i PRICES INCLUDE VAT * PROMPT DELIVERIES * FRIENDLYSERVICE T LABGE S STAMP FOR CURRENT I IST

vryrr r vyyLtt ^tytr Ltt tLfl tyt\.rULrLEg Nowenloya!vor d wrdP reputatlon for qualrtv rel ablltty and performance at a rea sttc once Four modeis

ava able to surt ihe needs ol the professtona and hobby market I e, lndustry Lersurelnslrumenta and l-l F etc When compalng prrces NOTE al models nclude TorordalpowerSUpply lntegra heaLsrnk.GIaSSftbrePCB. andDlvecrrcurtstopowercompatlbleVu meter Open and shorl c rcurl prool SUpplied ready bUilt and teSted.

OMPlOO Mk ll Bi-Potar Ourpur power 'l lOwatts R M.S. rnto 4 ohms, Frequency Res,ponse 15Hz-30KHz 3dB, THD 00j%.S.N.R l18dB, Sens f or Max output5O0mV at'lOK, Srze 355 X I l5 X 65mm.PRICE f33 99 + f3.00 p&pOMp/MF1O0 Mos_Fet Ourput power 110watts R M S rnto 4 ohms, Frequency Response 1 Hz IOOKHz 3dB, Damprng Factor80, Slew Rate 45V,'uS, T H D Typrcat0 002% lnput Sensrtrvrty 5OOmV, S N R-125d8 Size 3OO \ 123 r 60mm PRTCEPRTCE f39 99 + f3 00 P&P

r 2OORes

ponse 1 Hz 100KHz 3dB, Damprng Factor250, Slew Rate 50V,,uS, T.H.D Typrcat0 0O1"/", lnput Sensrrrvrty 500mV, S N R

130d8, Size 300;r 150 \ 1OOmm PRTCEPR|CE f62 99

- f3 50 P&P

OMP,/MF3O0 Mos,Fel Output power 3OOwatts RMS rnto 4 ohms Frequency Response 1 Hz 1OOKHz 3dB, Damprng Facror350, Slew Rare 6OV,uS, T H D Typrcat0OOO89t, lnput Sensrtrvrty 50OmV, S N R

13OdB, Srze 330 \ 147 ! 102mm PRTCEPBTCE f79 99 f4 50 P&P

NOTE: Mos Fets are supp ed as srandard il OOKHz bandw dtlr & npur sensrrrv y 5OOmV) j requ redPA versron(50KHzbandwdth & tnpul Sens I vrry 775mV) Order Sta|dard or pAVu METER Cornpatb e w Ih our four amol I ers deta e.l ahovp A

PRICE f8 50 - sOp P&P

Beller lo be Alarmed lhen terr ledThandar's famous Minder' Burglar Alarm SystemSuper or nr cro,rave pilnc ple Supp red as three un lscomp ele wrlh nlerconnecl on cab e FULLYGUARANTEEOControl Unit Houses m crowave radar un .a-traup lo I 5 melres adJUSlab e by sensrl v ty aa-. aThree pos lron key operaled lacra swrlch orr :a:'

armed 30 second er t and entry de aylndoor alarm E ectronrc swept lreq s ,a-1 04dB outputOutdoor Alarm Eleclron c swepl lreq srren

-.ti3outpul Hotrsed rn a tamper prool heavy duiy .;-'-caseBolh lhe conlro un I and outdoor a arnr conla n .ecnargeab e baller es wh ch prov de lul protect c-dur ng nra ns lar ure Power requ renrent 200,260 Va :AC 50/6OHz Expandabe wilh door sensors oanab!llons elc Conroiete vr lh nstrucl onsSAVE 1138.00 usual Pr ue t228 8sBKE's PRrcE f89.99 + f4,oo p&p

7 Why buy a c,, ect on of sell assemb y boards'

IDEAL for Workshops, Factones,O[frces, Home.etc Supplredready burit

OMP LINNET LOUDSPEAKERSThe very best in quality and value N/lade spec atl! ro s- ::_: :need for conrpactness wrth lr gh sound oulput evels : - :-:lhard wear ng black vynrJ. w lh protectrve corners gr E :-: .-hande Al modelsSohms Full range45Hz 2OKH2 :::-_15 12" Walls B M S per cabrnet Sensrt v1\ l!'

-' :iOMP 12-100 Watts 1OOdB. Price f 149,99per parrOMP 1 2-200 Watts 1O2dB. Price f 'l 99 99Pef Pair De very Secur cor i3 -'- :: ::

Professional 19" cased Mos-Fet sterecampsr Used the World over in clubs, pubsdiscos etc. With twin Vu meters, il'. -toroidal power supplies XLR connectic.sMF600 Fan cooled. Three models (Ra.r;.:sR.M.S into 4 ohms). lnput Sensitivity 775m!MF200 (100 + 100)W. f '169.00 SecuricorM F400 (200 + 200)W f228 85 Deliverv-MF600 (300 + 300)W. f322.00 f 10 00

19" STEREO RACK AMPS '1 ( iir l-St ]= : U!dER

" upro400WATTS R.M Sge selection of McKenzieuding Cabrnet Plans. Large

POWER BANGE8"50 S Hi-Fr,,?a oz " ai y vo round a ly i xrng esc|tctreon Res Freq 4OH? Freq Hesp ro6KHz PF CEfl abie w 1tr btack gr ile ft 1 99 p&p ft 50 ea12 1 .M.S Hi-50oz riagnel 2'alyvoce corl Gro!ndallylxngescutcheon Dr-.castchissrs Wl le.ofe RpcEr pq25Hz Fteq Resp to 4KHz Sens gsdB PRCE f2B 60 f3 OO p&p eaMcKENZIE1 .M.S. C1285cP Lead guitar/keyboard/Dlsco2 lAlly@ntredomeRe;Freq+Slzrreq.niip to65KHzSens gBdBpRtCE!299912" 85 WATT R.M.S. Cl285TC P.A,/Diso 2" afly yoice coit. Twin @ne.Res. Freq. 45Hz Freq. Besp. to 14KHz PRICE C31.49 + 13.00 P&p ea15" 150 WATT R.M.S. C15 Base cuita/Oisco3" a Pea10"2"v10"2"v15"Res15" i()o WATT R M.S. Cl5400 High Power Bass.Bes. Freq. 40H2. Freq Besp. to 4KHz. Sens. 102d8 pRICE C89 52 +f4 OO p&p

1" HzSens.SgdBpBtCEl22.OO+C1.5Op&pea1 " 92dB PRICE C32.OO + C1 .50 p&p ea.'101 " voice coil Res. Freq. 35Hz Freq. Resp. to 4KHz Sens. 92dB pRtCE t36 OO + f2 OO p&p ea12" I)0 WATT R.M.S. Disco/Sound re+nlorement ets.11/2" voice coil Hes Freq 35Hz Freq Besp to 4KHz Sens 94dB pBtCE 4700 + C3OO p&pea

SOUNDLAB (Full Bange Twin Cone)5 60 WATT R M S Hi,FilMulliple Array Disco etcl"vorcecorl Res Freq 63Hz Freq Resp to20KHz Sens B6dB pRCE1999 ftOOp&ped6,r" 50 WATT F M S Hi-Fr/Multiple Array Disco etcl vo.ecoil Bes Freq 56Hz Freq Resp ro20KHz Sens 89dB pRCEft099 ft 50p&pea8 60 WATT R.M S Hr Fr Mullrple Array Disco elc.l/i"vorcecorl Res Freq 38Hz Freq Resp !o20KHz Sefs 89dB pRtCEft299, ft 50p&pea1O" 60 WATT R M S Hi-FilOisco ercl%"vorcecori Res Freq 35Hz Freq Besp tolsKHz Sens 89dB pHlCEft649rf2OOp&p

BSR P295 ELECTBONIC TURNTABLEi E ec ron,c spoldcon rol45& 33,r p nr r P LrsM nls vani )lc prrclr contro a Belt dr ven r A urrr n rirf !l.1ter wrr i stiobed inr i Cue lever r A.lskdlc lbr.s dev cc) * Adlustable counler balance rManoa arr r Srandard , cartr qe frx ngs. rSrrppl eci coilrplere w lh cuL o!t re irp a e * -O 6Operation9l4vDC 65mA

Pnce f36 99 f3 00 P&P

DC 04 maq cartildge for above Pnce f4 99 a :a: :

HOBBY KITS. Proven designs including glass ifibre printed circuit board and hiqh oualitvcomponents complete with rnstrucllonb_ ' :

mony spplietions Rtreiver g0 x 70 x 22mm (9/12 voh) p7ic.; l:f 17 82 Transmitter 80 x 50 x 15mm (9/12 volt) Pricc: f 1 1 29P&P + 75p e8ch. S.A E. frrr complsre list.

STEREO DISCO MIXEB

lQs_Irtl_c-l-]4tqEs pER oRDER f 1.oo minimum. oFFtCtALg! pF]ls jglLg gIvr E, sc H o o LS, c o LL E G Es. c o Vr n iri r,ili'r,t i,!!r B o D r E s, E r c p B r c E b JN-c L u-sli! "o;V'A.J "ddiE;'c?iuiii ii

, VISA ACCESS C,Q,D, ACCEPTFD

ln alSaEte - Concentrution, @ ALPHA - Relaxation,problemsolving,-aclive pleasurc, lrunquilily,lhought.

-

\ Posilive feelings. /

This ill allow you to hotr your own brainwav(prog n to conlrol them and achieve peak pelrevo !The conlains: two PCBS; all components,amo for screening th bio'amplitier; attra(lesi; knobs, plug5, s@kets; leads and braassembly and use

vailable separate havB a range of a@ePlease send a st l-addressed envelopan SAE + e2 will sts, @nstruction deta

THE ALPHAPLANCan you really train yourmind lo think moreeffetively? Can you reallyexcl at things you're'rc good at ?The Alpha Plan hs allthe answrs!t2.fl, (m VAT)

@VYqoru.& G.in

"l',li

fe

7 V,T$ii_H.GSI f.2X"^-

aTHEIA - tnagination,, crealivily, hynagogic

tmaqer)t As leatured in ETl, January 1988 Theapproved parts set contains PCB,case, toroidal cores, class X and Ycapacitors, VDRs, lCs, transislors,I FDs all edmhnnanls end fi rll

TIONER--)

instructions

Parts are available separaPlease send a stamped,self-addressed envelopeit you want the listsOtherwise, an SAE + elwill bring you lists,construction detailsand lurtherinlormation.

For those who willUse with Hi-Fi, cotmains-borne inter10 LED logarithmiremovedMassive filter seclbalanced inductorBank of six VDRsSuppresses comr

Oo

o

o

Oo

te y.

Y;tru

!s! Through a simple traininglormance in all situatios lt will

including thre PMI prsisiontive instrument cas with tihingss electrodes, tull inslructions fol

tsories, professional el*lrodes,r il you just want ihe listsils and lurlher inlormation

Just wipe on with a cloth to PlatePCB tracks, connectors, wire,componnt leads, coins orhosehold ornaments with a layerol purc sllver lt must be magiclLARGE BOTTLE (150m1)SILVER SOLUTION eil.m + VAT!

I oooo

-

only be satistnputers, radioerence

o display shov

ion with thirtet

to remove trarron mode and

@@@

ed with the bests, TV sets to elir

rs the amount ol

)n capacitors an

rsients and spikdifferential moc

ninate

interrerence

d two current-

e interference

II SItiIr l6 iL(r lt )I I I '.J I qtFEATURED IN ETI,MARCH 1988

JUMPIN'JACKFLASHts4.........Lighhngwizard brings

any r@k band s stageperformance to lile!

. Sound operated llash -

pholographbullets in llghtl

. Voice swrtch and sound to actioncontroller wrlh endless applications

The parts set consists ot a high qualilyPCB and all components, los opto isolato(triac, heat sink, pots etc to build the circuitboard What you do next is up to youlThe ETI article, supplied lree with everysel shows how to make the most of J F scapabilities

JUMPING JACK FLASH 86.90 + VAT

THE DREAMMACHINEFEATUREO IN ETI,DECEMBER 1987AdJusl lhe conrols losud your m@d and lellhe gentle relarrngsound dflll overyouAr hrsr you mghr hearsotl rarn. sea sud, orlhe wtnd lhrough dtstanl lrees Almosl hypnolcsound draws you irresistably inlo a Faceful,rekeshrng sleepFor many, the thoughl ol waking rekeshed and aledkom perhaps lhe firsl truly r*flul sleep rn years rsexcrting enough in itsell For more advenlurous soulslhere are skange and mysleflous dream exp(enceswailing Take lucid dreams, lor inslarrce lmage bangrn conlrol ol your dreams and able lo change them alwill lo act out your wrshes and lanlasres, Wilh theDreamMachrne lseasylThe approved pais sel consrsls ol PCB allcomponenls. co oudspeaker. knobs. lampluseholde( luse powr surply preslige caseand lull rnslructrcns

DFEAM MAoHTNE coMpLETE pAFTs sEr C16.50 + varalso AVATLABLE wrrHourCASE FoB oNLy 811,90

' var

POWERFUL AIRIONISERFEATUBED IN ETI,JULY 1986lons have bsn dEcnbdas'vitamiEollh air bylhe hoalth magaines,and have kn crdildw{h everylhi4 Lomcunq haylever andaslhmatoimprcvngMtu dtrmg mdtoin$mnia Allhdghered kffi royberderated,lbre ism(Ml tulM UElddeanerand purei d ffi l'd Ee qdatngthan dead air

TbO|RECIlON6s(:Maglddexcileftnl when il af ed as a ffi trWtinETl Allast, anmrstutw6ry*6(bflrthan?) cotffi gtu e6 r*, godlobuild andlunrApdffi tuclff**atm,soreollh sqgdd exFr16red@s!

U/e can suply a ffi sd Fre, hriyappovd by hdqG. b M tu ry roidThe sel ircludes a r# lIG FEE {d Md,66@mpoMb,@ lftd deghpd6lorlhelosler Adg btr ffi, hg@sls'ahutehtrdolturedhffi@nEWhatmqe @ re sy,OIRECT loll PAFTS SEf lnslructimsWITH BI-ACK CASE El1.5O + VAT areWITH WHITE CAS. Al1.& + VAT rrcluded

AD7Y1 Prftision 12-bit multiplying DAC 8t.m + VAT1M3524 Switch mods 60ulator lC m,m+ VATcFs csldlaror lc tl 0o + vATLMsg GrEd @mparalor lC 3 lor El .m + VATMcl456Dualry-amp 3,or$.m+VAT4 los sHlid winh FREE DATA.ftE apdy sly dils 6l6ks l6st

1M2917EXPERIMENTER SETConsisls ol LM817 lC. 6@d pnntd c(di board anddeleilod rn6lMlonswilhd.la mdcr@its lor eighldlfrqenlprcids bbuild. Cen tE usd b 6xf.imntdh hecrcirils in tu'Nn Gd UUe lC ,eelur6(ETl,Oe@br1S)LM2917 Expeilmenter Set 85.00+vAr

RUGGEDPLASTICsuilau lorm6rG @tuilroffiad mains @d@ll6t

ONLY E1.65 + vrr

LEDsGIs Ect lgarla LEOSfd bd{raph dsdaF50b.e350 5O0b.C25l@bE6 10@he45dcrnrAro^tm EqflPrEr{TAsled 3.m LEOS rsd,gr@n, yellw andoaE 25 ol each (1(x) LEDS) lor e6.8O

3ETI APRIL

Geoff Bains: Editor

Jez Ford: Assistant EditorPaul Chappell: Prolects Editor

Fowler: TechnicalJerry Fowler:

Jeff Hamblin: Design

News

Diary

Fame &Fortune

r0Next Month

llRead/ffi

t7Capacity ForThoughtJohn Linsley Hood takesthe first of two looks atcapacitors - what they are,how they're made and howto use them properly inaudio design.

I Golden SquareLondon WIR 3ABTel:01-437 0626Telex: 881 1896

4

rssN theol42-722g au

on

Argustaken ipublish

publishconrained in adveftisements are accepted by us in good faithas .ofie(t ar rime of going to press Neither the advertisersnor rhe pub le however, for anyvarialions a which may occurafter the pu s,O Subscription rates UK {16.80. Europe:t2 I .00. Middle East: {2 I .20. Far East: f22.80.Rest: {21.50 or U5$32.00.

I I I ustrator

l5T-ShirtGompetition

r6Satel I iteAcolyteKeith Brindley rounds offour extensive look atsatellite TV with an in-depthreview of the top threesystems available.

70Count On ItMike Barwise continues theChip ln series with a look atthe Fairchild 74F525 pro-grammable divider and findsthat under the cloud of com-plex application notes liesan extremely useful chip.

77Special Offer

23CompetitionResultsResults of the SpecialistSemiconductor competitionfrom the January issue.

Ed Davis: PhotographyChristopher Harris: Ad ManagerJulie Capstick: ctassifted SatesAndrew Selwood: copy controtMark Webb: Grorp Editor

73Subscriptions

24Pulling lt AllTogetherPaul Chappell's CircuitTheory series finds somepractical use for all thecomplex numbers andphasor diagram theory ofrecent months. Perhaps themaths is some use after all.

76VirtuosoPowerAmplif ierHi-fi devotee Graham Naltydescribes his power amp toend all power amps. Thismonth he looks at thegeneral principles behindthe design and in detail atthe power supply.

37ClockwiseKeith Brindley kicks off ourFirsf C/oss project series forbeginners to electronicswith a simple but accuratetimer for the kitchen, oranywhere else come tothat.

ETI APRIL 1988

34Special Offer

36Gombo-LockChung Yiu Ko presents aremarkably simple butingenious design for a com-bination lock. Protect yourbelongings behind a wall olprobability.

40TransatlanticTime ZoneGorrectorAndrew Armstrong keePshis cycles intact whilejet-setting around the worldthanks to this essentialtravel aid. It will even tellyouwhich month it is.

47Special Offer

ETI APRIL

Backnumbers Ad lndex

Published by Argus Specialist Publications Ltd., I Golden Square, London W I R3AB. Tel: 0l-437 0626. UK newstrade distribution by SM Distribution Ltd', l6- l8Trinity Gardens, London SW9 8DX. Tel: 0l-274 861 l. Overseas and non-newstrade sales by Magazine Sales Department, I Golden Square, London W I R3AB. Tel: 0l-437 0626. Subscriptions by lnfonet Ltd',5 River Park Estate,Berkhamsted HP4 lHL. Tel: (04427) 76661. US subscriptions by Wise OwlPublications, 43 l4 West 238th Street, Torrance, CA90505 USA. Typesetting andorigination by Multiform Ltd., Cardiff. Printed and bound by Adlard & Son Ltd.'Letchworth. Covers printed by Loxley Brothers Ltd., Sheffield.

43SpectrumGo-processorGraeme Durant describesthe interface between hisco-processor and theSpectrum in this penul-timate article.

48Book Look

48Playback

49OpenGhannel

50

50Photocopies

5lPCB Foils

s3PCB Service

s3Oops!

54Readers' FreeAds

55Glassif iedAds

58

5

ranges to avoid overloading andcosts just t24.50 + VAT.

Contact Beckman, rli|-rl{l NewStreet, Birmingham 82 4L). Tel: 01-643 8899.

f he European Consumer Elec-I tronics Directory has beenpublished by Euromonitor publica-tions. The directory is not simply atrade directory and carries noadvertising. It covers all ECcountries plus Scandinavia.

The directory begins with anoverview of European electronics,examining trends of the first half ofthis decade and giving some fairlygeneralised predictions for thesecond half

-

audio sales to fall,Portugal to go hi-tech and so on.The next 150 pages are the mostuseful

-

Iisting publications, MRcompanies, trade associations andthe main electronics companies ofeach country.

The last section of the book dealswith statistics for each country

-mainly dealing with import, export,sales and production of electricconsumer durables.

It's a useful tome if you get it free,otherwise the [135 might limit itsappeal to PR and MR users. Con-tact Euromonitor, 87-88 TurnmillStreet, London EC1M 5eU. Tel:0L-251 N24.

nline Distribution has an-nounced its new catalogue of

passive, electromechanical andconnector products.

The catalogue has 48 pages withing of photos andprices and postagenents from various

manufacturers including Erg,General Hybrid, Harwin

"na l. -

The catalogue is free from OnlineDistribution, Melbourne House,Kin gsway, Bedford MK42 g AZ. T eli(0234) 2t7ets.

ultimate in surround soundI' _sensation is promised by Sharpin the form of the Optonica SM-

475. convert your living room into anacoustic Wembley or White HartLane, consider that for the tg79that Sharp are asking you couldspend every Saturday until 1993 onthe terraces themselves.

. .

_.Full details from Sharp UK,Thorp Road, Newton Heath,Manchester M10 9BE. Tel:061-2052333.

6 ETI APRII 1988

ew from Dataman of Dor-chester is the Softy 3 Pro-

grammer -

a PROM programmer,PROM editor, ROM and RAMemulator and development systemall crammed into what looks like adesktop calculator.

As a PROM blower the 53 has abuillin library for EPROMs andEEPROMs (25,27,28,87 series andso on) and can blow say an Intel27C256 in under 20 seconds.

It is also upgradable for handlingfuture PROMs, Ioaded via the ROMsocket (bottom left in the photo).

The whole thing can be remotelycontrolled by 2-way RS232 inter-face, with baud-rates up to 9600.

As a memory emulator the 53accesses typically in 100ns, provid-

ing a direct replacement forEPROMs, EEPROMs and RAMsup to 64K

Inside the dinky 7in by 4in casethe CMOS architecture has 4K ofBIOS ROM masked into the micro-

controller, with 8K RAM programarea and a 54K buffer for user data.

The 53 retails at 9495 + VAT.Contact Dataman, LombardHouse, Dorchester DTI lRX. Tel:(0305) 68065.

jBiBigtii

'Talking Books for the HandiI capped would greatly welcomeany volunteers who would be ableto service the Library's specially-designed tape machines or toproduce homemade inductionloops for listeners with hearing aids.

The organisation is based inLondon but operates nationwide.

If you could help please contactThe National Listening Library, 12Lant Street, London SEl 1QH. Tel:0r-407 9477.

7ETI APRIL

stations. The distribution mono-poly simply wouldn't bother orwould make little headway if it did.

The fourth monopoly option is tosplit the industry into regionalboards. Each board would have alocal distribution rnonopoly andwould genrate most of its ownpower, purchasing the remaindercompetitively. The boards wouldhowever favour their own powerstations and competition would beminimal.

This option would create severalbuyers of coal and would enablcoal to be privatised,

The report firmly informs theGovernment that 'it would be agrave mistake to adopt any of thesgoptions'.

The altetnative is competitiveprivatisation. The CEGB is brokenup into separate generating com-panies which compete to supplyelectricity to five Regional Distribu-tion Utilities (RDUs).

Components In Electronics 88 -

March 8-10thIslington Business Centre, London. Conference and exhibition. ContactNutwood Exhibitions on (0705) 264333Senior Management Seminar On Surface Mount

-

March lfthThe Royal Automobile Club, London. Seminar by Coopers and Lybrandfor DTl. Contact Katy Parkinson on 01-831 2858

Status Of Nuclear Fusion Research -

March lTthIEE, London. Lectures by Dr D C Robinson. Contact IEE on 01-240 lg71Video, Audio And Data Recording

-

March 2l-24thUniversity of York. Contact Institute of Electronic and Radio Engineers on0t-240 t87tCADCAM 88

-

March 22-24thNEC, Birmingham. Contact EMAP International Exhibitions on 01-4044844Electro-Optics & Laser UK

-

March 22-Z4thNEC, Birmingham. Exhibition running alongside the Optics-Ecoosa ,ggconference at the Birmingham Metropole Hotel. Contact CahnersExhibitions on 01-891 5051lnternepcon Production Show

-

lvlarch 22-24thNEC, Birmingham. Contact Cahners Exhibitions on 01-981 5051Into 88

-

March 22-24thNEC, Birmingham. Products, advice and services for industrial training.Contact EMAP International Exhibitions on 01-404 4844

Computer Show -

NIarch 22-24thAberdeen Exhibition and Conference Centre. Contact Offshore Con-ferences and Exhibitions on 01-549 5831

r Recruitment Fair -

March 25-26thRainbow Rooms, London. Contact Intro Ltd on (0491) 681010

ing In The Next Generation -

March Zi-27thr(F

This would produce comptitionat every stage except the finalregional distribution, would enablecoal prluatisation, would bring innew management (ten companiesstarting up), should ned minimalrgulation and should enable con-siderable cost savings.

The main drawback with thisoption is that it would take years toimplement, and the Governmentwill not wait that long.

The report therefore puts for-ward what appears to be the onlychoice for a Iast privatisation.

Private Transition to Com-petition (PTC) creates the sameRDUs for distribution but keeps theCEGB as a rrrcnopoly whollyowned by those RDUs.

This new CEGB is debt-financedand so cannot make a profit givingthe RDUs an incentive to gt com.ptitive generation moving as hstas possible, eventually leading tocomplete competitive privatisationover a longer period.

This is a radical reorganisationand will worry a good many people,not least the unions involved. TheTUC is firmly against any form ofprivatisation at all.

For the consumer, prices wouldbe expected to drop but since eachRDU would bea regional monopolythe consumer might suffer corner-cutting and artificially high priceswithout recourse to any regulatorybody.

Nuclear power presents prob-lems that are totally irreconcilablewith privatisation, evn the CPSreport admits that.

And would the ElectricityCouncil ceas to exist or becomean advertising agency for the com.bined industry?

The industry and its dependantsnervously await the PrimeMinister's decision.

'Current Choices -

Good andBad Ways to Privatise Electricity'isavailable fot 84.90 from CPS, BWilfred Street, London SWlE6PL.

privatisation of the electricityI industry should be abandonedrather than create a BritishTelecom/Gas style privatemonopoly, concludes the Govem-ment's own report from the Centrefor Policy Studies.

The report (entitled 'CurrentChoices') examines in detail sixmethods of privatising electricityand consigns four ol these to thewaste bin.

The first is the option urged bythe CEGB, to privatise the wholeindustry as a single entity subject tominimal regulation, as with BritishGas.

The report rejects this as failingto introduce any form of com-petition, incompatible with thecompetitive privatisation of BritishCoal and as unlikely to drawcommercially-orientated manage-ment into the industry.

The second option is to keep thesupply industry whole brit allowcompetition in the building and run-ning of new power stations.

The report rightly suggests thatfew investors would relish theptospect of putting f, 1%billion into a180OMW station that would take atleast seven years to build, only toIace competition from the still-existing and massively entrenchedCEGB.

Effectively then, this secondoption is the same as the first.

Next to hit the binliner is theproposal from the ElectricityCouncil. This involves creating twoseparate monopo.lies, one for dis-tribution and one for generation.

The distributing monopoly could(if it wished) compete in powergeneration

However the report points outthat the new CEGB would hold allthe expertise and experience ofbuilding and operating power

8

Annual conference of the British Computer Society's young professionalsGroup. Contact Julia Allen on 01-637 O47l for the venue

ters In Retail & Retail Technology Exhibition -

March 29-3lstMetropole Exhibition Centre, Brighton. Contact Focus Events on 01-g327717HF Radio Systems And Techniques

-

April 1l-l3thThe IEE, London. Conference organised by the IEE and The Institute ofMathematics and its Applications. Contact IEE on 01-240 1g71

Nuclear Power -

April 12thScarborough Lecture Theatre, University of Durham. Lecture by CSmitton. Contact IEEIE on 01-836 3357Scottish Computer Show

-

April 12-14rhScottish Exhibition Centre, Glasgow. Contact Cahners Exhibitions on 01-891 5051

on (0491) 681010

1010

Heathrow Penta Hotel, London. Contact Softsel Computer products on01.568 8865

TE 1988 (Automatic Testing & Test Instrumentation) _ April 26_Metropole Exhibition Centre, Brighton. Contact Network Events on

815 226sh Electronics lVeek

-

April 25-28thOlympia, London. Contact Anne Jackson on (0799\ 2669B

Second Power Sources And Supplies Conlerence -

Apri! Z6-2glhOlympia, London (at British Electronics Week). Contact AnnL Jackson on(07e9) 26699

tronics And The Stock Exchange -

April 28thIEE, London. Lecture bv D C Marlborough. Contact IEE on 0l-240 1g71

tronics And The Space Program -

May 4thLondon. Lecture by J Egan. Contact IEE on 01-240 1871

l9th88 (Technology In Tourism & Leisure Exhibition)

-

May l7-Business Design Centre, London. Contact PLF Communications on(0733) 60535

omputer North -

May 24-26thG-Mex Exhibition Centre, Manchester. Contact Cahners Exhibitions on

l-891 5051ineering Products And Technology North

-

May 2S-Z6thribition and Conference Centre, Doncaster. Contact Trinity exhibitions(08e5)58(}1

mation Technology And Office Systems Exhibition -

June T-

Exhibition Centre, London. Contact BED Exhibitions on (0932g

ETI APRIL 1988

mAN/flm,

That's what an incrediarticles to ETI. The resmoneY for a cou,le of

ions now!PROJECTSETI has built its reputation on novel, worthwhile

designed andave recently dee want to hear

first instance send us a brief description of yourmasterpiece along with a circuit diagram.

Whatever you can contribute to ETI, take the plunge now.We can offer a modicum of fame and a very reasonablefortune.

Write in to:The Editor

ETI1 Golden Square

London W1R 3AB

If you know what you're talking about and ithasn't all been said before, we want you to addto our wide ranging and informative features. Ifyou have a great idea for a feature or two, sendin a brief resum6. If you don't have the ideas butyou think you have a commanding knowledge ofa suitable subject area we want to hear from youtoo.

ruEUY FROM SAGE AUDIO - A OUANTUM PLUS LEAP FOR HI-FI IN SONIC AND TECHNICAL PERFORMANCEBuild your own 'Suprfi'class A syslom, superior to oll the big namos in Hi-Fi today

THE WORLDS HIGHEST TECHNICALPERFORMANCE, COMBINED WITH . s u P E R M O S 2':i'H'f,rHsT:i5'J3fiiBti".j"r'iffLor,aporuerurs

Po lOOW to 45OW (Dependant upon chosen PSL VoltageO/P PowerDistortionSlewrateFreq respO/P CurentO/P lmpedance

1 00W to 450W (+/-50V to +/-72V DC)0 0001% (1 ppm) A World firsll685/us (Zero transient compression)

All semiconductors ars selctd, prsision matched P/N types,Custom msd9 matched MOSFETS, completely symmetrical designHOLCO resistors for non-critical areas and tor critical areas we usethe very bost lazer trimmed SMD resistore, supr io bulk foils with 5times lower inductance, Minimal pacitor design using xtanded,oil and polypropolene types.

0,5H2 to 350kH2 -3dB (lin8oamps and total reactan ble0.0085ohms(dampingfa 80) M6NSTER;NTEGBA;HEATS;N1stzE24orlo0r,loomm

THE MOST ADVANCED POWER AMPLIFIER IN THE WORLDUltra linear 7Olo efficiency. PURE CLASSA Thtoughout

Don't buy another power amplifier until you've checked ALL these lactsl Take the guess work/risk out of power amp construction, our modules are ready built,tested, guaranteed and come with a signed certificate of performance comformity.

Some manufacturers build technically excellent amplifiers whilst turning a deaf ear to component sonic quality.Other manufacturers use audiograde components whilst totally ignoring BASIC technical performance.

THE SAGE AUDIO DESIGN PHILOSOPHY _ A UMAUE COMBINA|IONa ADDITIONAL SAGE EYCLUSIVE (World firstl TECHNICAL DESIGN FEATURES ENABLE THE SAGE SUPERMOS 2 T0 ACHIEVE A NEW HIGH LEVEL lN AUDIBLEOsorurc pERFoRMANcE urunanrcireo ev lrirv orxen powER AMpLTFTEB, EITHER MoDULE, Krr, BoARD oR coMHERcrALry MADE AMpuFtERs.1 Activg Class A Our unique MOSFET pure class A O,/P stage eliminates crossover distortion generated by cla$ AB stsgs along with all thermal tracking problems, BJT $orage

effects and will not switch to class B under high drive as conventional low eJficiency class A amps will2 Zerc THD All semiconductors are exclusive matched selected pairs operating in patented constant current/@nstant VCE for totally distortionless linear operation

ultre high linear symmetrical slewrate some 50 iimes the normal eliminating TIO altogether.

the 'Sound o, the PSU comps'5 Pfotgction Conventional Vl limiters which introduce distortion when driving real speakers have been eliminated The massive oulput stage can wilhstand a S/C this ensures no

sund degradalion whilst driving up to totally reactive loads with zero resistance6 BalanCed CifCUitry Symmetrical design gives inherent stability and lineararity eliminating the need ror DC *rvos which degrade LF disiortionAND THERE'S MORE . . .SONIC COLOURATIONS'ELIMINATION' ln addition, byColourations' associated with conventional amplif iers:-

and innovative engineering we have ELIMINATED 5 of the major causes oI 'Sound

yohage is allowed to appear across them ror the lowest possible source of'CAPACITOR SOUND'A N D STI LL M O R E . . . We can't possibly describe this unique Super Hi-Fi a mplif ier f ully in this ad, to receive a full size 8 page glossy brochure giving f ulltechnical details send f 1 and a large SAE (6 IRC) Supermos 2 modules f 14O plus f2 9/p ler module.

The original Supermos 50-1 50W f65 plus f1.50 p&p per unit. User manual f2.5O refunded on purchase.SPECIALOFFERS GradelaudioPSUcapsbrandnewMullardl'l422,OOOI63Vf18each,36amp6OOVbridgerecsf450,MOSFET'S10Al7OVT03Pmatchedpairsfl1/pai,MJ1 1 01 5/1 6 darlinglons, brand new matched pairs f8,/pair APRIL PSU KIT OFFER - 4x22,OOOu 63V CAPS, 2x36amp bridges f68 (normally f81 | or only f6O if ordered with twoSupermos 1 or SuDermos 2 modules Oflet for this month only

9ETI APRIL

Sales Construction House, Street, West Yorks. BD16 4JH, England.

Next month's ETI features even more ingenious projects andinstructive features. The comprehensive news section keeps you upto date with the best new products as wellas with the movements ofthe world and its stomach. As Napoleon said - "Not tonightJosephine, I'm trying to get this co-processor going".

It was in the early 1930s that ETI exploded onto the newstands ofdepression-ravaged Britain, although it only sold its first copy inl972.That copy was bought by a Mr. Sinclair of Cambridge and he'snever looked back. "l laughed so much I crashed my C5 into theHoover factory" he said, though no one bothered taking muchnotice.

Other world leaders say:"l cried until I got an Oscar" - Sally Fields."l'm glasnost'vailable in Russia" - Gorbachov & Madivar.

ETI is read in 44 different countries, and all by Mr. G. Smith ofCroydon. "Next month I hope to read it in Malawi" he said in anexclusive satellite linkup with Wofch with Mother'lbut I'm havingtrouble getting a copy."ELECTRONICS

TODAY INTERNATIONAL

Moke 9th April your Moy Doy with ETIThe artlclct llrted are planncd tor tha nGxt i3rcG but m.y not appcar duG to

WRITEhen studying the ETI Concept(October 1987) and the Heat-ing Management SystemDecember 1987) I noticed theof 1N4148 diodes across relay

coils.I consider this to be fundamentally incorrect

as this type of diode is not suitable for this type ofservice.

I have experienced failures (about 100) doinga similar function. I think designers have forgottenone of the fundamentals. When using transistordrives, especially on gates (lCs), their switchingtimes can be extremely fast.

If I remember my theory correctly, the peakinverse voltage (PIV1=6;76,. So with a coil current(di) of say 0.06A and a switching time of say 100nsthen PIV=600000V. Only 1N4007 diodes areappropriate.

Finally, the article Home CADMAT (October1987) was rather irritating because of the relativelyhigh cost of commercial CAD packages andquality assemblers for the BBC micro or IBM PCcompatibles. Has any thought been given topublishing a PCB design program in Basic?

R A J HowardNew Farnley, Leeds

Nice try but although your formula is correct,it's the wrong formula. PIV= di/dt only appliesto circuits utithout the diode. The diode isplaced across the relay coil to prevent justthis voltage build up. Withthe diode, inversevoltage is limited to 0.7V

-

the forwardvoltage across a diode.

The reason your diodes were failing wasprobably due to excessive instantaneouscurrent

-

not voltage.To solve fhot problem an 1N4007 would

be a good idea although these devices may betoo slow to conduct to protect the drivercircuit.

The idea of Home CADMAT was to writeprograms and build equipment of your ownrather than rely on the expensive commercialvariety. However, we shall look into thepossibility of a PCB design program in Basic.Any readers care to contribute?

t last! Other ETI readers are reveal-ing their true colours by coming outin favour of Doctor Who (Read/Write-, January 1988).I too was pleased to see the

return of the good Doctor to the realms of our FSTTrinitrons and equally appalled to see him stillaccompanied by Bonnie Langford (good thoughshe is at screaming).

The good news is she has left the series. Thebad news is the series has ended (three stories

-

aseries?!)

Come on BBC let's have less of Anne Thingyard Points Of View and some moreDoctorWho,immediately.

John 'Cyberman' GrahamGallafrey.

ETI APRIL

Yes, yes, but why tell us? Shouldn't you bewriting to the BBC (even'Anne Thingy') andheckling them?

By the way, how can impoverished ETIreaders afford FST Trinitrons?

'm glad to see you're taking a look at theworld of satellite television. This has longbeen a subject which has interested me.

However, I feel your introductory look atSTV (January 1988) was short of one

important aspect of the installation -

planningpermission.

Planning permission is required for any dish ofmore than 90cm diameter. Different localauthorities take different views on the obtrusivenature of satellite antenna but all tend to take eonsto express their views in the form of permission!

It can take months for planning permission togo through and stories of over a year are notuncommon. Some local authorities insist the dishis disguised with shrubs or even painted green!

So be warned, before you part with thereadies, you have the local bureaucracy tocontend with.

Mark PinnaWalkley, Sheffield.

Too true. It does seem ridiculous thatyou canput up a 15 foot greenhouse and paint it pinkand the council won't (and can't) do a thingabout it but put up a 4 foot aerial hiddenbehind your garden fence and they goberserk.

As the government is trying to encouragesatellite TV for the future (in the form of DBS)it is ridiculous it is now discouraging the fore-running service. Write to your local MP andlocal councillor.

n the August to December 1986 issues ofETI you published a series of articles on anIntelligent Call Meter project to automatic-ally log the cost of telephone calls.

The project began with a warning whichsaid that although the meter was not BABTapproved 'it can be used as the basis of a sub-mission for approval.'

Have any readers submitted one of thesedevices to BT and was it approved. If it wasrejected

-

on what grounds?S GreenSmethwick, West Midlands.

Devices can be submitted to the BABT eitherfor'one off approval or for approval of design(which includes the entire manufacturingprocess). So, any reader submitting the ICMfor approval would only obtain approval fortheir own particular model. Only if they gointo production (which they cannot withoutthe permission of the author and ETI anyway)can a generalised approval for the ICM begranted, and then only for those produced bythe approved method of production.

ln other words, you're on your own!

r-,frlFlFlrIEE"rl

iilu,it

i,I

.,

!IF

I

I

CAPACITY FORTHOUGHT

(b)Fig. 1 (a) Circuit symbol (b) Accuraterepresentation of a capacitor

(e) (b)Fig. 2 Physical construction (alwithout dielectric(b) with dielectric

position, how much space you can spare, andhowlittle bothered you are about wasting money.

As for polypropylene (the current favourite ofthe 'golden-eared' fraternity) the questions I wouldask are 'what type, how made and by whom, andhow used?' So, let us look at some technicalities.

Normally in circuit diagrams the circuitsymbol shown in Fig. 1a is used to depict acapacitor, but in reality it is more accurately repre-sented by the drawing of Fis. 1b, where 'C' is thecapacitance at some specified frequency, tem-perature and applied voltage, 'Rr' is the leakageresistance across the capacitor (which again maybe temperature, humidity, frequency and appliedvoltage dependent), 'Rr' is the equivalent seriesresistance due to dielectric Ioss (again not aconstant factor), 'R.' is the straightforward seriesresistance due to its method of manufacture, andfinally 'L' is the inevitable inductance of thecomponent.

Physical constructionIn principle, a capacitor is a pair of conductors inproximity to each other but not in electricalcontact, such as a pair of parallel conducting platesin a vacuum (as shown in Fig. 2a).

When an electrical potential is appliedbetween these plates, electrons will flow into thenegatively connected plate from the negative poleof the applied potential. An equivalent number ofelectrons will be repelled away from the oppositeplate and will flow towards the positive pole of theapplied potential. If there is some circuit resistancethis will lead to the familiar charging current stageshown in Fig. 3.

If the potential is removed and the wires fromthe capacitor are shorted together the sameprocess will happen in reverse, so the wires willprobably spark as they touch since there is now nolonger any reason for the asymmetry of charge onthe plates.

The capacitor LS amuch used butoftenmisunderstoodcomponent. JohnLinsley Hoodturns your doubtsto dust with a two-part looh, at allthings capacitiue

here is an old joke that a metallurgist issomeone who, given a choice ofmaterials, chooses wood . . . The point,I suppose, being that any specialist whoknows the snags inherent in his chosen

speciality is likely to be more enthusiastic aboutthe potential use of something else.

This is basically how I feel about capacitors.For some years I was involved in the manu-

facture of the polypropylene film used in makingcapacitors, responsible for the electricalevaluation of our own and competitive films ofvarious types to see how well they would perform.

This was quite an interesting project andinvolved visits to a large number of capacitormanufacturing companies to discuss the use ofpolypropylene and other films in this particularfield.

I don't think that this makes me a capacitorspecialist, but at least I have had a rather closeracquaintance with this topic than is normal forelectronics engineers. I know a lot of the unpub-licised problems.

So Say The Hi-Fi BuffsQuite a lot has been written in recent years in the'Hi-Fi' and electronics press about the differencesin sound qua.lity which can be brought about bychanges in the type of the 'passive' componentsused in the audio system, whether these beresistors, capacitors, connecting cables, mainstransformers, printed circuit board materials,solder, or even the fixing screws with which thecases are held together.

With most of these claims technicallyplausible explanations for the observed el{ects areusually only remarkable by their absence.

The tests on which they are based are alsoinevitably subjective in their nature and rely onlistening trials which, however extensive, canseldom be conducted on an instantaneous'A vs B'switch-over comparison. Where any length of timeelapses between two alternatives, the memorybecomes clouded and expectations begin tocolour the observations.

There may be basis for the claims, though Ifeel that these are often exaggerated or incorrectlyinterpreted by their discoverers

-

like the changein sound quality (sometimes even for the better,since it lessens 'crossover' distortion) whichhappens when an amplifier having a poor stabilitymargin is caused to oscillate at some ultrasonicfrequency by the unwise connection of high self-capacitance LS leads. I remain agnostic.

Nevertheless, in the case of capacitors andparticularly in the case of those used in thefeedback loop of an amplifier using negativefeedback (NFB), I feel that a good case can bemade for care in their choice, since there areeffects which are capable of being measuredinstrumentally as well as being heard.

But there is no blanket answer to the questionof which capacitor do I use

-

it will depend onwhere you want to use it, what are the particularqualities which are especially needed in that

e,-f,TffiffiFdF-{r*l

TJETI APRIL r988

The theoretical value of such a capacitor(ignoring the effects of fringe fields at the edges ofthe plates) is given by the formula:

C = AK/11.315d (pF)where A is the effective opposed area of the plates,K is the dielectric constant of the material separat-ing the plates (=1 for vacuum or air), and d is thegap separating them

-

all dimensions being incentimetres.

The practical problems of such a constructionare due to the need to prevent the plates fromtouching and the difficulty of getting any largeamount of capacitance.

These can be solved if some insulatingmaterial is fitted into the gap, as I have shown inFig. 2b. If this is thin and has a good electricalstrength, the gap d between the plates can bemade very small which increases the capacitancefor a given elfective plate area (see the formulaabove).

Dielectric For DivisionThe capacitance will also be increased because thedielectric constant K of the insulating material willbe greater than unity.

This comes about because all such insulatingmaterials will 'polarise' to some extent, either bythe displacement of orbital electron cloudssurrounding the atoms of the constituent material,or by the migration of ions, or by the physicalreorientation of polar molecules.

This has the effect of producing equal butopposing charges on the surface of the insulatorfacing the capacitor plates (Fig. 2b) which lessensthe effective spacing between the plates.

Unfortunately, the introduction of a dielectricbrings the problem of leakage (though this isn'tsuch a problem with modern materials as it waswith the old waxed paper insulated 'tar babies' ofmy early years in electronics!)

The insulation may break down electricallythough there are techniques tor reducing thishazard. The dielectric constant may not beconstant

-

certainly it will decrease with appliedfrequency and will also be affected to a lesserextent by temperature and applied voltage.

Finally the dielectric introduces'dielectricloss' which is represented by the term'Rr.'in Fig.1b. This comes about (understandably) becausethe migrations of electrons or ions or the molecularreorientations (which produce the eflect shown inFiS. 2b, and which cause the increase incapacitance) all absorb some energy when theyoccur, which is every time the applied electricalfield is reversed.

The more frequently the polarity of theapplied electric field is reversed (the higher theoperating frequency) the higher the loss. Materialssuch as the largely non-polar plastics (polye-thylene, polypropylene, PTFE, and polystyrene)don't have very high dielectric constants

-

whichdoesn't help very much to make compact highvalue capacitors. On the other hand very littlehappens when the field is reversed, so thedizlectric loss is very low and the dielectricconstant K doesn't alter significantly with:requency (up to the GHz range).

The thinking of the hi-fi purists is largelycoloured by considerations of dielectric loss, and'9p' is reputed to be very low and therefore verygrcd. However, the actual loss factor depends onjre purity of the material, on the way in which it is:nade (including additives included to assist in-=roduction and the extrusion temperature). I have

ETI APRIL

to TIM

Fig. 3 Typical capacitive charging current pulse

(b)Fig. 4 Foillfilm capacitors (a) construction (b)contacts

nEr-lHFlEIJEn

listed the major qualities of the most commondielectric materials in Table 1, but as I haveindicated these figures can only serve as a guide.

Non-polar ManufactureGenerally, plastics film insulated capacitors areeither of the film/foil type, or of the metallised filmconstruction. In the'F7/F'type, two long lengths ofaluminium foil (which should be scrupulously cleanand of high purity if the loss factor of the capacitoris not to be worsened) are sandwiched between apair of slightly longer strips of plastics film and thewhole thing is wound up in 'swiss roll'form, asshown in Fig. 4a.

Usually the foils are arranged so they extend abit beyond the edges of the film strips so thatelectrical end contacts can be made to them as

r-flTfY{HffiF*d$"**{CJ

LJ

shown in Fig. 4b. Sometimes (as is usually the casewith the small polystyrene capacitors) the foilsdon't overlap the Iilm but a pair of connecting wiresis simply trapped in the spiral while it is beingwound.

With larger capacitors it is helpful to make acontinuous edge contact since this lessens thespurious inductance value, because of the'shortedturn'effect. It also helps keep the electrical resis-tance of the plates low.

In all film/foil capacitors the electrical strengthand consequently the thickness of the film must begreat enough to prevent any possibility of electricalbreakdown at the rated working voltage. Suchcapacitors therefore tend to be bulky for a givencapacitance value.

In the case of the metallised film (MF) types,the problem of possible electrical breakdown issolved by using a very thin metallic conductinglayer, vacuum evaporated onto the surface of thefilm so that it leaves a clear strip along eachalternate edge. End contacts are then made byspraying a solderable metallic layer onto each endof the sandwich. Such MF capacitors will 'self heal'in that if there is a local breakdown of the dielectric,the instantaneous discharge of the stored elec-trical energy through the puncture will burn off themetallised layer in that region.

Such internal flash-overs cause a gradualworsening of the loss factor because of theaccumulation of combustion products in thewindings. They also cause a gradual decrease incapacitance. Both of these problems are lessenedsignificantly by not running the capacitor at morethan half of its rated working voltage.

The major problem with'MF'types howeveris that the metallised layer is so thin and has asignificant winding resistance Rs which cannot bedistinguished electrically from dielectric loss. Onthe other hand they are very small in size.

There has recenily been an increased avail-ability of stacked foil capacitors, a number ofpostage stamp sized pieces of film with eithermetallised layer or foil plates assembled into asmall rectangular stack, and then resin encap-sulated with projecting radial connection leads asshown in Fig. 5. These have the advantage of lowseries inductance and compact PCB assembly, butare otherwise similar in characteristics to the spiralwound versions.

Tantalum And AluminiumElectrolyticsIn these capacitor types, a large value ofcapacitance is obtained by chemically growing avery thin insulating oxide film on the surface of anetched metal plate or a pellet of sintered metalpowder, with a conducting electrolyte occupyingthe gap between this and the other plate. Thisavoids the problem of electrical failure throughbreakdown of the insulating layer because if thereis a puncture in the oxide film it is promptlyrepaired by local electrolytic action between theexposed metal and the electrolyte.

The snag is that this action is going on all thetime, with continuous small pulses of currentevened out by the capacitor itself into a fairlysmooth current flow. The electrolyte though quitea good conductor is not as good as a layer of metal,which is why the non-polar capacitors always havea lower series resistance value. The otherproblems are that the value of the series resistanceis dependent on voltage, temperature andfrequency, as is the capacitance itself.

Also the polarity of the capacitor must beobserved, and if any AC potential is likely toappear across it there must always be a continuousDC bias voltage which is greater than this. Thismeans that electrolytics are not very happily usedwith zero polarising potentials.

When 'tantalum bead' (sintered tantalumpellet, resin encapsulated) electrolytics firstappeared they were greeted with great enthusiasmsince they had a lot of factors in their favour. Thetantalum oxide dielectric was electrically andchemically very strong, and it had a much higherdielectric constant than alumina. This meant that amuch more acidic electrolyte could be used givinglower series resistance, and more capacitancecould be packed into a small volume.

In addition because of the strength of theoxide layer, the capacitor would even stand a small(0.5 lV) reverse potential which permitted use insignal lines. Unfortunately the instantaneous(though small) voltage dependence of conductivityleads to a complex behaviour pattern on transientvoltage steps, and this can give a rather'dull'soundwhen used as the blocking capacitor in a feedbackline.

The increase in the cost of tantalum beadcapacitors has stimulated research work on theiraluminium equivalents with the result thatphysically small high-capacitance aluminium typesare now available which are much to bepreferredin audio use such as DC blocking in signal or leed-back lines if high capacitance values are essential(though their quiescent working potentials mustbe carefully chosen).

Even so, non-polar types should always be thefirst choice, except in routine supply linedecoupling duty.Permanent PolarisationThis is the electrostatic equivalent of permanentmagnetisation and is a snag which is exclusive tothe plastics film dielectric types of capacitor.

As in steels, the durability of such apermanent polarisation is a function of the hard-ness of the material. It occurs much more readily inthose films which are biaxially stretched duringmanufacture such as polypropylene or polyester(PETP) since this greatly increases theirmechanical strength.

Those films which are made by casting from alacquer (such as polystyrene, polycarbonate, orpolysulphone) or by sintering a powder (such asPTFE) are much more limp physically and muchless prone to this defect which can have the effectof building in a permanent series potential withinthe capacitor dielectric.Circuit ApplicationsNext month I shall look in detail at the specificrequirements of audio circuits and how best to fulfil

Fig. 5 Stacked foil capacitor

14

them.

ETI APRIL 1988

!T'S NEW, IT'S HERE AND ONLY AVAILABLE THROUGH ETIMICRO MINIATURE SURFACE MOUNT PROJECTS

You've heord obout them, probobly seen ondmorvelled of them inside your colculotor or wotchond now you con begin using them with the lotestronge of surfoce mounted component kits fromSUMA DESIGNS. All the kits ore of well proven, in-house design ond come completewith oll compon-ents individuolly pocked for eosy identificotion, highquolity fibregloss, surfoce mount, solder resist P,C,B.,full ossembly ond circuit instructions, All you need is opoir of tweezers, o fine-tipped iron ond o steodyhond. Using osembly techniques develo@ tocope with our own prototype ond botch monufoc-ture you con be ossured of success, To moke thingseven eosier we hove developed o custom modecomponent jig which olthough not esentiol greotlysimplifies ossembly ond is well worth the investment,A full ronge of SM components ore olso strcked foryour own projects,

SM20: SOUND TRIGGERED $TVIICHT t13.9tlriggecfrom slightest sound, LE.D. indicotion,vorloble sensitivily

SM21: DIGITAL DICE: 9'15.9:Touch control, roll dNn speed SM L.E.D,reodout

SM22: VHF FM TRANSMIffiR: 9't2.9tComplete with mic, tunmble over 88-108bond

SM23: LI\IE WIRE DETECTOR: i,9 9{Non-contoct indicoiion ot llve wkes, blownfuses tc,

SM24: trHF FM BLEEPER; t{3.9{Tronsmits constont blep to neorby FMrodio Tuneoble

Clllz: Custom designed ossembly jig for speedy t6.9(ond rccurqle ossmbly

Send 9x4 S.AE, forfulldetoilsof these ond othekits plus oll components qvoiloble,All prices ore inclusive but pleose odd e 1.50 peorder to cover post & pocking.

i4 N *',1;ft"*cr barterre.q

T.SHI TCOMPETITION

Back in the halcyon days of 1980 ETI had a pretty nifty line inT-shirts for sale. Although those garments are long gone, theunclothed masses at the ETI editorial office want to resurrect thetradition and once again see the ETI logo across the chests ofmillions.

The only problem is we have no design (well, we wouldn't want touse the old one again would we). Since we're not only unclothedbut overworked too, we haven't the time (well, OK, we haven't theimagination either) to come up with a good design. That's whereyou come in.

We are giving away three ETI T-shirts of the winning design and ayear's subscription of your favourite mag (this one, silly!) to thecreator of the best design for the ETI apparel, The design shouldbe in just one colour, incorporate the ETI logo (as on the cover ofthis issue) and be suitable to fit on the front of a T-shirt. Althoughhumorous designs would be appreciated, please bear in mind it isillegal to send obscene material through the post!

Get your grey cells working on the problem, draw the design on alarge T-shirt shaped blank (like the coupon) and send it with acompleted coupon to arrive by 15th April to:

n

\\\\\I___

ETI APRIL 15

SATELLITEACOLYTE

here are so many satellite televisionreceivers around (see February ETI),with so many different features you'dbe forgiven if you are confused by allthe relevant specifications. It's for this

reason that the final part of this series on STVconcentrates on only three receivers, with theintention of an objective comparison

-

although,as you'll see, comparison isn't easy. The threereceivers looked at have been chosen simplybecause they seem to provide the widest range offeatures and represent some of the best availablesystems.

They are all 'future-proof' (a term used bymany manufacturers to show that their productscan be up-graded with add-on boxes to receivescrambled signals and MAC signals). Otherequally high quality receivers exist, however.

The three receivers reviewed are theRediffusion RSR50, the Salora SRV1150 (both wellknown and highly acclaimed) and the newcomerPace SR640.

Readers of the many satellite televisionmagazines will already be familiar with theRediffusion and Salora receivers but the Pace

receiver (designed and built by Pace Micro Tech-nology

-

perhaps better known for modems) willbe a stranger. Nevertheless the Pace receiver, asyou will see from the following, is a receiver toreckon with.

First, a brief description of the three receivers,listing a few relevant specifications and highlightingtheir quirks.

Rediffusion RSR50Last year the RSR50 took the STV fraternity bystorm. It was the first receiver which allowed theuser to fully store all relevant information aboutindividual satellite television channels.

The idea had been around for a while oIcourse, with some receivers able to store infor-mation regarding frequency, polarisation, audiosub-carrier frequency and skew, such that thetouch of a single button (or at worst, two or threebuttons) on the remote control handset wouldchange the channel displayed on the television.

What distanced the RSR50 from the rest isthat it allows the user to store antenna positionrelevant to particular channels too.

When channel changing from one satellite toanother, users of other receivers must first adjustantenna position as well as changing to therequired channel. RSR50 users merely have topress a single button

-

the receiver takes care of itall. This feature caught the eyes of the STV pressand prompted one magazine to call the RSR50 the'one to beat in 1987.'

The antenna position-controlling factormeans that, whereas most other receivers requirea separate antenna controller, the RediffusionRSR50 is effectively an integral system.

I say 'effectively' because the receiver is not,correctly speaking, a stand-alone unit

-

aseparate power supply is required, containing thetransformer and relays required to control theantenna actuator. This power supply is merely forthe antenna though, and the receiver still incor-porates its own internal PSU, which means it conbe used alone in a single-satellite installation(without an antenna actuator).

Without all the bulky antenna-controllirig

Keith Brindleyrounds off his looh,at satellite TVwith a detailedreuiew of hisfauourite threesystems

F{Erl

16 ETI APRIL 1988

electrics, the RSR50 has quite a low profile incomparison with most other receivers and is ofsuch a width, depth and colour as to be extremelyhandsome sitting on top of your video cassetterecorder underneath your telly.

Apart from its good looks the RSR50 also hassome specifications worthy of note. First it has areasonable carrier-to-noise threshold of only 8dB-

so even fairly weak signals willproduce pictures.Its IF bandwidth of 30MHz allows a good pictureresolution for most channels (only Superchannel,Teleclub and RAI-Uno have higher bandwidths

-35MHz -

than this).For transmissions in stereo sound the RSR50

has stereo audible outputs allowing direct connec-tion to a hi-fi system. Video outputs allow similardirect connections to video recorders or TVmonitors for higher quality pictures than the UHFoutput affords.

Whereas most other receivers accommodatedescramblers or MAC decoders simply byfeaturing a baseband output, the RSR50 also has adescrambler'loop-through' socket. Whendescramblers are available, only one inter'connection will be required, so installation will be adoddle.

The remote control handset is a delight to use.Single-handed operation (either hand) is possiblesimply by movingyour thumb around to reach onlya few logically placed buttons. This is the result oftwo things incorporated into the receiver; first, thestored details for antenna position, channelfrequency, polarisation and so on which mean thata sing[e button press changes received channel.Second, the receiver has two main operatingmodes, the handset buttons meaning differentthings in each.

Main mode is called memory mode. In this,channels are selected simply by pressing anumbered button

-

Premiere may be pro-grammed into channel 1, so pressing buttonnumber 1 selects Premiere. Logical, huh? Up to 50channels can be programmed into the receiver andas there are only ten numbered buttons, channelsabove number 9 require two button presses. Inboth modes, other buttons on the handset allowcontrol over antenna position, polarity, skew andaudio mute.

The other mode is search mode, entered fromthe handset by pressing the mode button thenbutton number 1. In search, the receiver createsan on-screen graphic displaying the frequency ofthe received transmission. Handset buttonsmarked channel- and channel+ allow tuningof thisfrequency up or down. The numbered buttons canalso be used to jump to a desired frequency (thelFfrequency range of 950-1750MH2 is split into eighty10MHz steps). Once a channel is located, and allrelevant details have been selected, it can bea.llocated a channel number by pressing the storebutton (inset into the handset) followed by thedesired number.

Once back in memory mode (accessed bypressing the mode button twice) that channelcomplete with relevant antenna position can beselected just by pressing its number.

In operation, there are few things to complainabout. A great deal of thought has obviously goneinto the receiver in a successful attempt to make ituser-friendly, resulting in the two operating modes.Single-button channel selection is a sheer delightand I'm sure many other manufacturers will berushing to adapt this feature into their products tocompete. On-screen graphics are clear andprecise and remain synchronised even when no

ETI APRIL

signal is picked up, say when the antenna istracking between satellites.

The only real niggle with the receiver was thelack of any facility for accurately metering signalstrength

-

useful when installing and setting upthe antenna. The integral bar-graph signal leveldisplay isn't sufficiently accurate.

Salora SRVI1SOThis receiver has been around for a little while andas such you might be forgiven for thinking thattechnology has overtaken it. However, quality-wise you'd be wrong to pass it over. Whereas theRediffusion RSR50 control-antenna-positionsdirectly, a Salora system actually comprisesreceiver and a separate antenna control unit (ACU1160)

-

this makes a 'feature-by-feature' com-parison a little tricky.

In overall specification terms, the Salorasystem is every bit as good as anything else around.Threshold is 9dB (not quite as good as theRediffusion RSR50 and a lot poorer than the PACESR640). Onthe plusside, an IF bandwidthof 32MHzmeans that picture resolution is theoretically muchbetter. The argument goes that a picture formedfrom a signal with a carrier-to-noise ratio less thanabout 9dB isn'tworthwatching, so youmightas wellhave a good picture resolution when a strongersignal is available.

The receiver itself is smartly designed but thisis slightly offset by the presence of the antennacontrol unit which, although the same colour,doesn't quite match in style. Back panel connec-tions for video, sound and baseband outputs areprovided for direct connection to video cassetterecorder or monitor.

Operation of the receiver is straightforwardenough. Front panel buttons for tuning (+ and -),sound (one of four preset subcarriers), channelstep and store are the only controls provided andonce the antenna has been pointed at a satellite,it's a simple job to locate channels and store themready for watching

-

much as you'd tune in a UHFtelevision tuner. Setting up the antenna controlunit to store details regarding antenna position,channel polarity and skew,'is somewhat moreinvolved. Nevertheless, lengthy instructions areprovided, albeit translated

-

presumably fromFinnish. Satellites are number from A1 to A6, 81 to86 through to E6

-

a total of 30 satellite positions.One of the trickiest concepts to come to

terms with is the fact you have to allocate differentpolarity signals into set channel nurnbers. Thisdepends on a rear panel switch setting. Either oddchannel numbers all cause the polarotor to be

17

F{E.n

vertically polarised and even channel numberscause the polarotor to be horizontally polarised orchannels 1 to 16 switch the polarotor for verticalpolarisation and channels L7 to 32 switch it forhorizontal polarisation.

A compact remote control handset is pro-vided which takes a while to get used to. Channelbuttons are numbered in white from 1 to 16,doubling up from 17 to 32. Choosing betweenhexadecimal groups is done with two otherbuttons marked 1-16 and 17-32. Buttons forantenna polarity and position are provided, as arebuitons used to initiate the satellite selectionprocedure and to return to channel selection.

Although a parental-lock isn't a feature of theSRV1150, it is possible to lock the antenna positionso mean mummies and daddies wishing to frus-trate their offspring could lock the antenna into aposition where no satellite is sited, effectivelyrendering the system useless.

Where the system falls down is in generaloperation. Changing channels on a single satelliteis easy enough, a maximum of two button presseswill select any channel stored. But, when you wishto change to a channel located on a differentsatellite, things begin to get tricky. First, by press-ing the SAT button on the handset, the satelliteselection procedure is invoked. Then the satellitewanted (A1 to E5) is keyed in using the bluelettered buttons on the handset for the letter,followed by its number

-

at this point the antennacontrol unit moves the antenna to suit. Then, bypressing the CH button to return the system backto channel selection, the required channel is finallylocated. Although logical enough, it is a lengthyprocedure, all of which could entail a total of sixbutton presses

-

try doing that after a Vindalooand five pints of lager!

That really is the system's only fault (if you cancall it a fault). It is, however, only apparent whenconsidered beside receivers like the RediffusionRSR50 and the Pace SR540 because they are sosimple to use. Most other receivers are equally ascomplex, if not more so, than the Salora 1150system.

Pace SR540Satellite television receivers are a new venture forPace and it's interesting to consider why thecompany has gone down this path. Their firstproduct, the SR640 follows in the RediffusionRSRSO's footsteps in that it has separate modes ofoperation, making use extremely easy.

AIso like the RSR50, it is an integral systemrather than just a tuner. It has a similar separatepower supply too, although Pace has gone thewhole hog with the SR640 and put the receiverpower supply in the separate black box. This factmakes the SR540 itself the smallest (to myknowledge) and lightest receiver on the market.Housed in a custom-adapted Vero Proteusinstrument case, the receiver is certainly neat.

Specifications are, like the other receivers inour comparison, pretty good. Featuring the lowest

threshold of all (6dB) the SR640 will produce apicture where the others haven't got a sufficientlyhigh carrier-to-noise ratio

-

not that picturequality at a carrier-to-noise ratio of 6dB is worthwatching (see March ETI).

On the other hand, an IF bandwidth of only27MHz means that picture resolution for thosechannels such as Superchannel, with bandwidthsaround 36MHz isn't as good as, say, the SaloraSRV1150's. However, the trade-off betweenresolution and sensitivity which Pace seems tohave made means that fewer sparklies will occurwhen a weak signal is received when comparedwith wider IF bandwidths, lower sensitivityreceivers.

Back panel sockets for the usual audio, videoand baseband outlets allow direct video cassetterecorder or television monitor interconnection.

Only three controls adom the SR640's frontpanel, two of which are labelled with left and rightgoing arrows (channel down and channel up

-

innormal mode) and a mode control which steps thereceiver through the three modes

-

NORM(normal), DISH and TEST.

The remote control handset is quite easy touse. Ten numbered keys 0-9 are used for channelselection and can be used for single-buttonchannel selection (see later), although channelsare generally in decades

-

1-9, 10-19, 20-29,and30-39. Two buttons, tune- and tunef adjustreceived frequency. Two buttons, decade- anddecadet select which decade of channels areaccessed. One button alters polarisation and anumber of buttons are used to store antennaposition and channels. As well as a single mutebutton, volume can be turned up or down (a nicefeature) with the vol- and vol+ buttons and finallythere is a standby button.

The receiver is factory pre-set so that differentsatellite positions are allocated to each decade ofchannels, which means that to change to a channelfrom a different receiver up to four buttons mayhave to be pressed in sequence. However, userscan store channels from different satellites within asingle decade if they wish, thereby making singlebutton channel selection possible.

Channel selection takes place in normalmodeand users will rarely need to access any othermode. In dish mode, for example, the two arrowedkeys on the front panel change function to adjustdish position east and west. Dish mode is also themode used when channeli are to be stored in thereceiver. Like the Rediffusion RSR50, all details ofa channel including antenna position are storedtogether with a corresponding channel number sowhen in normal mode, users can select a channel(and antenna position) merely by pressing thechannel's number.

There are three test modes. Test mode 0produces a black and white UHF test signal usedwhen tuning the television receiver upon installa-tion. The other receivers in this review also featuresuch test signals, via switches on the rear panels.Test mode 1 converts the digital dish positiondisplay on the receiver's front panel into a digitaldisplay of signal strength

-

extremely usefulwhensiting the antenna. Test mode 2 is a skew adjust-ment, allowing the receiver to adjust the antennapolarotor's absolute vertical and horizontalpositions

-

only undertaken once on installation.Quirks are few and far between. The only

significant hardware problem is the lack of a UHF'loop-through' facility which, a company spokes-man has concurred, is to be included in the nearfuture. Without 'loop-through'UHF, signals from a

ETI APRIL 198818

terrestrial aerial aren't passed through to thefollowing video cassette recorder and television.Either users have to swop UHF leads, or a UHFsplitter/combiner is required. The only otherminor niggle is the fact that on power-up, thereceiver jumps to the first channel of whicheverdecade it was in on power-down. When usingnormal factory preset channels and decades thismakes absolutely no difference but if you prefer tostore different satellite channels on single decadesand if you were watching a channel from a satelliteother than that stored at the first channel of thedecade on power-down, when powering-up againthe receiver will reposition the antenna before apicture is obtained. So a wait of quite a few secondswill elapse before anything is seen. This could beprevented with a software change to either power-up the receiver in the channel it was in on power-

down or to put the receiver on standby whenpowered-up.

In ConclusionWell, we said comparison isn't easy, didn't we?Each receiver in this review has its own pros andcons and deciding which is the best is well nighimpossible. Best value for money has to be thePace SR640. The Rediffusion RSR50 just pips theSR640 to the post when considering which iseasiest to use. Highest quality picture resolutionmust be the Salora SRV1150 (due to its IF band-width) but the SR540 with the narrowest IF band-width gives a picture which is sparkly-free with theweakest of signals when the other two show somemost irritating interference.

Ah well, you pays your money

'FG*rtuteslfaoilitis/ RMiffrision RSf,5O Salora BfrVt 150^-^-tai^-.i^h

" hpirt level . -6O to.iThr6$trotdt6/N)'. ', 8dB ':, gdBI " Hz ',' 32M!",a2

ygs .,

' yesyes

Y9s ' yes 'yes yes

Pace SR64O

6dB27MHz

yesnoyesyes

6-6.8MHz

6 5,26 6/6.65MH2mono

channels 32 to 40yesno rsee text

nonoyspu tse/IRTE36 volt + reed275(w)x6O(h)x2SOtdI125{w}xgqh)x1 25(d)good

yesnono

norm/dishltestno

39

yestl0

12

+see textnoyesyesUK

f41 0

yss.Ycsyesy6sYesYGs:yes

ETI APRIL 19

ErlFl

Mihe Barwisetakes a look at theFairchild 74F525and finds there's alot of countersqueezed into thischip

20 ETI APRIL 1988

COtlA{T ON ITew generations of logic deviceswithin extant generic familiesusually differ in terms of power andspeed. However, sometimes amanufacturer will use the new

generation to introduce additional new devices.Such a device is the 74F525 programmable

divider from Fairchild. This is a versatile devicewhich can be used wherever binary programmabledigital timing is needed. It is typical of newintroductions in logic in that it incorporates thefunctions of ibout half a dozen separateconventional chips in one package.

The 74F525 (Fis. 1) consists of a 16-bit downcounter, a 16-bit input holding register and somecustom control logic which Fairchild has notdocumented. The device has five modes ofoperation, three of which have alternate modes. Itcan act as a triggered digitally timed monostablewith retrigger or pause, or as a free-running binarydivider. The Fairchild modes are given in Fig.2.

The Fairchild operational documentationleaves a bit to be desired (l had to phone the USAfor control informationl). So, I am giving here acontrol resumd before discussing.

The mode pins M0-2 are set by CIP switchesor hard wiring to the literal mode number inpositive logic. So all lines low equals mode 0 and allhigh gives mode 7. The input clock is applied toCP. The inverter between CP and XTAL must nofbe used as an oscillator inverter. It cannot carrythe required current. XTAL can be used to carryforward an inverted clock in.

The data inputs are conventional TTLcompatible logic inputs into a 15-bit transparentlatch. This is loaded by taking WE high after theinput data is stable, but due to the transparency ofthe latches WE should not be kept idle in the lowstate unless stable data is available from anexternal source.

The safest data write is performed by a shortnegative-going pulse, the data being effectivelylatched on the trailing (rising) edge.

The RUN/LOAD alternatives of the counterare controlled by the state of XTR. This stands forexternol trigger (a misleading name

-

it is not atrigger edge signal. It is a sfofe controlled modesignal). While XTR is high, the counter loodssynchronously with the clock, when low thecounter counfs.

The remaining control input MR is a masterreset. There must be a master reset after power-upbefore any other operation. I do not know whatwould happen in each mode otherwise but I reckonthe results would be pretty unpredictable!

Controlling The 74F525Whatever the mode of operation you have chosen,the external control sequence is much the same.After power-up, MR must be pulsed low. There-after it can be ignored, except in mode 7 (con-tinuous divide by n) where it is used to stop andclear the counter.

Next, XTR is taken high and a 16-bit dataword (the divisor of modulo) presented to the data

1 Block diaqram of the 74F525

inputs and held stable while WE is pulsed low.When XTR is next taken low, counting will

begin. When the counter reaches zero (countingdown) the outputs Q and Q/2 change state inaccordance with the chosen mode of operation.

In all modes except mode 7, XTR must betaken high and low each time a sequence isrequired. In mode 7, alter an initial XTR fallingedge, the counter is automatically reloaded by itsown rollover through zero. In this mode, theoutput at Q is a positive-going pulse one clockwide. Q/2 just changes state (toggle5) on the risingedge of Q.

The preceding notes show the internal trans-parent latches are only essential to operation inmode 7. In all other modes they simply avoid theuser reprogramming the counter externallybetween one-shot operations.

Some source of external programming is stillneeded and this usually takes the form of binaryswitch inputs or registers programmed by a micro-processor. Assuming for the moment that this isthe case, a simple loading/control interface isshown in Fig. 3.

Using The 74F525Now let us look at the various modes of this cleverlittle chip in more detail. In all cases, let us assumethat M-R has been pulsed and a modulo loaded intothe input register already.

In mode 0, the F525 operates as a 'wait for it'delay device. Q goes high n clocks after XTR islowered. Q/2 toggles state as well but this is notvery useful unless you know its state before.

If XTR goes high, the counter reloads and Q istaken low. If Q is still low (because the sequencehas not completed) it stays low. Lowering XTRrestarts the sequence from the top.

This mode is basic but not very widelyapplicable. Precision timing is not possible due tothe delay of anything up to nearly one clock periodbetween XTR falling and the first active clock. Themain use of this mode which comes to mind is as atime-out microprocessor watchdog. Softwarewrites to pulse XTR normally often enough for thecounter to not time out. If the CPU loses its way,XTR is not pulsed soon enough and the outputfrom Q is used to either reset the CPU or cause anerror interrupt.

In this application DIP switches would be usedto set the required modulo. WE would be tied

While XTR is HIGH, the data in the data latches is loadedinto the counter upon the next positive-edge of CP Thenegative-edge of XTR enables the count-down to beginwith the next positive'edge of CP. When the countreaches zero, 0 is brought HIGH and 0/2 toggles state'Taking XTR HIGH at any time causes the data in thelatches to be loaded into the counter and the 0 output tobe cleared.

MODE O

MODE 1

MODE 2

MODE 3

MODE 4

MODE 5

The operation is exactly the same as in mode 0 exceptthat 0 is normally HIGH and goes LOW on a count olzeto. O/2 toggles on the negative-edge of 0

While XTB is HIGH, the data in the data latches is loadedinto the counter upon the next positive'edge of CP' Thenegative- enables the count'down to beginwith the e-edge of CP. When the countreaches z ght HIGH for a single period of CP'0/2 toggles state e edge o{ 0. Taking XTRHIGH at any time ata in lhe latches to beloaded into the co 0 output to be cleared'

The operation is exactly the same as in mode 2 exceptthat 0 is normally HIGH and goes L()W on a count of zerofor a single period of CP O/2 toggles on the negative-edge of 0.

While XTR is HIGH, the data in the data latches is loadedinto the counter upon the next positive-edge of CP' Thenegative-edge of XTH enables the count-down to beginwith the next positive-edge of CP. When the countreaches zero, 0 is brought HlGl'lfor a single period of CP'0/2 toggles state on the positive-edge of 0. Taking XTRHIGH before the counters reach zero causes the datacurrently in the counters to be held.

MODE 6While XTR is HIGH, the data in the data latches is loadedinto the counter upon the next positive-edge of CP The

counter. but will not affect 0.

MODE 7

The operation is exactly the same as in Mode 4 exceptthat d is normally HIGH and goes L()W on a count olzerc. A/2 toggles on the negative-edge of 0.

The negative-edge of XTR enables the count-down tobegin with the next positive-edge of CP. When the count,rrirhrr.rro, 0 is brought HIGH for a single period of CP'0/2 toggles state on the positive'edge of 0 Thepositive-edge of CP upon which 0 goes low also causesihr dat, in the data latches to be reloaded into thecounters Taking XTR HIGH at any time causes the datain the data latches to be loaded into the counter and the0 output to be cleared. However, after an initial XTR thismode can run continuously until stopped by MR'

Fig. 2 The Fairchild mode descriptions

permanently low and the clock would drive CP.Mode 1 operates the same except the Q

output polarity is reversed. It goes low on time-out.

ETI APRIL

UPDATE-1F----+Lr-_3-_

Fig. 3 A simPle loading,/controlinterface

ml-lFJ{i-{.dz

TA +25'C T^, Vtc = Tr, Vcc =Vcc = +5 0V Mit ComCL = 50pF Cr = 50PF CL = 50PF

Min Tgp Max Min Max Min Max

f.,, Maximum 50 65ClockFrequencY

t(L) D. to WE 4 5th(H) Hold Time. 2 0

HIGH or LOWD" to WE 20

) D"toCP 11 5t'(H) Hold Time, 0

HIGH or LOWh(L) D" to CP 0t(H) Setup Time, 7 5

HIGH or LOWr(L) XTR to CP 13 s

Hold Time,HIGH or LOWXTR to CP

Setup Time,HIGH or LOWMode to CP

MR Pulse 7 5width. LowCP Pulse Width 4 0HIGH or LOW 9 0Recovery TimeMR to CP

Table 1 Timing constraints ol the 74F525

Modes 2 and 3 are very similar. The onlydifference is that the Q output is active for only oneinput clock period when the counter reaches zero.This output is therefore more suitable for drivinglevel triggered flip-flops, where only one input may

21

t-

Fig. 4 Using the 74F525 for precisionpulse generation

be active at one time for valid outputs.The addition of a simple one-shot tripped by

XTR and clocked synchronously with the F525allows this mode to be used for precision pulsegeneration (Fig. a). For this application, the inputof data can be the same as that specified for modes0 and 1 if binary switches are used but if a micro isused for programming (a good idea!) an additionalset of external registers is used.

Modes 4 and 5 are the positive and negativelogic variants of ahold-off monostable. This actslike a retriggerable monostable in that the periodmay be extended during the delay but instead ofreloading the delay from the registers, the F525simply freezes the counter when XTR goes high.This has the effect of applying an arbitraryasynchronous extension to the monostable periodwhile XTR is high.

Mode 6 is an internal implementation of myrather clumsy attempt at precision pulsegeneration using modes 2 and 3. After a fallingXTR, Q goes high on the first active clock (the firstclock to decrement the counter) and goes lowagain when the count reaches zero. This is the truedigital monostable. It is also retriggerable by takingXTR high and low again u-,rithin less fhan one clock

period (to avoid screwing up the timing precision)without affecting the state of Q.

The timing constraints (Table 1) suggests that

three PCB pulse generator87) could be built using onlymodes 6 and 7 plus a few flip-

Mode 7 is unique in that it is a free-runningmode. Once set up it will continue dividing by th!same modulo until either a master reset stops it, itfinds a new modulo in its registers or XTR is takenhigh. It is important to note that in mode 7 the realmodulo used by the counter is n+1. The output atQ is a clock-wide pulse and at Q/2 a square wave oftwice the period.

If you are going to modify the modulo on thefly, the external registers and the control interfaceof Fig. 3 are obligotory. This interface reloads theexternal registers, writes the internal registers andthen pulses XTR.

The only possibility of a restart glitch is whereby chance the atthe momentupdated. A littl ::by synchronisi tothe leading edge of Q (or the falling edge of theinput clock) allowing the update to be iterformedeither sufficiently early or late for this lrazard to becovered.

That's about it on the T4F52S,butrt.,villalmostcertainly crop up in the future as a supportingdevice for other new chips. It might be u,orth men-tioning that it is the perfect solution (in nrode 7) forthe MFS switched capacitor filter clock (ETI

I ma'".. I

SPECIAT OFFERFOR SPECTRUM AND BBC MICRO OWNERS

Now your computer can take control foran affordable price. These tried and trustedinterfaces from DCP Microdevelopments areoffered at t20 otl the normal price.

Both units are extremely easy to use fromboth Basic and assembler/machine code andare supplied ready built and complete with allthe documentation you need.

To order by post fill in the form below (ora copy) and send it with your remittance to

ASP READERS'SERVICES (RO ET5/6)9 Hall Road, Maylands Wood Estate,

Hemel Hempstead, HP2 7BHPlease make cheques payable to ASp Ltd.Overseas orders add f5 (lnterspec) or f10(lnterbeeb) for airpost.Access and Visa card holders can also placetheir order by phone on (0442) 41221Allow 28 days for delivery.

INTERSPEC f29.95The lnterspec unit plugs directly onto

the expansion edge connector of theSpectrum to provide a full range ofi nterfaci ng facil ities.

The unit is housed in a plastic caseapproximately 4/zx3x1in which contains thetop quality double sided PCB and interfacecon nections.. 8-bit input port. 8-bit output porto four or inputso four ed 1T lA outputs. e!8h ultiplexed anaiogue to

digita 15-way expansion bus

All sections of the interface are l/O portmapped and designed for maximum com-patibility with existing Spectrumperipherals. Power is supplied through theSpectrum edge conneclor.

The expansion bus provides all the dataand address/control signals for the additionof further DCP modules or home-builtdevices. Connection is by multi-way pCBconnector and all the information requiredfor adding further devices is given.

November 1987).

INTERBEEB f49.95The lnterbeeb unit connects to the BBC

micro's lMHz bus expansion connectorand is supplied complete with its ownpower supply unit.

The interface unit is housed in a plasticcase approx 4/zx3x1in which contains thetop quality double sided PCB and interfacecon nectors.o 8-bit input porto 8-bit output porto four switch sensor inputso four relay-switched 1fl 1A outputso elght channel multiplexed analogue to

digital convertero precision 2.5V referenceo external power supplyo 15-way expansion bus

interface are memoryma expansion map forma and compatibilitywit als.

The expansion bus provides all the dataand address/control signals for the additionof further DCP modules or home-builtdevices. All the information required forusing additional devices is included.

Please debit my ACCESS/VISA cardNo..........,...,...... to the sum ofr

SS COMPETITIONRESULTS

The Specialist Semiconductors kit competitionback in the January 1988 issue attracted a largenumber of entries.

Most readers managed to correctly calculate thecurrent in the given network as 1mA (to the nearestmA). The wordsearch caused a few problems withsome entries listing every combination of threeletters possible! However, from the entries with themost real words the first out of the hat were fromHarry Clough of Hitchin and Mark Thomson ofBarnet. Congratulations to the lucky winners andcommiserations and thanks to everyone else whoentered.

The t75 Specialist Semiconductor vouchers willsoon be winding their way to Mr. Clough and Mr.Thomson.

The wordsearch square with most of the realelectronic words is given here to show you whereyou went wrong! Don't forget the T-shirtcompetition in this issue!

GITTYB

DAER

oTALLIcso

Ect{BF

EEDBAC

VIU'IVA

oTSIsNAR

IucRI

oTIcAPA

UY

xEITv

Ht{oKoa

JsAFILNP

QI,CDUJTJ}TUTSIJ"

AKI

ETI APRIL

Paul Chappellshows how thetheory of recentmonths can be putlo practical use

EtrJt-{FJ-{E-{

F-{t-{l--JEJEHEJ

PULTI}{G IT ALLTOGE

ince we've spent most of the past sixmonths looking at phasors and complexalgebra, I think it's almost time to give it arest for the time being. Before takingourleave of the topic, let's just look around

and see what we've achieved so far.One of the problems with explaining the

theoretical side of electronics is that most demon-strations of any new idea involve very simplenetworks far removed from the complexities ofpractical circuits. The idea, of course, is to showup the new concepts as sharply as possible withoutbecoming involved in irrelevant detail. But thebetter it works, the more remote the theory seemsfrom real life!

Believe it nor not, if you've followed thearticles so far you already have a useful tool at yourdisposal. It's not a complete and self-containedmeans of investigating circuit behaviour but thenno other single technique can make that claimeither. Used in conjunction with your intuition andgeneral electronics knowledge it can help to shednew light on obscure areas and act as an aid todeveloping new ideas.

In an earlier article I had occasion to mentionamplitude modulated radio signals to make thepoint that an amplitude modulated sine wave is nota sine wave at all, rather a combination of threesine waves. At the time the best'proof'l could offerwas to refer you