99548277.23501 VAN' Moto, Assy 9 754.5279.51000 5,55 Capston Moto, 29 95143.0.4904.00900 FVHP6I5 Gear 104, Assy 595143.2.4571.05900 VTC9455 Stoppe, Reel Base 0 55143.0.5457-01701 VARIOUS Pinch Flolle, Assy 8 951438545181700 VARIOUS Pesch none, Assy I 9514385511 VTC9300 Idle, Assy 095143.0,6611%03800 VARIOUS Loadtng Roue, 095143.0.6627,01201 VTC5150 Reel Drive Pulley 850143-0.9974.00100 FVHP615 61:.. 8' If finf, ,s; .1 1 95

Video Head Cleaning Tape (VHS Automatic wet/dry) 6.50Video Head Aerosol Cleaner 0.85Video Copying Lead and Connector Kit Z/V 7 95

VIDEO ALIGNMENT TAPES49 5049 50

E H T MULTIP.IERS VARICAP TUNERS PUSH BUTTON UNITS

i TT CS, . 635iTICV030 6.35PHIL IPS G8' 616HANK 120A 6.91THORN 3000350U 7.57THORN 8500 80013401314 9000 100UNIVERSAL TRIFLER 545

. .. ,

Ecclu43061401.1.4 D 8.65U321 8.250327 8 2511324 1100

7.95:TT 9(... 10 ' W4, 10.19PHIL IPS GI3 ,55173 6 WAy 14 49

200MA QUICK BLOW FUSES

THERMISTORS 8pcach200MA 5 AMP Sp each

REPLACEMENT

ELECTROLYTIC CAPACITORS

VAI,,4, 0 23VAIOS65 0.23VA1104 0VA8683 0

9545

041097 025

20MM ANTI SURGE FUSES

15p each14 -r8511. 12p each

DECCA 30140040035091 2 85DECCA130 1001400,305W 2.99DECCA 17001700,200,400-35091 3 55GEC 21101600100V1 2.25ITT C VC203200,4000/ 1.80PHIL 1750160030051 225PHIL IPS G912200 639 1.19PHILIPSGI I t 470 2505. 2.35

SPARE & AIDS

HEAT SINK COMPOUND 1.00FREEZE IT 0.95SOL DA MOP 0.64SWITCH CL E ANER 0.85W040 1.75

PUSH PULL MAINS SWITCHiDECCA GFC RANK THORNETC I 102POE IF GAIN MODULE 6.99614001 CAP 1274V) 0.69


PHONE0474 60521

4 LINES

P. M. COMPONENTS LTD TELEX

SELECTRON HOUSE SPRINGHEAD ENTERPRISE PARK 966371SPRINGHEAD RD, GRAVESEND, KENT DA11 8HD TOS-PM

A SELECTION FROM OURSTOCK OF BRANDED VALVES

*1714 24.50 EBF15.1 095 El513 6.95

*1834 750 60689 070 61802 165

41998 11.50 66E93 0.95 61821 15042087 11.50 EBL 1 2.50 61822 12.95

42134 14.95 6E121 2.00 61180 2150

A2293 6.50 ECS2 0 75 EMI 9.00

*2426 35.00 EC70 1 75 6144 900

42599 37.50 EC81 7 95 60400 070

*2792 27.50 6C86 100 EMIR 070

42900 11.50 EC88 100 EMS. 165

A3042 24.00 6190 110 EM85 395

A3283 24.00 EC91 550 EM87 25040/4013 400 EC92 1 95 EN32 15.00

ACiTHI 4.00 EC93 150 6891 1.95

ACT22 5915 EC95 700 6892 1.50

*11221 39.00 EC97 1.10 6035 235.00

AH238 3100 EC8010 12.00 6051 0.90

*160 COO ECC32 3.50 6081 2.3S

AN I 14 00 ECC33 3.50 EY83 1.50

ARP12 0.70 61135 3.50 EY84 5.95

ARP34 1.25 ECC81 1.50 508687 0.50

ARP35 200 ECC81 Special EY88 0.55

AZ11 4.50 Onalay 2/5 6091 5.50

AZ31 250 ECC82 0.46 60802 0.70

BL63 200 ECC82 Phelps 6240 2.75

85814 55 00 115 6241 275BT17 25 00 ECC83 085 6280 0.75

CIK 2 50 ECC83 Brom,9 6281 0.75

C3E 22 00 .35 6290 1.50

C3JA 70 00 ECC83 Phnp 6806 2.95

C64 20 00 FWA 800 2.95

C1108 65 00 ECC8.3 S.499a G55. IK 9.00

C1134 32 00 so 0I80 2121 6.95

C11484 115 00 ECC85 75 0240 2D 9.00

C11491 195 00 ECC86 15 GC106 17.50

C11501 13500 ECC88 95 GC1130 17.50

01534 32 00 Eccs, 00 GC 10 48 17.50

C36 2950 ECC 180 72 GC104E 17.50

CC3L 0.90 ECC189 95 GC12 48 17.50

CK1006 150 ECC80IS 95 0086W 600CV Nos Prices ECC1303S 95 GOT120M 5.00

on request ECC804 60 GN4 100034 27 SO ECC2000 95 GNIO 15.00

D63 1.20 ECF80 IS GR1OG 4.00

DA41 22.50 ECF82 15 GSIOC 16.50

DA42 17.50 ECF86 I 70 0510H 12.00

DA90 4.50 ECF200 1 85 0512D 12.00

D4100 125.00 61E202 105 GTIC 14.00

DAF9I 0.70 ECF801 0.85 GTICSS 13.00

DAF96 065 ECF805 2.50 011117514 6.00

DC70 1.75 61E806 10.25 GTRISOW 1.00

DC90 3.50 ECH3 2.50 GU20 35.00

DCX.4.5000 ECHO 100 GU50 17.50

2500 ECH35 150 GXUI 13.50

DETI6 28 50 ECH47 1.50 G0U3 24.00

DE118 26 50 ECH/31 1.00 GXUSOSS 1/50DET22 35 00 ECH83 100 00501 1.50

DE123 35 00 ECH84 1.00 GY802 1.50

DET24 39 00 ECH2000 150 0230 2.50

DET25 22 00 ECL80 0.60 GZ31 2.50

DET29 32 00 ECL82 0.79 0232 2.50

DF9I 1 00 EC183 150 6233 4.50

DF92 060 60.84 0.74 0234 2.50

13F96 1 75 ECL85 ass 0237 45013697 1 25 ECL86 0.95 84491 100

D1163 1 20 ECL805 0.69 HABC80 0.90

DH77 090 66374 2.50 HBC90 0 75

0879 056 6E39 1.50 HBC9 I 0.00

D11149 200 6640 4.50 HF93 0.75

DK9I 1 20 F142 3.50 HF94 1.50

0692 50 FF50 250 11690 1.05

0135 250 7155 495 11121 1.95

D163 1 00 EF70 20 14123130 1043

DL 70 250 6171 150 8141 1500173 150 EF72 20 814200 1500192 125 6E73 100 8190 0.70

D193 t 10 EF80 0.55 8192 1 50

0194 250 6683 3.95 01.13130 3500(96 250 EF85 0.50 1112 400

DC S10 13.50 EF86 2.25 11090 1 00

04.S16 moo EF86 Mul180 HVR2 30000470 150 4.50 13118 65.00

014160 4.50 6689 150 KR6 3 4500DY51 1.50 EF91 1.96 KT8C 7.00

DY8687 0.75 EF92 2.15 KT3X 3.50

00802 005 EF93 0.95 1136 2006551 4950 FF94 0.96 1144 4006801 29 50 6195 195 1(145 4.00

6811 12 00 F197 090 1161 500E82CC 4 50 F198 090 6163 2.00

6831C 450 F1183 065 6166 USA 995EIL3F 5 50 EF184 065 616.6 GEC 17.50686C 950 66730 1 80 KT66 Sp Yellow

038C 7 95 F6731 350 Spy, 19.50

688CC 350 66732 3 50 6167 9.00

E88CC.On 66800 11 00 K177000 Lion%%tiara at 695 668045 19 50 1095688/1 S.ernans EFEIOSS 19 50 KT8I 7.00

SPeoai 5.95 EF806S 19 SO 1198 USA 10.95

E9OCC 7.95 66812 065 11788011on

690F 7.95 661200 150 1195

69111 4.50 61190 0 72 KTW61 2.50

E92CC 3.95 E190 095 KTW62 2.50

E99F 6.99 6132 095 KTW63 20061301 16.50 6133 500 61263 2506182CC 9.00 6134 2S0 1102 26 6.95

6180r 650 6134 Mu8367 1102,26 12.00

611361 850 Phos 4.50 107-20 95.00

61880. 750 El. 36 1.95 LS9111 6156280F 19 50 6136 1.531490 MS024 600062113/1 12 00 2.50 045374 60.00

E288CC 17 50 EL37 9.00 045143 155.00

E8106 29 50 E138 6.50 148079 10081148 I 00 6141 3.50 1480132 7.50

EASO I 00 E142 too M8083 1256452 35 00 6181 6.95 M8091 7.50

6476 1 95 6183 7.50 048096 3006479 1 95 6184 Brand/ 0.03 M80913 550EABC80 0 70 61.841.40.10 29S 148099 500EAC91 750 6185 450 148100 550EAF42 I 20 6186 1 75 M8136 70064E801 200 6190 1 75 148137 7956834 1 50 E191 600 808161 6.50

6641 395 6195 1 75 148162 550EB9i 005 61153 12.15 048163 550EB/33 2 50 EL 1836 350 148190 4.50

EB/4 I 1 95 6118P 350 ME119S 1506E1181 150 E1360 675 048196 550EBC90 090 FI 500 195 118204 5.5068C91 090 71504 1 95 148223 4506136130 095 Fl 509 525 148224 2.00

3.55ME 1401 29.50ME 140.i 29.50ME1501 14.001.1114 3501116 400

4501.4P25 195.00MS4B 5.5004014 150M21.100 125.00N37 1250N78 9.85042 1.190A2WA 2.50043 250087 1.15OB2WA 2.500C2 1500C3 1.50003 1.70ON44 1.0001/1513 1000046 1.79ORP43 2.50011650 3.95P61 2.50P41 2.50PABC80 0.75PC86 oisPC88 0.75PC92 3.50PC97 1.10PC800 1.10PC900 1.25PCC04 esoPGG85 0.55PCC88 0.70PCC89 0.70PCC189 0.70PCC805 0.70PCC806 0.80PCE82 asoPCF80 essPCF82 aeoPCF84 0.65PCF136 1.20PCF87 0.40PCF200 110PCF201 110PCF800 0.40PCF801 1.35PCF802 085PCF805 1.25PCF806 1.00PCF008 1.25PCH200 1.50PC1.82 0/5PC1.83 250PCL81 075PCL85 010PCL86 085PCL800 0.90PCL805 090PE1.100 69.00PEN25 2.00P6644000 2.50PEN45 3.00PEN451/0 100PE N46 200PE06.4oN 42.00P61200 aosP121 250P136 1.75P138 1.50P181 0.72P1814 0.72PI. 82 060P183 052P184 078P108 100P195 2.95P1302 1.0061345 12.50P1500 1.10P1504 115P4 S08 1 7561509 485Pi 519 4.95P18021 3.50PLE120 295PL5557 29.50P032 0.60P033 050P081 070P082 070P083 0.70P088 015570500A 196P0800 079P0801 0.79013.3300 72.0001130750 139.500853500 595.000E03-10 4.950E08.200 145.000140 65.00OP25 11000E02.5 19.5000E03-12 7.9500E0320 35.0000E0140 450000V0?6 195000003-10 55000003-10Moilard 1500001103.20 25.0000006.404

27.5000006.404Muliaid 39.5000007-50 55.0000203-20 42.500575 20 1500575 40 3000592 10 5.000595 10 166OS 10645 COO

0515015 6.950515030 1.15OS15045 7400S1200 195OSI 202 3.950S1203 4 15051705 3.95OS1206 1.05OS1707 0.90

001208 090OS1209 3.15051210 150OS1211 150051212 120OSI213 5.00001215 2.10OSI218 5.000037 9.500003.12 5790005-25 2500006.20 29.500008100 14500002-2500 4500003.125 6500004250 70.00004.400 76.00R10 4.00R16 12.00

4417 1.50R113 250R19 250R20 1.2091169 55.00501-125 4.95501.2404 14.50RG3-250A 3.5056112504 35905121(25 62.5051.208 12.005116 15051,116 12 00RP013 25056043 2.50RPY82 2.50551250 1500553.1250 35.00RS613 45.0055685 54.95RSS688 52 1556617 59556633 2195SIIE12 31.00A3026 12.00S1041K 10000109 1K 15005130 5.59S130P 595SC1 800 100SC 1 1100 6.00SC1 1200 500SC1 1300 6.00SC, 200 900SP41 500SU42 495182 5 300 85.00182.300 45001837000 39500101.1004 2500T003.101 35 00TD3.I2 4.00TDD4 590TP25 1 50TSP4 700TT, I ISO1721 45.00T122 45 00TT100 69.00TTI1 31940 65.00102.1254 6500104 400 6500108.600W

365.00TYS2 250 375.00U1020 775U19 11 95U74 200U25 090U26 090U37 9.00U41 6.95050 ZOOU82 100U191 0700192 1000193 0 650251 1 000801 0.751148080 0.65UAF42 100UBF80 aoo()BC, 2.25URCA1 1.50u13189 I 00UM., 175UC9.' 1.20UCC6.1 0.70UCC85 0.60UC1130 100UCH.'1 1.20UCH, 250001142 2.50UCHP, 100110181 17518:183 2.5011141 1 ISU147 1 151/180 1 7511185 120uF89 20011141 5.50UL 44 150U184 150UL85 0.85UU5 350UU7 800ULM 90011041 350U085 0.70V235A 1K 250.00V238A 1K 295 008246471 315.000240616 225.00V241C II 195.000339 3500453 1200010631 10.950P48 4.50VP133 200VR75.30 3.00VR101 2.00V10530 1.500015030 1.1S2152 2.50VU29 4.50VU39 1 50W21 4.50

' 500W770 I 00W739 1.50024 4.50066 X65 195X761.4 1150C24 1.500125 0.50XFW47 1.50XFW50 1.50501-2500 75.00X05-500 22.500162861 7 50XNPI2 1500141.1600A 4950X91.32004 79.50X111.64304

149 500502 25 00Y65 695601101. 75000J1060 265.00611021 42.50611060 195.00011070 195.00611071 195.00011296 65.00Z77 1.202302C 12.002359 9902505S 1500252014 100752180 80077000 1002749 9607759 19952803U 1595ZA1001 12.50741001 1 502M1005 5002M1020 6.00ZA11021 800791023 7.957041041 14.0021411012 9002M1054 10.002M1177 900/1.112C2 55.00IM12133 4.00

1822 10 001827 55 00113354 45 0018.614 75.00ICI 2.50IC5GT 2501601 250I G3GT 25013307 250163 /50042 4.5018501 2S01P28 25 001639 19 50IRS 0.90114 1.00104 1.75105 10011(2B 140122 1952ASI i4 11.50287 1.502822 69.502C39A 32.50213984 39.502C40 37 002042 29.502C43 60 002151 0 752153 45002105 1.507021 1.952D2IW 1152E2? 49002E26 7.952342 93.002355 295102625 350021(25 Raytheon

750021(26 95 002625 250.002648 140.002/(56 250002020 5.0034.10713 12003A 1084 9.003A 13913 11.003.1111061 12.003411116 11.5034.1373 7.503A15704 moo343A 195344 1.10345 4.503115 0953417 3 353132 300383 495387 450382a 10.033826 24003829 15003826 1 50304 1003C23 19 003C4- 24.003CB6 ISO3CN384 2.503CS6 0953003 2503095 150306 45030214 29.503E22 49 503E197 1953EJ7 1.95304 250304 1 753W (GT 25048`518 115 0049sA 65 004.2s04 79.50.1400A 97 504.10004 425.00

.1832 35 00413074 1754026 1 954C27 25 004C28 25 064C35 145 004C X10004

425 004C)(40004

1100 004C X2506 49 004C075081WA.

59 504032 125 004C)(250614 75.064CX2501( CiPAAc

95.004CX125C EIMAC

125 004103504 95 004CX350F 79 504XCI50013

398 504E27 195 004057 2 254352 75 004JC64 2 9541(16 t 50401504 35 00SA 1020 950541521.1 9005A1631 10 005A170K 6.2554-180M 90054-2068 10 0054M8 21554N8 I 205454 200SAUS 1505811004 10 00513.2541./ 14 5056-25514 19 5058.25614 15 005835714 15 00513.2581.1 14 505C22 125 00501.84 2505.11806 2950 0059408 55051140Y 4 95514 5 955U4G 2955U4013 4 505040 I 5050301 19552407 I 5063012 0 706A 203K 900647 4 95648G 1 SO

64C7 20064G5 15064G7 1956486 25064.14 2006417 20064K5 1 95

6416 25064115 06064444 3 256/0.15 60064516 19564N5 4506AN8A 3506405 1 756408 0 8564115 S 9564S5 1 506456 25064070 4506476 0 7564T8 I 7564114 2006AU5GT 45064176 0 956496 0 1564W84 25064X407 195640313 1 9564213 4 50687 280688G I506810 1956846 0 956841 45068484 15068E6 09568060 3006486 I 9568H8 1506816 150681(4 40068676 1950816 85 006818 115681.46 115.0068808 0586E154 165613N6 1 6568197 45C6858 3 956605 0 7568074 0.726817014 34568145 0706897 4956898 2.15613984 2.196E157 55068W4 1.5068W6 5.3568W7 ISO68W6 COO6806 048681001 3506876 2.506877 2.956C4 1256C5 1.956C6 2.506080 1.506C11 2506C18 2536C44 4.956C47 1506C85 3.956/66 1.15

6106GA6CF61'67

H6

84.3.,

51'

4501 507256 953 953 25

501 602 950 750 95

.(.1 6 502 502 35095150695

,B 25050

.54 2153 954 952 50

011 2501 752502501 951 752 957954502004 952 502 005 501 50300

002 752 500601 25I 25I 251 25

17 001952 95550395

.961 0801 50I 95

516 2 652 152 502502 503 959 502 501 951 955 502 504 952153152502004 15

4 4506 502 506 504152 003 006 502 50

*^ 2 95250

6115 3156719 395

.C.0 295,tC1Crf 575

6 50,1 195

I IS ,15 750

6GAV 550SYLVANIA6138 2506106 65060701 1706970 3156S44 I SO6SA70T 1356SC 7 1506507 2506587 I 35653707 1206567 1 35656701 1 35651701 0856SN7GT 1356507 I 356SS7 I 956U401 1 7561J6WA 3506U8 115608A 1506V6G 1256V6GT I50

003 951 001 50100

002 254 50200I .'5

1 SO

3 502 503 SO

2 502 503 50550

'54 2 50888 2 508810 2508805 1 95

8607 1 951002 12510067 25010EW7 2951061 0 751001(6 1 9510P14 25010618 0 76101012 065III' NI 45 00u 1 55 00.11) 55046 395406 150'10761 495

10031'6 1 25

115A i'15./. 250AU6 1500117 065

1 95A07 2 5037 085.1 X.5 795

2 303 951 954 501 50

1,66 1 95bH7A 2 95

6 1 751)74 275

25065 19556 20

.9361 350NI4/: 350

19 50)4 36 00917 3952.7 450

.6".74 450-.GT 3 95'GT 350

2951 951 501 951 954 75

9595so958595502020205095

14 00509550

1.11,1 959550

. 1 00.635 50

405 501,0461 50.0 00.3 1 00

04 3 00.1.19 2 00

3 5000

I 5070

16 959555609515

4 951.8 3 75

39 50064 2.95h(31 1 75

'8)6 1.75

6001,60606X2N6 - 46 - 5GT6 - 5GTv6.88746747*407'AU7'1167677C5706-67

AUDIO TAPE MEADS

MONO HEADAUTO REVERSESTEREO HEAD

250195150

VALVE ANDCRT BASES

84 3 95 31.101. 35 00850 5.50 01(410 25 00B7G 925 SK6I0 35 00B7G SKTID 0.25 UX5 I 75

138G 1.50 1167 1 78811 070 VALVE C3NS6194 0.35 B7 and BS 0.35B9A SAIL 060 ANODE COPS890 0 75 0.508108 020 00V066138 0.50 SOCKET 1000814A 100 813 SCICKFIOCTAL 035 II SO

19506 500 40

1 48095I 001350 95

1 10125045060060

001 00

.'18 060"19 1 00PL 1 2501'111 0.60('114 1 75' .IS6C 550415880 19 5035 45016GT 200

1 85381167 4 5040106 5 5042 6 9547 6005045 1 SO

5005 09550006G 11550E145 15050306 29552111 20053CG 15 00615PT 4507581 3507511 25080 45083 85083AI 15084 3008541 6508542 25090AV 15 0090CI 3509000 13 5090CV 12 5091AG 90092AG 19 5092AV 15 009541 650

00E1 10 0008C1 1505082 65050C2 2505014 2 155SUG 25 008587 I50

211 33 502744 15 00307 5.003284 15 003884 17 5042545 800431U 4505720 55 00'05A 800'084 800"7.14, 2,7500

75 00'591* 650'075S 6 95803 14 95805 59 00807 250810 85 008114 15 00812A 35 00813 27 50813 Ph, 35 008296 14 508334 95 008664 650872* 20 00873 60 00884 550930 9 959314 13 95954 1.00955 1.009584 1 0012994 0601619 2501625 3001676 3002050W 6952050 5502051 5503534 400

109515 00

1977 25 0046874 950.1212E 25000.13I30 40043280 9005636 5.505647 9505651 2505654 1.955663 1.955670 3.255672 4505675 24 005678 7505687 4505692 3505696 4.505'04 1505718 6.155725 25057215 2505727 2505749 2505750 1 855751 2.955763 5.7558146 3 255873 9505829WA 6.505840 1505842 11 005847 10 955879 5.0058813 11955894 39.505899 4505963 1755065 2256005 1.856012 11006021 1656057 1756058 1956059 3.756060 2.251,062 4.506063 2006064 3 256067 7006072 4206080 95018080WA 9.505132 10.506136 25061466 9506155 72 006156 72006157 2506158 3206201 6 456205 6 956211 2 SO6267 4506350 3506360 4506386 14 506463 7506545 9506550* GE 10 956550 RCA 15 00

6870 11 506887 95068838 9.95697'3 5957025 25070274 6507032 2007059 2 SO'167 395'189 3507193 7.507199 7507747 4957360 13 507462 15 007475 5007486 155 00"527 85 007551 8507558 9 457586 15 007587 29.507609 47 007733 550"7E93 29.5071325 7295'868 5.955012 15008950 10 5010042 10.5018045 moo18046 1150

OVER 4 MILLION ALVES IN STOCK4.000 - Doll 1001 Types

P'"w I9PeS not 1151 Id

>1004114 selected 110 matched vas...available n request

CALLERS WELCOMEOPEN MON-THUR 9AM-5 30PM

FRI 9AM-5 OOPM'24 -HOUR ANSWERPHONE

SERVICE'ACCESS & BARCLAYCARDPHONE ORDERS WELCOME

UK ORDERS P&P ElPLEASE ADD 15°. VAT

EXPORT ORDERS WELCOMECARRIAGE AT COSTPLEASE SEND YOUR

ENQUIRIES FOR SPECIALQUOTATIONS FOR LARGE

REQUIREMENTS4



861

Siemens UKdevelopments

Woodley, near Reading is the sitefor the new Systems Develop-ment Group of Siemens. Its aimis to provided system planning ofmedical facilities, factory auto-mation management, and theintegration of components intotelecommunication and in-formation systems. One of thegroup's initial projects if theinterfacing of Sinix, Siemens'version of Unix, with their hard-ware running under MS-DOS.providing a 'transparent' link be-tween the two operation sys-tems. The linking of Sinix to ahigh-speed graphics terminal isalso high on their list of priori-ties. Another project is to de-velop an emulator that allows aSiemens PCD-2, or any IBM-PCclone, to function as a high-speed terminal connected to acentral processor via Ethernet.an RS232 interface. V11 inter-face or a simple modem. TheGroup is also currently develop-ing a symbolic debugger for the'C' language, with a windows -oriented user -interface to im-prove usability.

In addition to developingSinix-based systems, the groupalso carries out systems pro-gramming for the Siemens 7,500series mainframe computers,which use the BS2000 operatingsystem. Here. particular areas ofapplication are job control lan-guage and compiler develop-ment.

The group. indeed Siemens asa whole, is firmly committed toconforming to, and helping todevelop, standards in operatingsystems, communications and inprocessor architecture.

Ban on manycordless phonesSince the 21st May, all cordlesstelephones that do not conformto the DTI conformance speci-fication MPT1322 and operate onfrequencies below 853MHz mustnot be sold. It is illegal to use,import, or manufacture suchphones. By outlawing certaincordless telephones the DTI aimsto remove a considerable sourceof interference to a wide range oflegitimate radio users. includingemergency services. Because of

[-UPDATEpoor technical standards and theuse of incorrect frequencies,their higher output power candisrupt other radio users.

Other measures to amend thetechnical definition of cordlesstelephone apparatus have alsocome into force. Copies of theStatutory Instruments relatingto this new legislation (SI Num-bers 774/87 and 775/87) areavailable from HMSO. Copies ofthe DTI performance specifica-tion MPT 1322 are availablefrom: The Library. Radiocom-munications Division, DTI,Waterloo Bridge House, Water-loo Road, London SE I 8UA.

Touch -sensitiverobot

A joint project team from BritishAerospace and HatfieldPolytechnic are developing arobot with a sensory feedback,i.e. a touch system. One systembuilt at Hatfield uses a Fanuc Alrobot with an Astek six -axis forcesensor fitted to its wrist joint onwhich is mounted a grippingdevice. This system is controlledby a BBC -B computer. A closedloop control system responds tothe data fed back from the sen-sor. The robot is programmed toinsert a peg into a hole. As thepeg is not in perfect alignmentwith the hole, the magnitude ofthe force exerted on one side ofthe peg is resolved into x. y and zcomponents by the Astek sensorand is transmitted to the compu-ter. A control algorithm is usedby the computer to process thedata and provide responsivemovement commands so thatthe contact forces are reduced. Itis an incremental process. thepeg being moved in small stepsuntil it is freely inserted.

Such a system has the poten-tial of providing precision assem-bly with improved quality con-trol; a detailed record of assem-bly operations would be gener-ated for subsequent analysis inthe event of problems. The pro-ject is part of a cooperative agree-ment between HatfieldPolytechnic and BAe for a pro-gramme of joint endeavourcovering research, education.training and the exchange ofstaff. Another initiative was theestablishment of a British Aeros-pace Research Fellowship inv.I.s.i. circuit design.

Human factorsin electronics

designThe operator actually has to useequipment and, if it is not easy, asystem is likely to be rejected ornot used to its full potential. Thisis the message of a pilot schemeaimed at improving 'human fac-tors' design in electronics -basedequipment and software whichhas been launched by the DesignCouncil and the Alvey director-ate.

This new initiative recognisesthat the success of the rapidlygrowing variety of products thatrely on electronics depends in-creasingly on the way that theymeet the real needs of the userrather than simply relying onsuperior technical performance.Human factors design is becom-ing increasingly important nowthat most such equipment isbeing used by people who are notcomputer specialists and that anever wider range of products,from teaching aids and videorecorders to machine and pro-cess control systems, areelectronics -based.

'Support for Design - HumanFactors' is available to com-panies with up to 500 employeesin one operating unit (whichallows subsidiaries of largercompanies to participate) andwill provide up to fifteen days ofdesign consultancy with two-thirds of the cost being paid forby the Department of Trade andIndustry. Companies that havealready benefited from a 'Sup-port for Design' project can applyfor a human factors design pro-ject.

The pilot scheme is intendedto use skills that are normallybeyond the scope of conventionalergonomics and industrial de-sign and the scheme will drawparticularly on the expertise ofthe three human factors centresfunded by the Alvey Directorateand a small number of indepe-dent consultants.

Within the scheme, consult-ants will be able to advise on thedesign process in general or con-centrate on specific aspects. Pro-jects are likely to range fromanalysis of requirements, spe-cifying and designing the userinterface, building and testingmock-ups and prototypes, de-signing training and operating

EXHIBITIONS& CONFERENCES

28 August -6 SeptemberFunkausstellung: Internationalaudio and video fair Berlin (in-corporating MediaForum). In-ternational Congress Centre,Berlin. Details from AMK Berlin,Postfach 19 17 40, Messedamm22. D-1000, Berlin 19, FRG.

2 -4 SeptemberMediaForum Berlin '87 Con-gress and exhibition on telecom-munications, including broad-casting. Details as above.

8 - 10 SeptemberDigital signal processing; a

short course at the Institute forInformation Technology. Uni-versity of Sheffield. Details fromMrs C. Scown. Tel: Sheffield768555 Ext. 5100.

9 -11 SeptemberElectrostatics summer school.University College of N. Wales,Bangor. Tel: 0248 351151 Ext.2749.

15 - 18 SeptemberDesign Engineering Show andconference, NEC, Birmingham.Cahners Exhibitions. Tel: 01 8915051.EED 87, electronics in en-gineering design. NEC, Birming-ham. Cahners as above.Test and transducer; interna-tional conference and exhibition.Wembley Conference Centre,London. Trident International.Tel: 0822 4671.

22 - 23 SeptemberComputer Networks; shortcourse at the IIT, Sheffield. See8 September for details.

23 - 27 SeptemberPCW 87: 10th Personal Compu-ter World Show, Olympia,London.

29 September - 1 OctoberNAV 87. Navigation data, dis-semination and display confer-ence and exhibition. HeathrowPenta Hotel. Organized by theRoyal Institute of Navigation.Tel: 01-589 5021.Semiconductor International;design, assembly, test. materialsand chemicals. NEC Birming-ham. Cahners Exhibitions. Tel:01-891 5051.

5 -8 OctoberHDTV 87: International collo-quium, Ottawa, Ontario, Cana-da. Details from HDTV Colo-quium, Journal Tower North,300 Slater Street, Ottawa, Ontar-io KIA OCB, Canada.


procedures, and evaluating theeffectiveness of equipment orsoftware. For further informa-tion about 'Support for Design -Human Factors' contact: AndrewHartley, Support for Design, TheDesign Council, 28 Haymarket,London SWIY 45U Tel: 01-8398000.

* * *

Human factors at a morepsychological level are the sub-ject of other Alvey research pro-jects. A 'programmable usermodels' project will evaluateseveral psychological approachesto human/computer interactionsby assessing how well eachapproach addresses a sample ofproblems experienced by compu-ter users. From this study, it ishoped to improve understandingof the desired characteristics of acomputer language that is easyto understand.

A further phase will involveprogramming and evaluatingprototypes of the language.These will have built-in con-straints corresponding to the hu-man limitations of understand-ing. It is intended that eventualusers of the facilities will in-corporate similar constraints atan early stage in the design ofnew software.

This research is being coor-dinated by Logica Ltd in col-laboration with the AppliedPsychology Unit of the MedicalResearch Council and STL Tech-nology Ltd. Another new projectwhich Logica is coordinating,with the collaboration of com-mercial and academic researchfacilities, is to develop an Intelli-gent Computer Aided Instruc-tion (ICAI) system and know-ledge base for teaching the op-eration, maintenance and pro-gramming of a computer -controlled machine tool. Theproject, known as the KnowledgeBased Engineering Trainer(KBET), is based on TUTOR,Logica's intelligent tutoring sys-tem, produced in collaborationwith the Royal Signals and RadarEstablishment. Using principlesof knowledge -based technology.the system provides a next -generation improvement incomputer assisted training.TUTOR has been demonstratedsuccessfully as a Highway Codetraining aid, but has been de-signed to help teach a wide rangeof subjects.

rUPDATEOne of the most important

features of the KBET project willbe the integration of video -discequipment into the ICAI system.By this means the inclusion ofvideo in the training materialwill provide improved learningopportunities.

Emergencynetwork

An alarm communication sys-tem is under test in Sweden.After only a short period of use ithas demonstrated that the timebetween alarm and action can beconsiderably reduced. The sys-tem, called Coordocom, inte-grates and coordinates informa-tion for alarm centre operators.It does this by providing onesingle channel for all emergencycommunications and concen-trating them into a single audioand visual channel. The operatorhas control at all times over thecommunications using radio,telephone, teletex and facsimileequipment to control field per-sonnel while keeping in contactwith those who need help. Adisplay screen also shows whichcalls are waiting and which ser-vices are available. The systemcan extract addresses from theSwedish telecoms subscriberlist.

The system is built from threemain parts; a communicationsexchange, computers and theoperators work stations. TheNorwegian Nerion exchange wasoriginally designed for air trafficcontrol. Three networked Vaxcomputers control the links be-tween the exchange and work-stations and provide informationstorage, as well as back-up foreach other. The work can bedistributed to as many work-

stations as are needed to copewith the emergency. Preparedprocedures can be displayed onthe screen for the operator tofollow. One example is an airportemergency where 40 pre -planned procedures are alreadyprogrammed. There are also alarge number of general-purposeemergency plans, e.g. for housefires. The system includes anumber of databases including,for example, information onpoisons. A resources modulecontains information on all theemergency facilities, includingvehicles and equipment. Theoperator's screen is divided intotwo sections with one part show-ing communications equipmentlogging, such as which calls arewaiting and for how long. Anoperator can deal with up to fivedifferent events at the same time.The rest of the screen is theoperator's notepad and form -filling area. Messages to/fromother operators can be written toand read from this area. Such asystem is applicable to a widerange of traffic control opera-tions, such as taxi and haulagecompanies or other authorities.

User's guideto IT

A useful guide to implementingIT projects in a business environ-ment is provided by a report:"Information Technology - Thefuture user environment" pro-duced by the Alvey IT users'panel. It expresses a user's viewof where IT is going over the nextfive years.

The panel had been formed togive user guidance to the Alveyresearch programme. Beforeconsidering user needs in soft-ware engineering, it became

All bids made at a Christie's London auction are recorded on tapeby this 22 -channel voice logging system from Philips whichrecords the incoming bids from 17 telephone channels as well asthe auction room bids and the auctioneer's voice.

apparent that a preliminarystudy was needed to formulate astrategy for harnessing IT tobusiness needs. This report is theresult. It is divided into threesections; an analysis of trends, adescription of the kinds of pro-ducts and services which areexpected to be the buildingblocks for future systems and adiscussion of the ways of definingand implementing systemswhich will meet users' needs.The report concludes withappendices which discuss thetechnical strategy for software inmore detail.

Copies of the report are avail-able from the Institution of Elec-trical Engineers.

Faraday for allThe IEE Faraday lectures havetraditionally toured the countryand been given at theatres orlecture halls to sixth -formpupils. The most recent lecture,presented by ICL on the techno-logy and future of computer sci-ence, has been produced in avideo form and a copy of videotape has been sent to everysecondary school. This has in-creased the availability of thelectures as limited space or ex-cessive travel prevented manyschools from attending the lec-tures. This video production wasfunded by the DTI and the Man-power Services Commission. Acompetiton related to the themeof the lecture has been designedto heighten awareness of IT andits appication in modern indus-try. Called "The Faraday 2001Time Capsule Competition" itinvites teams of 16-18 year oldcompetitors from different disci-plines to look into the future andenvisage the roll of IT in the year2001. The winning entry will besealed into a capsule and placedin the London IEE headquarters.For the purpose of judging, en-tries must be submited beforethe end of the calendar year.They will be divided between tengeographical regions which willbe judged in 1988 with cashprizes to the top five teams ineach region. These regional win-ners will qualify for a nationalfinal to be held in April 1988.Further information can beobtained from the competitioncoordinators. Hobson Pub-lishing plc. Bateman Street.Cambridge CB2 1LZ.


RAEDEK ELECTRONICS102 PRIORY ROAD. SCRIBERS LANE. HALL GREEN. BIRMINGHAM B28 OTB. ENGLAND

Telex No 333500 CROSAL GTel 021-474 6000

(4.

I

ElectronAMPEREX Tubes AMPERITE

AleCETRON CTC Du MONT EIMAC

ELECTRONS GENERAL ELECTRIC GENALEXITT LITTON MACHLETT MULLARD

NATIONAL OMNI WAVE R C A RAYTHEON STC

AdrSIEMENS SYLVANIA TELEFUNKEN THERMOSEN

VIC TOREEN VARIAN TUNGSOL BURROUGHS WESTINGHOUSE

11118600

R. F. Power TransistorsdrMOTOROLA

AEROTRON QUINTRON REGENCY WILSON T.R.W.MSCACRIAN TOSHIBA NEC MITSUBISHI PHILIPS AMPEREX

GENERAL ELECTRIC R.C.A. JOHNSON

R/F CERAMIC CAPACITORS


DM 105 - Pocket DVM fromA meter to literal l suit allpockets - including the onethat holds the wallet! itIts small size offers easy one hand operation, with large, clear.wide-angle display.Rut small does not mean delicate; the DM105 can withstand a 12inch drop without damage, due to its glass epoxy PC boards,shock mounted display and through thermoplastic case.Overloads are handled with similar certainty; the resistancerange is protected up to 5()0 dc, AC voltage range to 800V rmsand DC voltage ranges up to 1000V.Battery life is typically 2000 hours and a LO BATT indicatorsignals 20% battery life remaining.RangesI)C Volts: 2V, 20V, 200V, 1kV\C Volts: 200V, 750VFreq Range: 45Hz - 450HzDC Current: 2 mA, 20 mA,

200 inA. 2AResistance: 2K, 20K, 200K, 2MI)C Accuracy:

± (0.5% reading + 2 digits)size: 130 x 75 x 28mmWeight: 195g\ccessories: Battery, test leadsand manual.

Price £24.75 inc VATPlease add 70p for post and packing.

This is just one of our range of quality Analogue and Digitalmultimeters.

Full details and specifications from:

Cirkit Distribution LtdPark lane. Broxbourne, Herts EN10 7NQTelephone (0992) 444111 Telex: 22478 VISA

TRADE ENQUIRIES WELCOMEAIL

RACKMOUNT CASES

--eiwymtssussmositil"'

tg" Seit-Assemoly Rack Mounting Case wal i It oft Covers Front Panel 10 guage, BrushedAnodised Aluminium, Case 18 gauge, Plated Steel with Removeable Rear 8 Side Panels In1U & 2U Types, a Subri s is Mounter Bottom Cover In 3U Type the Subplate is,ocated on two Rails Mr The So. ..s

U (13/4) height, 230m depth £28.302U (31/2) height, 308m depth £33.603U (51/4) height, 230m depth £41.00

Width Behind Front Panel 437m (All Types).All prices Postage 8. VAT Cheques, P. Orders Payable to

J. D. R. Sheetmetal, 131 Grenfell Road,Maidenhead. Berks SL6 1EX. Maidenhead 29450.


HIGH QUALITY - LOW PRICES

4+11GHz SATELLITE TV RECEIVING EQUIPMENTRECEIVERS. LNITs, LNC's, FEED HORNS.

ANTENNAS, ANTENNA POSITIONERS,POLOROTORS, LINE AMPLIFIERS. ETC.

I or further is

HARRISON ELECTRONICSCentury Way. March. Cambs PE 15 8C)W Tel 10254151289



APPOINTMENTSAdvertisementsaccepted up to

12 noon August 3for September issue

DISPLAYED APPOINTMENTS VACANT: £25 per single col. centimetre (min. 3cm).

LINE advertisements (run on): £5.50 per line, minimum £45 (prepayable).BOX NUMBERS: £12 extra. (Replies should be addressed to the Box Number in theadvertisement, c o Quadrant House, The Quadrant. Sutton, Surrey SM2 5AS).PHONE: SUSAN PLAITS, 01-661 3033 (DIRECT LINE)

Cheques and Postal Orders payable to REED BUSINESS PUBLISHING and crossed.

BLOOMSBURY HEALTH AUTHORITYDepartment of Audiological PhysicsApplications are invited for two posts based at the Royal NationalThroat. Nose & Ear Hospital. Grays Inn Road. London WC1.

SUPERVISING MEDICAL PHYSICS TECHNICIAN IIDuties include supervision of technical staff engaged in themaintenance, repair and calibration of audiometric and audiovisualequipment and other activities related to the field of AudiologicalPhysics.The successful candidate will be an MPT II or will have served aminimum of two years' as a Technician III.A knowledge of acoustics or audiometric equipment would be anadvantage but not essential.Salary: £9.758-£11.875 pa inclusive

AUDIOVISUAL AIDS TECHNICIAN - MEDICAL PHYSICS

TECHNICIAN IVThe post carries responsibility for the provision of audiovisualsupport and the production of audiovisual material for both theHospital and the associated Institute of Laryngology and Otology,University of London.The successful candidate will have the appropriate qualificationsand experience - details on application.Salary: £7,193-19,078 pa inclusive according to age andexperience.Informal enquiries will be welcomed by Mr G Frost, PrincipalPhysicist on 01-837 8855, ext 4112.Application form and job description available from the PersonnelDepartment on 01-837 8855, ext 4131.Closing date for receipt of applications: 31st July 1987.WORKING TOWARDS EQUAL OPPORTUNITIES

nncela iLroHnOduosnei\171471silrighby

La,AppoiNTmENTs

Telephone 01-808 3050

THE UK's No. 1 ELECTRONICS AGENCYELECTRONICS ENGINEERS if you are looking for a

job in DESIGN, FIELD SERVICE, TECHNICAL SALESor SOFTWARE ENGINEERING.

Telephone NOW for one of our FREE Jobs lists orsend a full cv to the address below.

Vacancies throughout the UK to £18000 pa.

Co i i i i i i i i NoCapital Appointments Ltd.. FREEPOST London N17 OBR.

Please send me your list for Engineers

Name (ww)

Address

Post Code

01-808 3050 - 24 HOURS

Vanderhoff Communications Ltd.

The Open University

OPPORTUNITY FOR ENGINEERINGGRADUATE IN TELEPHONE LINE TESTING

A major opening to' a science and/or engineering graduate, givinga rare opportunity to combine industrial R&D experience with

collaboration with the Open University. Salary up to £11,500.The Open University is collaborating with Vanderhoff Communica-tions Ltd (subsidiary of Vanderhoff PLC) to investigate advancedtechniques in the Testing and Analysis of Fault Conditions onTelephone Lines, in a Teaching Company Scheme, partially

supported by SERC/DTI.As a result we have an excellent opportunity for a young graduatewho wishes to advance h s or her career by participating in a jointventure which offers significant responsibility for product de-velopment, combined with the full support and resources of theOpen University. The successful candidate will have the opportu-

nity to register with the OU for a Higher Degree.Under the conditions of the SERC/DTI Teaching Company Schemewe exoect that applicants will be aged 28 or under and will have agood undergraduate or higher degree, but older candidates will

also be considered.The work will be performed using a variety of measurement andcomputer equipment for simulation and modelling enhancedtechniques which will then be incorporated into a working line test

system for live trials.A background in Digital Signal Processing and Circuit Analysis

Techniques would be particularly appropriate.The successful candidate will be employed on the scheme for aninitial period of two years, with the possibility of permanentemployment with the company after this time. The applicant will bebased mainly at the Vanderhoff Communications factory in

Nuneaton.The post offers an attractive salary combined with the usualbenefits of a major employer. For further particulars and an

app ication form, write toM. Fordham, Faculty of Technology, The Open University,(5087 3) Walton Hall. Milton Keynes, MK7 6AA, or tele-phone Milton Keynes (0908) 653941; there is a 24 -hour

answering service on Milton Keynes (0908) 653868.Closing date for applications: 10/8 87.

Hardware/Software/Systems£9,000 - £25,000

As a leading ieucitment iAinsultuncy we hove a wide selection of opportunities(or high calibre Design, Development, Systems and supporting staff throughout the UK

If you have experience in any of the following then you should be talking to us for,,r next rnreer move

ARTIFICIAL INTELLIGENCE IMAGE PROCESSING ANALOGUE DESIGNMICRO HARDWARE & SOFTWARE GUIDED WEAPONS C PASCAL

ADA RF & MICROWAe E ELECTRO-OPTICS SIMULATION C'lREAL TIME PROGRAMMING SYSTEMS ENGINEERING ACOUSTICSSONAR RADAR SATEI LITES AVIONICS * CONTROL ANTENNA

VLSI DESIGN

Opportunities exist with National, International and consultancy companies offeringexcellent salaries and coreer advancement

To be considered for these and other requirements contact John Spencer orStephen Morley or forward a deeded CV in complete confidence quoting Re. WW/66.

ISTS Recruitment, 85 High Street, Winchester,Hants S023 9AP. Tel: (0962) 69478 (24 hrs).

Recruitment

ELECTRONICS & WIRELESS WORLD titi5

BROADCAST SYSTEMS

Basingstoke

For a growing number of Broadcast SystemsEngineers there is only one company that cantruly satisfy their ambitions. One company withthe scope, range of products, reputation forconsistent quality and the track record inadvanced engineering to stimulate theirprospects.That singular company is Sony Broadcast. Ourspecial skill is to turn Sony's top products intototal systems. We are responsible for buildingsome of the most sophisticated broadcastingstudios and OB units.It is this engineering excellence that draws topprofessionals to Sony. Like you, they relish theopportunity of managing projects from initialtender to final commissioning. They welcomethe chance to work on any size of project -from a single radio car to a million pound TVStudio complex. They also derive a specialsatisfaction from motivating the most ableteams and tackling all aspects of a systemsassignment - audio, video, mechanical,electrical, air-conditioning, power etc.If you would like to join them, Sony wouldcertainly like to meet you. However, we sethigh standards. We look for good relevantbroadcast experience either from anoperational or a systems environment, coupledwith a minimum HND in Electrical/ElectronicsEngineering. Obviously proven powers ofcommunication - at all levels are also crucial.In return, the competitive salary, generousbenefits (including relocation assistance andfree health insurance), promotion potential,chance of international travel and theexhilaration of working in such a uniqueenvironment are very powerful attractions.Please send a detailed cv to: Dawn Swarbrick,Recruitment and Training Manager, SonyBroadcast Ltd., Belgrave House, Basing View,Basingstoke, Hampshire RG21 2LA. Ortelephone our 24 -hour answering service foran application form on: (0256) 59583.

SONT.Y.Broaccast

TechnicalRecruitmentINTERNAL SERVICE ENGINEERCalibrate and repair scopes. dvm s eicTo 03.000 Middx. Ref: 51499

PROJECT ENGINEERTo build and commission testequipment and repair large disc drives.To £15,000 Hants. Ref: 51419

0.A ENGINEERMaintain quality system based on13S5750. update O.A. procedures.

To £13,000 Surrey. Ref: 51489

COMMUNICATIONS ENGINEERUnder!: ke feasibility studies for trafficcontrol communications.To £17,500 Surrey. Ref: 51489

TEST SYSTEMS ENGINEERDesign test systems forcommunications equipment.To £14,000 Middx. Ret: 51481

ApplyCliveden Technical

Recruitment,92 The Broadway,

Bracknell RG12 lARTel: 0344 489489

John DixonCliveden Technical

Recruitment163 Bitterne Road,

SouthamptonTel: 0703 229094

NORTH LINCOLNSHIRE HEALTHAUTHORITYST GEORGE S HOSPITALDepartment of Medical Physics andComputing

Basic GradePhysicistA graduate electronic engineer orphysicist is required to assist in thewide spectrum of work covered bythe Medical Electronics Section ofthe above Department.In addition to the design anddevelopment of medical electronicsand the assessment of commercialproducts. the section is responsiblefor a range of electrophysiologicalmeasurements on patients.Applicants should have a degree inelectronics or an appropriatephysical science subject with astrong interest in electronics, butexperience in the medical field is notessential.Salary (under review) £7,452 to129.996 (1st or 2nd class honoursdegree)Informal enquiries would bewelcomed by Dr N. Gravill on Lincoln(0522) 512512, ext 7246.Application forms and jobdescription are available fromThe Personnel Department,St George's Hospital. Long LeysRoad, Lincoln LN1 1EF.Telephone: Lincoln (0522)512512, ext 7001.Closing date 31 July 1987

Full TimeMaintenance

Engineer required forbusy recording studio

Relevant experienceessential, must be

prepared to work overtime.

Call Karen on:01-354 2525 ,7


Electronic EngineersWhat you want, where you want!TJB Electrochemical Personnel Services is a specialised appointmentsservice for electrical and electronic engineers. We have clients throughoutthe UK who urgently need technical staff at all levels from Junior Technicianto Senior Management. Vacancies exist in all branches of electronics andallied disciplines - right through from design to marketing - at salarylevels from around £8,000 - £25,000.

If you wish to make the most of your quaifications and experience andmove another rung or two up the ladder we will be pleased to help you.All applications are treated in strict confidence and there is no danger ofyour present employer (or other companies _you specify) being made awareof your application.

TJB ELECTROTECHNICALPERSONNEL SERVICES,12 Mount Ephraim,Tunbridge Wells,Kent. TN4 8AS.

Tel: 0892 510051(24 Hour Answering Service)

0064,00111,0041114811141011411

Please send me a TJB Appointments Registration form

Name

Address

49

/,

of

TRAINEERADIOOFFICERS

Are you coking for o secure snore-oosed ^IC" o^_ -

a career in the forefront of mode-communications technology... then . -

loining GCHQ as o Trainee RadioTraining involves a 32 weer residential

course, (plus 6 weeks extra if you con,7ttouch type' after which you will be

appointed RADIO OFFICER and undertcxeo variety sceciclctdutses r-ovenno

We offer you: Job Security GoodCareer Prospects Opportunitiesfor Overseas Service AttractiveSalaries... and much more.

To be eligL - - -Peto obtain on MRGC or -INC in 3Telecommunications subject with

on ability to recd morse at 20wpm

.Jne MPT or 2 years reevont radio operating experience is also eligible.-Ceh.e is on equal opportunities employer

Salaries: .3torting pay for trainees is age pointed to 21 years. For those aged 21or over entry will be at £6,572. After training on RO will start of £9,758 rising by 5

annual increments to £14,387 inclusive of shift and weekend working allowance.Write or telephone for on application

four' to -

.........

laCHQ THE RECRUITMENT OFFICE, GCHQ, ROOM A 1108OAKLEY, PRIORS ROAD, CHELTENHAM, GLOS GL.32 SAJ

OR TELEPHONE 102421232912 3

ilea Working inEducation

LEARNING RESOURCESBRANCHTelevision & Publishing Centre.Thackeray Road. London SW8

TelevisionEngineer00,335-02,639 inc.

A Television Engineer isrequired to work in a sectionwhich is responsible formaintaining a high level ofperformance of a wide range ofsound and vision equipmentrelevant to the production ofcolour television programmes

Applicants mustdemonstrate a thoroughunderstanding of andcommitment to the Authority'sequal opportunities policy.

This post is suitable for lobshare

For further details and anapplication form. to bereturned by 28th August 1987.please write to PER/PS4A.Room 362A. The County Hall.London SE1 7PB. Pleaseenclose sae.

Inner LondonEducationAuthorityLEA iS AN EQUAL

OPPOR' 'Y EMPLOYER

THE PAPUA NEW GU NEA UNIVERSITY OF TECHNOLOGYDepartment of Electrical and Communication Engineering

. the following f Department , I

LECTURER I/II aopointment r I ., field of communications computer applications Thew.[1 be required to teoch at undergraduate level and carry research in a field of

relevance Candidates should have a von: honnurc decree with teaching and Inductrfal research

SENIOR TECHNICAL INSTRUCTOR/PRINCIPAL TECHNICAL INSTRUCTOR

communications Some experience with computer, information systems would be an advantage Thesuccessful applicant will teach on courses leading to Diplomas in Communication Engineering andElectronic Engineering. Candidates shoulo have a higher diploma degree qualification with teaching anindustrial .xperienre

SENIOR TECHNICAL OFFICER/PRINCIPAL TECHNICAL OFFICER r onic

should be wet qualified and have substantial experience of electronic workshop and 'dull -findingtechniques.Salary Level. Lecturer 1/11 K19.2304(21.030 per annum. STI PTI K19 230-1(22.930 per annum. STO/PT(17.840-K21.080 per annum. (K1 =Stg 0 6728 approx) Level of appointment will depend uponqualification and experience.Initial contract period is for three years Other benefits include a gratuity of 24% taxed at 2%. support forapproved research. appointment and repatriation fares. settling -in and settling -out allowances. six weekpaid leave per year, education tares and assistance towards school fees. Free housing salary protectionPlan and medical benefit schemes are available. Staff members are also permitted to earn fromconsultancy up to 25% of earnings annually Detailed applications(two copies) with curriculum vitae andthe names and addresses of three referees. including telephone numbers and indication of earliestavailability to take up appointment. shoull be received by the Registrar. Papua New Guinea University ofTechnology. Private Mail Bag. Lae. Papua New Guinea. by 8 August 1987.

Applicants resident in the United Kingdom should also send one copy to the Association ofCommonwealth University lApPts). 36 Gordon Square. London WC I H OPF from who further informationmay be obtained 467

ELECTRONICS WORKSHOPTECHNICIAN (Grade 5) required in theUniversity of Reading, Department ofPsychology. From mid -September 1987.The successful applicant wi'l be required toadvise staff and students on electronicsproblems and to design and construct awide range of specialised equipment. TheDepartmental research and teachingactivities depend heavily on the use ofcomputers. The ideal candidate w.11 havehad a recognised apprentice and at least 2years of varied experienced in electronics.Knowledge of BBC BASIC and ct 6502would be an advantage. Salary scale£7696to £9086b p.a. Application form availablefrom the Personnel Officer. University ofReading. Whiteknights. PO Box 217.Reading. RG6 2AH. telephone (0734)875123 ext. 220. Please quote Ref. 7.30A.

466

BOX NosBox number replies

should be addressed to'

Box No.cio Electronics &Wireless World

Quadrant House,The Quadrant,

Sutton,Surrey SM2 5AS

ELECTRONICS & WIRELESS WORLD s

LABORATORYr

London, W1

We are an equalopportunities employer

Two vacancies exist for Laboratory Engineers in the Radio Frequency Section ofDesign Group. The section designs a wide range of RF equipment for broadcastuse including low -power transmitters, receivers, transposers, radio links, radiomicrophones and specialised portable test equipment. The section has access to afull range of modern test and measurement equipment and makes comprehensiveuse of in-house CAE and CAD facilities.

Laboratory Engineers contribute to all stages of the design process includingbuilding and testing prototypes, preparing and checking documentation etc, andare expected to work with the minimum of supervision.

Experience of RF work in the frequency range 10 MHz - 1 GHz would be anadvantage, possibly in the fields of RF measurements, mechanical design andscreening, manufacturing methods or surface -mount technology.

Laboratory Engineers should possess an HNC, Higher TEC or equivalentqualification and are normally appo nted in the salary range £8,026 - £11,043;progression to a roof of £12,927 beiig subject to proficiency reviews. Applicantswith a TEC or equivalent, or those studying for HTEC will be considered for appoint-ment as Laboratory Technicians in the salary range £6,780 - £8,775. (Salariescurrently under review).

Design Group is currently located in Central London but will be moving toChiswick in Spring 1988.

Benefits include subsidised restaurant, social club facilities, contributorypension scheme and five weeks annual leave.

For further information contact John Sykes on 01-580 4468 ext. 4143.For application form contact us immediately (quoting ref. 3750/WL and

enclose s.a.e.) BBC Appointments, London W1A 1AA. Tel: 01-927 5799.

Electronics Technicians, Electricians

II

IBut you can make a world of difference.'Skills taken for granted in this country are vital to

members of poorer Third World communities.In a world which by and large does nothing about

work directly with those seeking to become more self -poverty and injustice (let alone hunger), VSO volunteers di

Ireliant:No one can change the world. But as a VSO volunteer,

you could make a world of difference.I work for February 1988 departure overseas. Most of IIWe have several jobs in electronics and electrical

the electronics jobs are in the repair andImaintenance of a variety of control systems. The

school leavers in basic electrical installation in Ielectrical jobs are principally in the training of

Idomestic and light industrial work. All the jobsrequire people who enjoy working with others andpassing on their skills. If you have a relevant skilland qualification please apply now.I for the local rate of pay. Postings are for two years -

You should be without dependants and willing to work

although many choose to stay longer - and most U.K.employers should be prepared to grant leave of absence.

For more information please complete and return.

I'm interested I have the following

1

I loin theAddress -

j

Name Ihuman race.

Cnanty no 313757

YOU CAN'T CHANGE ITHE WORLD.

PLOW 8.87

Post to Enquiries Unit, Voluntary Service Overseas,mil9 Belgrave Square, London SW1 X 8PW. (24p S A E appreciated)

ARTICLES FOR SALE

TO Al' 'U t URERS, WHOLESALERSBULK BUYERS. ETC.

LARGE QUANTITIES OF RADIO. TV ANDELECTRONIC COMPONENTS FOR DISPOSAL

SEMICONDUCTORS, all types, INTEGRATED CIRCUITS, TRANSISTORS,DIODES, RECTIFIERS. THYRISTORS, etc. RESISTORS, C/F, M/F, W/W, etc.

CAPACITORS, SILVER MICA, POLYSTYRENE, C280, C296, DISCCERAMICS, PLATE CERAMICS, etc.

ELECTROLYTIC CONDENSERS, SPEAKERS, CONNECTING WIRE, CABLES,SCREENED WIRE, SCREWS, NUTS, CHOKES, TRANSFORMERS, etc.

ALL AT KNOCKOUT PRICES - Come and pay us a visit ALADDIN'S CAVE

TELEPHONE: 445 0749/445 2713R HENSON LTD.

21 Lodge Lane, North Finchley, London, N.12(5 minutes from Tally Ho Corner 11611

BR I DGES wa veform n/transi storanalysers. Calibrators, Standards.Millivoltmeters. Dynamometers. KWmeters, Oscilloscopes. Recorders.Signal generators - sweep, lowdistortion, true RMS, audio, RM,deviation. Tel: 040 376236. (2616)

RF ENGINEERING SOFTWARE FOR BBC(B) ONTAPE, WITH NOTES

smitlahart: Comprehensive chart matching program(C413

Threeprobt Computes impedance etc at load, usingsimple system with 3 voltage probes SuitableVHF/UHF (P42 incl). (Hardware described but notsupplied

Other RF software shortly. Details from.

SCIENTIFIC AND ENGINEERING SOFTWARE

PO Box 416, Marlow 517 IkU (06284)2508 474

TEST EQUIPMENT FOR SALE:Signal generators, TV patterngenerators, scopes, power supplies, y -trecorders, DMM's etc, too many to list,write (with sae) for mere informationand prices to Box No.478 478

G.W.M. RADIO LTD40 42 Portland Road. Worthing. Sussex.

Tel: 0903 34897

Constantly changing stock of interest-ing items for callers. Receivers, MarconiApollo £250. Racal RA17 £200 for call-ers. Datel 2412 modem units 2/TG 2507-8modules 1, 2 & 3 in modem case assem-bly £30 inc p&p. Other telephone partsin stock. Ex -navy quartz chronometersdear 7cm dial in wood case £27 inc p&p.

When replying to

classified advertisements,

readers are recommended

to take steps to protect

their interest before

sending money

Laboratory Retractometer. Bellingham &Stanley/Pulfrich £185 Vacuum pump.Edwards. rotary £98 Monocular & stereo-scopic. binocular microscopes. Vinyl lac-quer 85p/tin Micro -spot-welding nead £59.Hand rivets for assembly of prototype cases£1/100. Pneumatic nail/staple gun £89. Diffu-sion pump £39. 15V 275 MAH Nicads £2each. Bandsaw £79. Crystalline pure siliconslices, mirror finish. for experimental semi-conductors. also make miniature or microheating elements for microscopy etc., whenused in large or small pieces. Alternativevises in artistic. jewellery & decorativefields. 1" recording tape on NAB spools B.Blue Steel rocket sections £50 for 5. 60Wintegrated circuit amplifiers. will make light-weight loudhailer units, compact sounddistribution systems. boost low power sys-tems etc. £6.50 each. HOUSE/FACTORY/STORAGE/OFFICE PREMISES in W. Sussex.Make two dwellings or one very large, orhome/works. or one large works. Officeequipment - steel drawer unit, steel slidingtop unit. filing cabinet, multi -drawercabinets (15 -drawer). Run of steel labora-tory benching.TEL: 040-376236. 2016


SERVICES ARTICLES WANTED WANTED

CIRCOLECTHE COMPLETE ELECTRONIC SERVICE

Artwork, Circuit Design, PCB Assembly, Test & Repair Service. Q.AConsultancy, Prototypes, Final Assembly. Full PCB Flow SolderingService.Quality workmanship by professionals at economic prices.Please telephone 01-646 5686 for advice or further details.

TAMWORTH MANOR302-310 COMMONSIDE EAST, MITCHAM

113911

PCBs MANUFACTUREDFor your circuits

* No lob too big or too small* Single or double sided* Real professional finish* Low cost * Delivery 1 to 3 weeks* Precision artworks/photography* 5 years experience

Talk to Graham Slee or Mark Fisher on

0226 200717undoGrange Lane Industoal EstateCAI,W004 RoadBarnsleySouth Yorkshae S701SEEniOnd

PRINTED CIRCUIT BOARDSsingle and double sided manufacturedto your specifications. We offer qualityand fast service (same day onprototypes) at very keen prices. Wesupply copperclad material cut to size.Artwork service available. MondoCircuits Ltd. 35 Grosvenor Rd,Twickenham,Middx. 01-8915412. 413

TURN YOUR SURPLUS i.cstransistors etc. into cash, immediatesettlement. We also welcome theopportunity to quote for completefactory clearance. Contact COLES-HARDING & CO, 103 South Brink,Wisbech, Cambs. 0945 584188. 1921

TEST EQUIPMENTFOR SALE & WANTED

Buyers and Disposal Officerscontact:

COOKE INTERNATIONALUnit 4, Fordingbridge Site

Main Road, BarnhamBognor Regis

West Sussex P022 OEBTel: 0243 685111/2

Stock list available, send 50pto cover p&p.

179

P.C.B's DESIGNED. Artworkcapacity available for single/doublesided, P.T.H. and multi -laver P.C.B.'s,also silk screens, solder masks, labelsetc. For C.A.D. Photoplot. artwork &photography, contact Mr. Williams, 49VVestbourne, Honeybourne. Evesham.Wares, WRI1 5PT. Tel? 0386 832152.

409

WANTED. Used test equipment for 9-40 GHz such as: Signal source,frequency counter etc. WALLFASS,P.O.Box 1244, D-405 Moenchen-gladbach, W -Germany. Tlx 852582. Tel2161-88555.

WANTEDTest equipment, receivers,valves, transmitters, com-ponents, cable andelectronic scrap and quan-tity. Prompt service andcash. Member of A.R.RA.

M & B RADIO86 Bishopsgate Street

Leeds LS1 4BB0532 435649

PLEASENOTE:

with effect from theSeptember 1987

issue, classified rateswill be

Displayed:£25 per single column

centimetre

Linage:£5.50 (minimum £45)

Box No.'£12 extra

STEWART OF READING110 WYKEHAM ROAD.

READING RG6 1PL.TEL: 0734 68041

TOP PRICES PAID FOR ALLTYPES OF SURPLUS TESTEQUIPMENT, COMPUTER

EQUIPMENT, COMPONENTSetc. ANY QUANTITY

IGOLLEDGEI

IELECTRONIC_f] --jQUAlt 1 /. (' 1{1',1 kLSOSCILLXI ORS AND Fli.FERS ofall types. Large stocks of standarditems. Specials supplied to order.Personal and export orders welcomed -SAE for lists please. OEM support thru:design advice, prototype quantities,production schedules. GolledgeElectronics, Merriott, Somerset YA165NS. Tel: 0460 73718. (2472)

BUSINESS OPPORTUNITIES

WANT TO RUN YOUROWN BUSINESS?

hew .1,ii-tr I:1, TIL

and goodwill of a well-known on -going series of

international semiconductor reference manuals with

worldwide circulation. Established 10 years with

standing order subscriber lists.

Requires good (but not extensive) knowledge of

semiconductor usage with minor graphic and word

processing. Runs on desktop micro. Sales £38-40k

po. Easily expandable. Good profit margin

Operational location unimportant.

£20,000 all -in for this going concern.

Your premises.

Box No 472 472

ADVERTISEMENTSfor your

ADVERTISEMENT

Advertisement Dept.,

NAME

ADDRESS

Sales andINDICATED ON FORM

Quadrant House, The Quadrant,

CLASSIFIEDUse this Form

PLEASE INSERT THE

To "Electronics & Wireless World" Classified

Rate £5.50 PER LINE. Average six words perline. Minimum £45 (prepayable).

Name and address to be included in charge ifused in advertisement.

Box No. Allow two words plus £12. Cheques, etc., payable to "Reed Business

Publishing" and cross "& Co." 15% VAT tobe added.

WantsBELOW

Sutton, Surrey SM2 5AS

iiI

REMITTANCE VALUE ENCLOSED

PLEASE WRITE IN BLOCK LETTERS. CLASSIFICATION NUMBER OF INSERTIONS


MATMOS LTD, 1 Church Street, Cuckfield, West Sussex RH17Tel: (0444) 414484/454377

COMPUTER APPRECIATION, 111 Northgate, Canterbury, Kent CT1 1BHTel: Canterbury (0227) 470512. Telex: 966134 COMPAP GPLESSEY Model T24 V22./V22 bis MODEM. Compact, automatic modem featuring the latesttechnology and the highest possible data rate over the ordinary phone system. Offers both V22and V22 bis compatibility. 1200'2400 Baud operation with auto bit rate recognition, operation on bothordinary phone (PSTN) and private circuit (PC). auto call and auto answer. duplex operationallowing simultaneous transmission and reception of data at 2400 Baud in both directions over asingle phone line. compact size IT T , 2' 2"). BT approved BRAND NEW £350.00VICTOR SpeedPac IBM PC ACCELERATOR. Half -wide 80286 accelerator card for IBM PC andcompatibles. With 8kbyte cache memory and 7 2MHz 80286 processor Offers AT/PersonalSystem 2 performance for a fraction of the cost Runs up to 7.5 times faster BRAND NEW £199.00ITT SCRIBE III WORD PROCESSING SYSTEM. Professional system originally selling ataround 06.000 without printer. Now offered with software Included and a variety of differingoptions (including hard disc. comms.. etc) available This system is available from us ALLBRAND NEW at the cost of current budget systems. but with office -quality performance andfeatures With dual processor workstation (TI 9995 and 2130H). 12" green display with slowscrolling. 128kbytes RAM. dual 500kbyte SHUGART 514" floppy disc drives and comprehensivesoftware £290.00DATA GENERAL Various items just arrived. Including CS60 system Dasher VDU terminals,tape drives. Winchester systems and drives etc. etc. Please enquire for details.TOSHIBA Model T300 PC SYSTEMS. 8088 (6MHz) based system with 192kb memory. 2 v halfheight DSDD floppy disc drives each of 655kb capacity (unformatted), high resolution greendisplay 25.80 characters. RS232 and CENTRONICS parallel interfaces. MS-DOS Ver.2 andT -BASIC are included together with a diagnostic disc and manuals We have low cost packagesavailable now. including Word processing, Spreadsheet etc. BRAND NEW .. £295.00AS ABOVE. but with single floppy disc drive £249.00TOSHIBA Model T100 CP/M SYSTEM. With 64kb RAM. dual 500kb 5' 4" floppy disc drives.serial and parallel interfaces. BRAND NEW £149.00MATMOS TERMINAL MATMOS PC with Ver 2 EPROM for terminal emulation Probably thelowest cost terminal available anywhere. With set-up menu and with data rates up to 9600 BaudMachine is easily modified for split Baud rate operation, 75/1200B. EPROM plugs into ROM socketaccessible from exterior Emulation is VT -52 compatible for cursor addressing, and for characterattributes as far as the MATMOS PC allows MICROSOFT BASIC is still available from thekeyboard (EPROM card on its own. £20.00. EPROM card with socket only for user's ownEPROM, £8.00 I BRAND NEW £69.00HITACHI Model 3055/SX r disc drives. With SHUGART compatible interface as for 5t4"drives. Uncased. 125K (single density) or 250K (double density), 40 track, 100 tpi: soft sector,3ms track to track time. standard 34 way edge connector. 12V and 5V powered (standardconnector) with overall 3.7W typical power consumption These drives have been tested by us onthe BBC whh DFS. on the AMSTRAD 6128 and on the TATUNG EINSTEIN. and are alsoknown to be suitable for the AMSTRAD 664 and as a second drive for the AMSTRAD 464Single -sided. 250kb unformatted BRAND NEW Data cables are available from us for theAMSTRAD 6128 and BBC at £7.50, and an installation pack including data and power cables withinstructions is available for the TATUNG Einstein at £12.00 £24.95KIMTRON Model ADM 85H VDU 24 v 80 intelligent VDU with TELEVIDEO 920/925 emulation.Features include. non-volative set-up with status line, auxiliary port. all Baud rates 50.19200Baud, cursor addressing, etc £185.00DEC PDP 11173 SYSTEM with DZV11 asynchronous multiplexor. OLV11 serial interlace.BA1 1-SB 9 v 4 backplane. TANDON 8' floppy disc drive with DILOG Model RXV-21 controller.2 AMPEX PYXIS 27 5r 4" Winchester disc drives with DILOG D0614 Controller All containedin DEC cabinet with DEC power control. 512kb memory £3,950.00DEC PDP 11/23 SYSTEM with DLV11-J quad serial interface. TANDON 8" floppy with MICRODEV. ASSOCS. Model MXV-21 controller. FUJITSU 8" Winchester disc drive with DILOGD0401 controller. 256kb memory Contained in 2' 19 rack units in portable carrying case.Supplied with XENIX operating system complete with all original manuals and distributiondiscs £1,385.00

ATARI Model 1024 STF personal computer with Model SC1224 medium resolutioncolour monitor. With flight simulator plus all usual software Originally purchased 1987 and asnew £565.00CALCOMP Model 463 AA graph plotter C325.00TOSHIBA HIGH RESOLUTION COLOUR MONITOR RGB input BRAND NEW £100.00TOSHIBA LOW RESOLUTION COLOUR MONITOR IBM compatible with RGB input BRANDNEW £95.00APPLE PROFILE 5mbyte WINCHESTER disc drive for APPLE II computers Complete withProDos operating system and additional interface for APPLE III £250.00APPLE Ile MUSIC SYSTEM incorporating 64k computer with dual disc drives and extensivesoftware. ALPHA SYNTAURI music system includes 5 octave keyborad and extensive softwarefor voice creation and 16 track sequencing £650.00FUJITSU Model M2230AS 514' WINCHESTER disc drive 6.66mbyte capacity unformatted.16/32 sectors 320 cylinders With ST506 interlace BRAND NEW £75.00DEC Model VT101 VDU with many leatures including non-volatile set-up, slow scrolling etc £150.00DEC Model BA11MF box power supply and 8 slot backplance for 0 -bus BRAND NEW £1 85.00ADAC Model 1822 128kbyte CMOS memory for 0 -bus. With battery backup.... £175.00TEKTRONIX Model 7603 100MHz OSCILLOSCOPE MAINFRAME with CRT readout £590.00TEKTRONIX Model 7403N 60MHz OSCILLOSCOPE MAINFRAME £250.00TEKTRONIX Model 7A18 75MHz 5mV dual trace amplifier £175.00

2911.0° °

TEKTRONIX Model 7A18N As above, but without CRT readoutTEKTRONIX Model 7853A 100MHz dual bmebaseTEKTRONIX Model 7D12 AID converter with M2 sample/hold module. . £550.00TEKTRONIX Model 5403/D40 60MHz oscilloscope with 2., 5A48 & 5B42 £750.00TEKTRONIX Model 5103/D15 2MHz single beam storage oscilloscope with 2 5A113N andSBION £550.00HATFIELD INSTRUMENTS Type 2105 50 Ohm attenuator£10.00BRYANS Model 45000 UV recorder with 6 v 45001 amplifiers 1 mVicm-50Vicm and chartspeeds from 1 mmimin-500mmis Timing line interval is adjustable from 0.02s -10s and RecordDuration 0 5s -20s With remote control facility £195.00TEKTRONIX Model 851 digital tester 'One knob lets you dial 22 functions . 11 functionsmeasure timing. two register plus and minus peak voltages. three carry out DMM measurements

and one reads line voltage at the outlet the 851 measures its 4 input thresholds to adjust tothe logic levels of the equipment being serviced 512 digit readout. 35Mhz frequency'timg,autoranging etc £95.00HEWLETT PACKARD Model 59307A dual VHF switch. DC to 500MHz 50 Ohm switch for HP-IB E185.00HEWLETT PACKARD Model 5045A digital IC tester with CONTREL H310 automatichandler With IEE interface and print out of test results either passlf ail or full diagnostic includingpin voltages at point of failure With full complement of pin driver cards and complete withsubstantial library of mag card test programs for 74 series TTL and other ICs CONTREL handlerallows fully automatic testing of ICs which are sorted into 2 bins Price includes a secondHP5045A (believed fully operational) for maintenance back-up £1,950.00TEKTRONIX Model 577 SEMICONDUCTOR CURVE TRACER With 5 heads for variousdiodes and transistors In excellent condition £1,900.00TIME ELECTRONICS Model 9810 programmable power supply £190.00TIME ELECTRONICS Model 505 DC current source, 0 05% With leather case £75.00

PLEASE NOTEVAT & carriage must be added to all prices. Please enquire for details.'All new equipment is guaranteed for 6 months.


INDEX TO ADVERTISERSAppointments Vacant Advertisements appear on pages 865-869

PAGEAdenmore Ltd 768Airlink Transformers 816

BAF Box Ltd 772Brigden Technology 854

Carston Electronics Ltd ....773,848Cavendish Automation 849Channel Micro Products Ltd 770Cirkit Holding 864Clark Masts 772Colomor Electronics Ltd 781Computer Appreciation 870

Data Acquisition Ltd 825Dean Microsystems 768Display Electronics 856/857Dowty Maritime Systems 771

Electronic Brokers Ltd... IFC,OBCElectronic Source, The 848

EMS Manufacturing LtdEssex Electronic CentreE&WW Databank

PAGE

780779848

Fibre Optic Links 779Field Electric Ltd 848Fluke (GB) Ltd Advert Card

Happy MemoriesHarris Electronics Loose InsertHarrison Electronics 864Hart Electronics 851Henry's Audio Electronics 770Henson, R. Ltd 852Hilomast Ltd 851

ICOM (UK) Ltd 768Ines GmbH 781

JDR Sheetmetal 864Johns Radio 768

PAGE

Kestral Electronic ComponentsLtd 772

Langrex Supplies Ltd 826

MA Instruments 780Maplin Electronic Supplies 781

848 M+B Radio (Leeds) 772Micro Concepts IBCMicrokit Ltd. 854

Number One Systems 852

Parasitic Research Inter 851Pineapple Software 837P M Components Ltd 860/861PSI Systems 816

Quart Hand Ltd 852

PAGE

Raedek Electronics 864Ralfe Electronics 837

Second Test Ltd 838Semicon Indexes 848Sherwood Data Systems 780Stewart of Reading 779Surrey Electronics Ltd 770

Target Electronics 780Taylor Bros (Oldham) Ltd 825Technomatic Ltd 814/815Thurlby Electronics Ltd 854Time Electronics 771Triangle Digital Services 770

Valradio Power Ltd 852

Webster Electronics 816Withers, R. Ltd 816

OVERSEAS ADVERTISEMENT AGENTSFrance and Belgium: Pierre Mussard, 18-20 Place de la Madelaine, Pans 75008.United States of America: Jay Feinman, Reed Business Publishing Ltd.. 205 East 42nd Street, New York, NY 10017 - Telephone (212) 867 2080 - Telex 23827.

Printed in Great Britain by E.T. Heron (Print) Ltd, Crittall Factory, Braintree Road, Witham. Essex CMS 3Q0, and typeset by Graphac Typesetting, 181'191 Garth Road. Morden. Surrey SM4 4LL.for the proprietors, Reed Business Publishing Ltd, Quadrant House. The Quadrant, Sutton, Surrey SM2 5AS. 0 Reed Business Publishing Ltd 1987. Electronics and Wireless World can be obtainedfrom the following: AUSTRALIA and NEW ZEALAND: Gordon & Gotch Ltd. INDIA: A. H. Wheeler & Co. CANADA: The Wm, Dawson Subscription Service Ltd.. Gordon& Gotch Ltd. SOUTHAFRICA: Central News Agency Ltd: William Dawson & Sons (S.A.( Ltd. UNITED STATES: Eastern News Distribution Inc.. 14th Floor. 111 Eighth Avenue. New York. N.Y. 10011.


IMAGE -10HIGH PERFORMANCE SINGLE BOARD GRAPHICS COMPUTER

High performance? they all say that.Ah yes, but. . . .

Image -10 has three 16 bit processors workingconcurrently for dazzling speedandImage -10 has a resolution of 768 by576 pixels for intricate detailandImage -10 has 256 on -screen colours givingphotographic quality.In fact Image -10 has all of therefinements asked of graphics intensiveapplications. Like hardware windows,hardware drawing, hardware zoom and pan,hardware character generation and hardwareblock moves. The only thing the cpu has to dois to run your applications.

Board Price £1295Image -10 includes:

* 68010 cpu * 82786 gdc* Floppy disc controller* Dual RS232 * Centronics

* Battery backed clock* 512K dram * 64K sram* Winchester disc controller

* Mouse * transputer ports* Stereo sound generator* 256K eprom

SOFTWAREDEVELOPMENT

SYSTEMS

Versatile, low cost, developmentsystems based on Micro Concepts'single board computer, Microbox 3.A choice of powerful disc operatingsystems is available including Os -9/68K, TRIPOS. and CP/M-68K

Features include:68000 microprocessor, 8MHz clock.512K dynamic ram. 64K static ram,128K epromOn -board graphics controller offers:80 column, 24 row, 16 colour text.640 by 480. 4 colour graphics.320 by 480, 16 colour graphics.Floppy disc controller.Winchester disc controller.Dual serial RS232 ports.Centronics printer port.Up to 48 lines of parallel input/output.Battery backed clock/calendar.I/O expansion capability.Double eurocard board format.

Hardware prices:Fully built and tested Microbox 3 singleboard computer £650System with dual 1M byte floppy discdrives £1195System with 20M byte Winchester andsingle floppy drive £1895

Access

UNIVERSALCROSS ASSEMBLERS

for allMSDOS OS -9 FLEX

computers

Supports the following devices1802 809 80496502 8022 80806803 8039 87516808 80C48 63038021 8051 680280C35 8749 68HC058048 68000 80208050 6301 80358748 6801 8040Z80 6805 80C491805 68HC11 80856800 8031 Z86804 80C39

Extersive directives supportmodular, conditional andstructured program-ming. Powerful Macro Pre-processor, xref generator andoutput conversion utilitiesprovided.

COMPLETE SUITE £295INCLUDES FULL 'C' SOURCE CODESNOT 68000)

MICROPROCESSORCONTROLLED

EPROM EEPROM andSINGLE CHIP MICRO

PROGRAMMER

Programs the following devices:2508 2516 25322564 27(C)16 27(C)322732A 27(C)64 2764A27(C)128 27128A 27(C)25657(C)256 27(C)512 275132758 27011 2816A2817A 2864A 2825648Z02 521313 52132352B33 68732 6876468766 8741 87428744 8748 87498748H 8749H 87518755 9761 CY7C282CY7C292 DS1225AT -ROM PC -ROM XT -ROM

No Personality Modules required.Controlled via RS232 interlace.Accepts Intel, Motorola, Ascii-hexand binary data. Ultra -fast fast andstandard programming modes.Low and high byte programmingsupported.Completely self contained butuncased. Price £295

All prices shown exclusive of VAT and carriage

Micro Concepts 2 St. Stephens Road, Cheltenham, Gloucestershire G L51 SAATelephone (0242) 510525


ELECTRONIC BROKERS

MARCONIINSTRUMENTS2958Total Access ControlSystem (TACS) adaptorcombines with 2955 toallow automatic testingof UK cellular radiosystems. Non-volatilememory allows users toset their own testprogrammes. Adaptorfits to 2955 using GPIBmodule.

£9160

MARCONIINSTRUMENTS2019A80KHz-1040MHzAM/FM signalgenerator, withfrequency resolution of10Hz up to 50MHz.Excellent output levelaccuracy with output to2V emf. Comprehensivemodulation indudesauxiliary FM. GPIBoption with Talk/Listencapability

£4450Also: 2018A,80KHz-520MHzversion £4150

MA NTS2955 Radio Communications test set with comprehensive testingfacilities for AM/FM/PhM radios to 1GHz. Modes include TX, RX andfull duplex. Menu -driven CRT display with store/recall of 38 standardsettings. Digital readouts for precision, plus bar charts and digitalstorage scope, built in self test and help facility. GPIB option for full

operation for benchtop and radio versatility. E5900instrument control. Single and two -port

MI. 7.1701

1000000

MARCONIINSTRUMENTS -

2022A10KHz-1000MHzAM/FM signalgenerator, withnon-volatile memory of100 settings.Comprehensivemodulation coveringAM/FM/PhM. Singlepush-button operation,large LCD display,optional GPIBprogrammability, lowharmonic distortion.

£2950

MARCONIINSTRUMENTS2305Modulation meter with0.5MHz to 2.3GHzfrequency range.FM/AM/PhMmeasurements. High0.5% basic accuracy.Fast testing withmicroprocessoroperation andauto -tuning.Frequency and powermeasurements inaddition to modulation.Programmableoperation with GPIBoption.

£5600

UK's No. 1 Test Equipment Distribution Companyrl o O t- q

LI

Electronic Brokers )

° Electronic Brokers, 140/146 Camden Street, London NW1 9PB

Fax: 01-267 7363. Telex: 298694. Tel: 01-267 7070.ENTER 3 ON REPLY CARD All prices exclude VAT and carriage and are correct at time of going to press.

miiminm10111:420=0111m:101m110111m electronics...tf 2604 voltmeter £395 tf 2702 inductor analyser...

Documents