electronics on-line · designing electronic circuits almost invariably calls for extensive...

electronics on-linedatasheetscomponent documDesigning electronic circuits

almost invariably calls forextensive descriptions of theoperation and typical appli-cation of the more complex

components in your design.Although a fair amount of

documentation may befound in databooks and CD-

ROMs (like Elektor'sDatasheet Collection Vol. 1

and 2), the Internet is also avast resource when it comes

to finding datasheets.

These days, many manufacturers ofelectronic components use Internetsites to offer vast amounts of data ontheir products. This service is usuallyfree of charge and available to anyone.Most datasheets are available in the so-called PDF format which can be readand printed by the popular AcrobatReader from Adobe. This reader pro-gram is free of charge, and may bedownloaded from www.aclobe.com/sup-portservice/custsupporadownload.htmlOften, the main problem is to locatethe datasheet of a particular compo-nent. The good news is that specialsearch engines are available to helpyou find what you want. Partminer isa good example, it may be found atwww.partminer.com/partminer/index.html.Using a helper program (which may bedownloaded free of charge), you canstart looking for type numbers anddescriptions. Partminer then makes theconnections with the relevant manu-facturer sites, and starts looking there.You can report new manufacturers toPartminer for inclusion in the searchoverview.Another datasheet search engine iscalled WebStir. You may find it atwebstirinfoquick.comliq-home.html.Like Partminer, this engine offers exten-sive information and search options

PartM in er(tm) MEMFile Options News Industry Forums Help

Ready - Enter a Part # or Description, select e(o) a TP

!1a t r el . 33

1 (eg. L1/13177

Request. Data Sheet ..ir

`To select more than one site, hold [Ctrl] key down and click mouse

Adaptec '''Advanced Hardware ArchitecturesAlteraAMDAnalog Devices

C-Cuhe MicrosystemsCherry SemiconductorDiodes Inc./Vishay-LiteonFairchild SemiconductorGennumHarrisHitachi

rInternet Quick Launch

Go to - I

Expand -)

Ihyvo

Today's NewsFor Mimi; Bum

4110,TITIT

Start Search

Stop

View Results

Add New Sites

SHORTAGES?CLICK HERE

MICROCOMLINE CARD I

EMBMEEEEDistributor

EDTN

.r.miedtn.ccm/embapps/e.ba016.htri, 's Big Move

with many manufacturers. It is, how-ever, not a free service. None the less, itis still of interest to the hobbyist becausethe program may be used on a 30 -daytrial basis, and up to 10 datasheets maybe downloaded free of charge.There are also sites providing theirown component databases or links forthat purpose. An example is the Com-ponent Database Server of the Centerfor Electronic Design, Communicationsand Computing (CEDCC) at PennState University, USA. Currently, thisserver holds mainly Motorola and Har-ris component datasheets.Component suppliers, too, are startingto offer datasheets by means of onlineservices. Although the English-lan-

Info Quick - Electronic Designer Interlace - NotscapoFile Edit Vie

a

guage web site of Conrad has been bnthe air' for some time atwww.conracl-electronic.comit does not (or not yet) offer thedatasheet service available on the website run by the mother company inGermany, atwww.conracl.de/index/htmlSearching for a particular part is noteasy, and the Conrad catalogue shouldbe used alongside the information onweb pages.Finally, there is Leeds -based Farnell atwww.farnell.com/uk/index.htmlTheir web site contains datasheets onno fewer than 1500 components, and iscertainly worth a visit.

(995005-1)

MDR

-Bookmarks A. Loce.tion.lbttp,ifwebsti fo-quick.comf ee_usertdefaul eb asp

GetD,s-taS beets

88110 se nodal

webst, - 1.',.quponent Search

Function: I (ALL)

l(ALL)

Manufacturer Max

lid PART_NUMBER

Filter17808

Interface Analog to Digital Converters Binary OutputAdd Part #

ADS7808

Dots

13 1.711242040

Description I ManufacturerA/D Converters, Data Acquisition- 12 -Bit 1O. Serial

Burr -BrownCMOS Sampling Analog -to -Digital Converter

Interface Digital to Analog ConvertersAdd Part # Docs Description

ILI AD78081? 2 121 1242040 13.3 V to 5 V Quad 10 -Bit DACs

A AD7808S A 211242040 13.3 V to 5 V Quad 10 Bit DACs

Linear Power Supply Circuits

ManufacturerAnalog Devices

Analog Devices

Add Part # I Docs Description ManufacturerL7808CP 1.2i 2012029030(POSITIVE VOLTAGE REGULATORS SGS-Thomson

I=1 L7808CT Mi 2012029030 POSITIVE VOLTAGE REGULATORS SGS-Thomson

L7808CV It2i 2012029030 'POSITIVE VOLTAGE REGULATORS SGS-Thomson

E L7808T 1[7] 2012029030 (POSITIVE VOLTAGE REGULATORS SGS-Thomson

1=1 L7808ABVIM 2012029030 PRECISION IA REGULATORS SGS-Thomson

Document, Done OIDMIEM=. sz

Elektor Electronics 1/99 11Elektor Electronics 1/99

with many manufacturers. It is, how-ever, not a free service. None the less, itis still of interest to the hobbyist becausethe program may be used on a 30-daytrial basis, and up to 10 datasheets maybe downloaded free of charge.There are also sites providing theirown component databases or links forthat purpose. An example is the Com-ponent Database Server of the Centerfor Electronic Design, Communicationsand Computing (CEDCC) at PennState University, USA. Currently, thisserver holds mainly Motorola and Har-ris component datasheets.Component suppliers, too, are startingto offer datasheets by means of onlineservices. Although the English-lan-

guage web site of Conrad has been ‘onthe air’ for some time atwww.conrad-electronic.comit does not (or not yet) offer thedatasheet service available on the website run by the mother company inGermany, atwww.conrad.de/index/htmlSearching for a particular part is noteasy, and the Conrad catalogue shouldbe used alongside the information onweb pages.Finally, there is Leeds-based Farnell atwww.farnell.com/uk/index.htmlTheir web site contains datasheets onno fewer than 1500 components, and iscertainly worth a visit.

(995005-1)

electronics on-lineelectronics on-line

These days, many manufacturers ofelectronic components use Internetsites to offer vast amounts of data ontheir products. This service is usuallyfree of charge and available to anyone.Most datasheets are available in the so-called PDF format which can be readand printed by the popular AcrobatReader from Adobe. This reader pro-gram is free of charge, and may bedownloaded from www.adobe.com/sup-portservice/custsupport/download.htmlOften, the main problem is to locatethe datasheet of a particular compo-nent. The good news is that specialsearch engines are available to helpyou find what you want. Partminer isa good example, it may be found atwww.partminer.com/partminer/index.html.Using a helper program (which may bedownloaded free of charge), you canstart looking for type numbers anddescriptions. Partminer then makes theconnections with the relevant manu-facturer sites, and starts looking there.You can report new manufacturers toPartminer for inclusion in the searchoverview.Another datasheet search engine iscalled WebStir. You may find it atwebstir.infoquick.com/iq-home.html.Like Partminer, this engine offers exten-sive information and search options

component documentation

datasheetsDesigning electronic circuits

almost invariably calls forextensive descriptions of theoperation and typical appli-cation of the more complex

components in your design.Although a fair amount of

documentation may befound in databooks and CD-

ROMs (like Elektor’sDatasheet Collection Vol. 1

and 2), the Internet is also avast resource when it comes

to finding datasheets.

component documentation

GENERAL INTEREST

home alarm systemprogrammable burglar deterrent

with a PC interface

Crime in general isstill on the rise, and

having an alarm sys-tem installed is no

longer a privilege ofthe wealthy. This arti-

cle shows that anadvanced alarm sys-

tem to protect yourhome and valuables

can be very compactindeed. Using the PC,the authorized user is

able to program themain parameters of

the alarm system.

Main featuresProcessor:Serial port:Inputs:Outputs:User indication:

Alarmcontacts

phone dialler

Interfacefor option

Key switch to arm system

Actuation delay (default):

Siren on time:

Adjustableactuation delay:

`Armed' indicator

Battery back-up

PIC17-6F8419,200 bits/s

direct and delayedrelay for wen

buzzer

normally closed

20

seconds

1 to 255 seconds

to 99 seconds

Design by H. Sommen

412 Elektor Electronics 1/99Elektor Electronics 1/99

Crime in general isstill on the rise, and

having an alarm sys-tem installed is no

longer a privilege ofthe wealthy. This arti-

cle shows that anadvanced alarm sys-

tem to protect yourhome and valuables

can be very compactindeed. Using the PC,the authorized user is

able to program themain parameters of

the alarm system.

12

Design by H. Sommen

home alarm systemprogrammable burglar deterrent

with a PC interface

Main features

Processor:

PIC17-6F84

Serial port:

19,200 bits/s

Inputs:

direct and delayed

Outputs:

relay for siren

User indication:

buzzer

Alarm contacts:

normally closed

Interface for optional phone dialler

Key switch to arm system

Actuation delay (default):

20 seconds

Siren on time:

1 to 255 seconds

Adjustable actuation delay:1 to 99 seconds

‘Armed’ indicator

Battery back-up

GENERAL INTEREST

By popular demand, PIC processorsand alarm systems are two subjectswhich are often covered in this maga-zine. In this article, you find the twocombined in a programmable alarmsystem for home construction. Here, aPIC processor is employed as the logic‘glue’ between the various sensors(detection devices) and alarm actua-tors. Besides this function, the PIC alsohandles all communication betweenthe alarm system and an (optional) PCof the IBM/compatible type.

Using a simple RS232 link and astandard terminal program, the mainparameters of the alarm, includingalarm time, may be programmed.

Optionally, the alarm may beextended with a separate telephonedialler, which allows ‘silent signalling’to be implemented. The relevant hard-ware is not discussed in this article, butwe intend to cover it in a future issueof Elektor Electronics. A commerciallyavailable dialler unit may be connectedto a dedicated output on the alarm sys-tem. If the alarm goes off, transistor T1will conduct for about one second —long enough to actuate an externaldialler.

T H E A P P R O A C HThe circuit diagram of the home alarmsystem is given in Figure 1. The con-nections to the various alarm detectiondevices have been kept as simple anduniversal as possible. Assuming thatdevices with normally-closed (nc) con-tacts are used, buffer inputs (IC3a-IC3d) are pulled low via a group ofconnection terminals (K3, K6, K7, K8).A pull-up resistor and a 470-nF affordample noise suppression.

The connectionpoint of the main switch(K3) has an extra indica-tor. If LED D6 is on, thealarm system is armed.This input is thereforebest connected to a keyswitch.

The detection device connected toterminal block K8 is ‘interpreted’ withan adjustable delay (for example, 20seconds). When the alarm is armed, anintegrated buzzer starts to sound toindicate that a contact in this group isopened within the delay period. Thisdelay is needed to leave the buildingwithout setting off the alarm. The dooryou normally use to leave the buildingis therefore connected to the ‘delayed’input.

When you enter the building, thesame delay is available to de-activatethe alarm. However, the buzzer willnot sound during this period to pre-vent it giving away the location of thealarm control unit.

The alarm contacts connected-up toterminal block K7 produce an instantalarm when opened. The upshot is

that all ‘nc’ detec-tion devices areconnected in seriesand to this input. If

any one of the switches is opened, thedetection loop is interrupted, and thealarm is set off.

The last input is the ‘tamper’ input.The switch protecting the alarm enclo-sure is connected to this input. If some-one attempts to open the enclosurewhile the alarm is in the ‘armed’ state,this will not go unnoticed because thealarm will go off.

The alarm actuator (for example, asiren) is switched by relay Re1. Thisrelay will be energized for a program-mable period when an alarm conditionis detected. The default ‘on’ time is180 seconds. Using the PC, however,you may set any period between 1 and255 seconds.

If an alarm condition has occurred,and the siren has been switched off,LED D8 will remain on. In this way,

when you come home, you areinformed that the alarm went off atleast once during your absence. TheLED activity can be cleared by pressingreset button S1.

What remains to be discussed at thispoint are the connections of the powersupply and the lead-acid battery. Thepower supply (approx. 13 V) has to beconnected to terminal block K1, whilethe 12-V lead-acid battery goes to K2.The battery is kept topped up via resis-tor R3. The three-pin regulator in thepower supply section provides the cor-rect supply voltage for the microcon-troller in the alarm control unit.

The battery used in the alarm sys-tem should have sufficient capacity topower the alarm control unit, the siren(or flashlight) and the telephone dialler.The battery is incorporated in thedesign to ensure that the alarm systemkeeps working when the mains elec-tricity system is tampered with by bur-glars.

13Elektor Electronics 1/99

PIC16C84

OSC2

IC2

OSC1

MCLR

RA4

RA1

RA0

RA2

RA3

RB0

RB1

RB2

RB3

RB4

RB5

RB6

RB7

17

18

13

12

11

10

16 15

14

1

3

9

8

7

6

2

4

5

C5

100n

X1

8MHz

C7

22p

C6

22p

K5

1

2

3

4

5

6

7

8

9

R2

1k

D3

BAT85

D4

BAT85

13 121

IC3f

R1

1k

R3

1kD7

ON/OFF

C8

470n

K3

3 41

IC3b

R9

12

k

C10

470n

K7 R10

100k

IMMEDIATE

1 21

IC3a

R11

12

k

C11

470n

K8 R12

100k

DELAYED

T3

BC547

RE1D5

1N4148

K9

R13

10k

T2

BC547

R5

10k

R6

47

Ω

BZ1

T1

BC547

R4

10k

K4

DIALER

561

IC3c

R7

12

k

C9

470n

K6R8

100k

TAMPER

K1

K2

D1

1N4001

D2

1N4001C4

100µ63V

C3

10µ63V

C1

100n

C2

100n

7805

IC1

IC3

14

7

5V

5V

5V

IC3 = 4069

12V 12V

5V

5V

12V

12V

12V

5V

980091 - 11

D8

R14

1k

R15

10

k

S1

9

8

1

IC3d

11

10

1

IC3e

5V

Bt1

1

Figure 1. The programstashed away in the PICprocessor allows thehardware to be kept verysimple indeed.

14 Elektor Electronics 1/99

980091-1(C) ELEKTOR

C1

C2

C3

C4

C5

C6 C7C8

C9 C10 C11

D1D2D

3D4 D5

D7

D8

IC1

IC2

IC3

K1K2

K3K4

K5

K6

K7

K8

K9

R1R

2

R3

R4

R5 R6

R7

R8

R9

R10

R11

R12

R13

R14

R15

RE1

S1

T1

T2

T3

X1

DEL

AY

IMM

TAM

PER

ON

/OFF

DIA

L

0 +-+BT1

+-BZ1

980091-1

RS232on_off = OFF

On_off = OFF

Dial = OFF

Dial = OFF

Isloopcnt

> 250?

Loopcnt +1

Wait 1 sec

Siren = OFF

Repeat = 180

Alarmbit = ON

Siren = ON

Dial = ON

Repeat = 1

Siren = OFF

Repeat -1

Dial = ON Repeat -1

Bit on_off = OFF

Alarmbit = OFF

Siren = OFF

Del_rep2 = 20

Del_rep = 20

Del_rep -1

Beep 1 second

Bit on_off = OFF

Bit on_off = OFF

Loopcnt = 0

Alarmbit = OFFDel_rep2 -1

Bit on_off = OFF

Loopcnt = 0

Del_rep2 = 20

Del_rep = 20

Repeat = 180

START

Iscontact

ON?

IsRS232on_off

ON?

Isalarmbit

ON?

Istamper

ON?

Isloopcnt

< 21?

Isrepeat

0?

Isrepeat

10?

Isdel_rep2

0?

Isgroup1

ON?

Isgroup2

ON?

Isdel_rep

0?

Isbit on_off

ON?

Loopcnt = 249

Bit on_off = ON

Y

N

Y

N

Y

N

Y

N

Y

N

Y

N

Y

N

Y

N

Y

NY

N

980091 - 12

Y

N

Y

N

Y

N

COMPONENTS LIST

Resistors:R1,R2,R3,R14 = 1kΩR4,R5,R13,R15 = 10kΩR6 = 47ΩR7,R9,R11 = 12kΩR8,R10,R12 = 100kΩ

Capacitors:C1,C2,C5 = 100nFC3 = 10µF 63V radialC4 = 100µF 63V radialC6,C7 = 22pFC8,C9,C10,C11 = 470nF

Semiconductors:D1,D2 = 1N4001D3,D4 = BAT85D5 = 1N4148D6,D7 = LED, red, high efficiencyT1,T2,T3 = BC547

3

2

Figure 2. Copper track layout and compo-nent mounting plan of the printed circuitboard designed fore the alarm system.

Figure 3. Flow dia-gram of the mainprogram written forthe home alarm sys-tem. The relevantsource code file isavailable on floppydisk.

101 = PIC16F84(order code 986519-1)

IC2 = 7805IC3 = 4069

Miscellaneous:K1-K4,K6-K9 = 2 -way PCB terminal

block, raster 5mmK5 = 9 -way sub -D socket (female),

PCB mount, angled pinsBz1 = piezo buzzer, 5V DCX1 = 8 MHz quartz crystalS1 = push-button, 1 make contactRe1 = 12-V relay, 1 change -over con-

tact (e.g., Siemens V23057 -130001-A002)

PCB, PIC and floppy disk (set),order code 980091-C

PCB only, order code 980091-1Source code disk only,

order code 986028-1.

OPTIONALPC CONTROLSerial communication with your PCtakes place via 9 -way connector K5. Onthe printed circuit board, all relevanthandshaking signals are available,allowing a standard (commerciallyavailable) RS232 cable to be used. Resis-tor R2 and diodes D3 and D4 convertthe RS232 signal levels on the interfaceinto TTL levels that can be processedby the alarm circuit.

CONSTRUCTIONBuilding this circuit should not presentproblems because a ready-made high -quality printed circuit board (PCB) isavailable from the Publishers (see theComponents List, and the Readers Ser-vices page elsewhere in this issue). Thecopper track layout and component -mounting plan of this compact boardare given in Figure 2.

SILICON HEARTThe PIC processor in the present pro-ject underlines the fact that microcon-trotters allow loads of functionality tobe combined in a single compact case.Here, the processor ticks at a rate of8 MHz.

The control software programmedinto controller ROM consists of a largeloop which is repeated every second.As shown by the flow diagram in Fig-ure 3, the basic structure of the pro-gram boils down to continuous testingof bits and/or input levels. Based on the

'itoco it it it

results of these 'interrogation' activi-ties, the processor determines whetheror not an alarm condition has to be sig-nalled. Because the program has a rel-atively simple structure, new featuresare easily added with the aid of thesource code file, which is available sep-arately on floppy disk (order code986028-1).

Serial communication between thePC and the alarm control unit is onlypossible in 'standby' mode, i.e., whenthe switch contact on K3 is open. Runa terminal program on your PC (or

connect a real 'dumb' terminal), andselect these communication parame-ters:19,200 bits/s, even parity, 8 databits and 1stopbit.`Local echo' should be enabled on theterminal. Press the Return key and fol-low the instructions on the screen. Thenew parameters are effective after youswitch the alarm control off and onagain. The screendump in Figure 4shows the information which shouldappear on the terminal display.

(980091-1)

Figure 4. The home alarm system is easy to program ifyou have a terminal or a PC running a communicationprogram like Telix, ProComm or HyperTerminal.

Direct° pood 19200 E 8 1 - HyperTerminal P117 El

Destand De2e:ken Deeld Oppepen Livethrengen Help

JSiren -on time in seconds is120

(001..254 [3 digitsl] ) ,

Modify? Y/N CENTER>Actuation delay in seconds is (00..99 [2 digits 1 ] ) :

20

Modify? Y/N CENTER>Closedsiren -on time in seconds is (001..254 [3 digits 1 ] ) :

100

Modify? Y/N CENTER>Actuation delay in seconds is (00..99 [2 digits 1 ] ) :

20

Modify? Y/N CENTER>Please input en press EnterActuation delay in seconds is (00..99 [2 digits 1 ] ) :

12

Modify? Y/N <ENTER>Closed

]]Verb:olden:0:0026 1.4,Tedelectie 0 Kw IN .1 SCROLL FCAPS MT 1 Domalumn lAHmkecho

Elektor Electronics 1/99 15L

O P T I O N A LP C C O N T R O LSerial communication with your PCtakes place via 9-way connector K5. Onthe printed circuit board, all relevanthandshaking signals are available,allowing a standard (commerciallyavailable) RS232 cable to be used. Resis-tor R2 and diodes D3 and D4 convertthe RS232 signal levels on the interfaceinto TTL levels that can be processedby the alarm circuit.

C O N S T R U C T I O NBuilding this circuit should not presentproblems because a ready-made high-quality printed circuit board (PCB) isavailable from the Publishers (see theComponents List, and the Readers Ser-vices page elsewhere in this issue). Thecopper track layout and component-mounting plan of this compact boardare given in Figure 2.

S I L I C O N H E A R TThe PIC processor in the present pro-ject underlines the fact that microcon-trollers allow loads of functionality tobe combined in a single compact case.Here, the processor ticks at a rate of8 MHz.

The control software programmedinto controller ROM consists of a largeloop which is repeated every second.As shown by the flow diagram in Fig-ure 3, the basic structure of the pro-gram boils down to continuous testingof bits and/or input levels. Based on the

results of these ‘interrogation’ activi-ties, the processor determines whetheror not an alarm condition has to be sig-nalled. Because the program has a rel-atively simple structure, new featuresare easily added with the aid of thesource code file, which is available sep-arately on floppy disk (order code986028-1).

Serial communication between thePC and the alarm control unit is onlypossible in ‘standby’ mode, i.e., whenthe switch contact on K3 is open. Runa terminal program on your PC (or

connect a real ‘dumb’ terminal), andselect these communication parame-ters:19,200 bits/s, even parity, 8 databits and 1stopbit.‘Local echo’ should be enabled on theterminal. Press the Return key and fol-low the instructions on the screen. Thenew parameters are effective after youswitch the alarm control off and onagain. The screendump in Figure 4shows the information which shouldappear on the terminal display.

(980091-1)


980091-1(C) ELEKTOR

IC1 = PIC16F84(order code 986519-1)

IC2 = 7805IC3 = 4069

Miscellaneous:K1-K4,K6-K9 = 2-way PCB terminal

block, raster 5mmK5 = 9-way sub-D socket (female),

PCB mount, angled pinsBz1 = piezo buzzer, 5V DCX1 = 8 MHz quartz crystalS1 = push-button, 1 make contactRe1 = 12-V relay, 1 change-over con-

tact (e.g., Siemens V23057-B0001-A002)

PCB, PIC and floppy disk (set),order code 980091-C

PCB only, order code 980091-1Source code disk only,

order code 986028-1.

Figure 4. The home alarm system is easy to program ifyou have a terminal or a PC running a communicationprogram like Telix, ProComm or HyperTerminal.

IO, TELEVISION & VIDEO

general-coveragereceiverpart 1: circuit descriptions

\NA

This two-part article

describes anAM/FM/SSB

receiver for thefrequency range

0.15 -32 MHz,which is generally

(but incorrectly)referred to as 'the

shortwave bands'.The receiver is micro-processor controlledand avoids many of

the pitfalls traditionallyassociated with RF

construction.

Main SpecificationsI Double conversion superheterodyne receiver, 1st IF 45 MHz,

2nd IF 455 kHzI Microprocessor control of synthesizer tuning and other receiver

functions1150 kHz to 32 MHz tuning range in 1 -kHz steps.I Selectable selectivity: 3 kHz (narrow) or 12 kHz (wide)I Internal 6 -band preselector with automatic band switchoverI 12 -key keyboard for frequency entry,

mode and bandwidth selectionI 16 -character LCD shows receive mode, bandwidth, frequency

and preselector bandI Memory for 21 frequencies, incl. bandwidth and modeI Spurious product rejection > 50 dBI Audio output power approx. 1 W into 8 n.I Power supply 15 V, max. 400 mA (approx. 90 mA without audio

and LCD backlight)Design by G. Bears, PE1GIC

1 8Elektor Electronics 1/99Elektor Electronics 1/9918

Design by G. Baars, PE1GIC

general-coveragereceiverpart 1: circuit descriptions

RADIO, TELEVISION & VIDEO

This two-part article

describes anAM/FM/SSB

receiver for thefrequency range

0.15 – 32 MHz,which is generally

(but incorrectly)referred to as ‘the

shortwave bands’.The receiver is micro-processor controlledand avoids many of

the pitfalls traditionallyassociated with RF

construction.

Main Specifications Double conversion superheterodyne receiver, 1st IF 45 MHz,

2nd IF 455 kHz

Microprocessor control of synthesizer tuning and other receiver

functions

150 kHz to 32 MHz tuning range in 1-kHz steps.

Selectable selectivity: 3 kHz (narrow) or 12 kHz (wide)

Internal 6-band preselector with automatic band switchover

12-key keyboard for frequency entry,

mode and bandwidth selection

16-character LCD shows receive mode, bandwidth, frequency

and preselector band

Memory for 21 frequencies, incl. bandwidth and mode

Spurious product rejection >50 dB

Audio output power approx. 1 W into 8 Ω.

Power supply 15 V, max. 400 mA (approx. 90 mA without audio

and LCD backlight)

The receiver we’re about to describe isthe product of many hours of design-ing, testing and programming by theauthor, a licensed radio amateur fromthe Netherlands. Throughout thedesign process, the emphasis has beenon repeatability, ease of constructionand avoidance of many of the pitfallscommonly associated with buildingradio equipment. As many of you willavow, the two best known pitfalls arewinding your own coils and non-avail-ability of specialized test equipment toalign the receiver, or, indeed, any otherRF project you may want to build.

So how are these problems solved?Well, the present receiver has only oneinductor you have to wind yourself,and the use of ready-made filters and

transformers in the IF sections obviatesthe need for complex constructionsand adjustments. If you are a carefulbuilder with some experience in RFtechnology, then the receiver shouldwork spot-on, and a minimum ofadjustments is needed to tweak it foroptimum performance. The goodnews is that these adjustments onlyrequire the built-in S meter, your hear-ing ability, and possibly a voltmeter.

T H E C O N C E P TThe block diagram of the general-cov-erage receiver is shown in Figure 1.The design is that of a double-conver-sion superheterodyne receiver with a‘high IF’, which means that the firstintermediate frequency (IF) is wellabove the highest receive frequency.

The antenna signal is first takenthrough a preselector stage whose

main purpose is to reduce the risk ofinterference and cross-modulationproducts caused by very strong signals.The preselector is manually tuned forbest performance. The second functionof the preselector is to make thereceiver input virtually independent ofthe antenna used: in fact, anythingranging from a simple telescopeantenna to a full-blown ‘beam’ (with acable impedance of 50 Ω) or a long-wire may be connected. Alternatively,for indoor use, consider a small mag-netic-loop antenna such as the superbDJ8IL design described in the Septem-ber 1998 issue of Elektor Electronics.

The preselector is followed by a pre-amplifier stage with manuallyadjustable gain. Here, again, one of the

most important design considerationsis to keep strong signals away from theinput of the next stage, the mixer. Ifyou are new to shortwave reception,then remember that your main con-cern is not dredging in the noise to getthe weakest possible signal into thereceiver, but to keep multi-megawattsignals out.

The local oscillator (LO) signal forthe first mixer is supplied by a synthe-siser circuit which can be tuned insteps of 1 kHz across the range45.150 MHz to 77.000 MHz. The syn-thesizer consists of the usual ingredi-ents: a VCO (voltage-controlled oscil-lator) a prescaler, and a loop filter for

suppression of the reference frequency(here, 1 kHz). Like a number of othersub-circuits in the receiver, the synthe-sizer is digitally controlled by a centralmicroprocessor.

The output signal of the first mixeris taken through a 45 MHz filter with abandwidth of about 15 kHz. The mainfunction of the filter is to suppress theimage frequency of the second mixer,which occurs at 44.090 MHz(44.545–0.455).

The first IF signal (45 MHz) is het-erodyned down to 455 kHz by meansof the second mixer and the second LOsignal, which is supplied by a crystaloscillator operating at 44.545 MHz. Themixer is followed by two bandpass fil-ters, one with a width of 3 kHz for ‘nar-

row-band’ mode (SSB), and one with awidth of 12 kHz for FM and AM recep-tion. The gain of all IF amplifier stages(45 MHz and 455 kHz) is controlled byan AGC circuit (automatic gain con-trol). Because the AGC voltage is ameasure of the received signalstrength, it can also be used to drivethe S-meter.

The last 455-kHz amplifier drivestwo demodulators (for AM/FM recep-tion), and a product detector (for SSBreception.) The oscillator in the prod-uct detector can be pulled a little toallow USB/LSB selection. The relevantcontrol is labelled BFO (beat frequencyoscillator). Analogue switches are used


Keyboard

LCD

PRESELECTOR

RF GAIN

S meter

VOLUMEBFO

TUNING

BandSelect VCO

MIXER 1IF 1

MIXER 2

Synthesizer

FMDEM.

AMDEM.

SSB DET.

÷ 64÷ 65

45 MHzIF 2

455 kHz

455 kHz

12 kHz

44.545 MHz

455 kHz

3 kHz

WIDE

980084 - 11

AGC

FM

AM

SSB

NARROW

rotaryencoder

Figure 1. Block diagram of the general coverage receiver.The design is a double conversion superheterodyne withhigh-side injection for the first LO. The use of a ‘high’ firstIF (45 MHz) guarantees a minimum of in-band generatedspurious products while also reducing the risk of IF break-through. Note that many functions are controlled by a cen-tral microprocessor.

1

to feed one of thedemodulator/detectoroutputs to the input ofthe audio amplifier, byway of a ‘speech’ filterwith roll-off points at 450 Hz and3.3 kHz.

The microprocessor circuit controlsthe preselector, the synthesizer, the IFbandwidth (wide/narrow), the modeselection (AM/FM/SSB, and the LCD(liquid crystal display). Its ‘inputdevices’ are a rotary encoder for thereceiver tuning, and a small keyboardfor direct frequency entry and severalother functions like channel memorycontrol, manual bandwidth selection(3 kHz/12 kHz), etc.

P R A C T I C A L C I R C U I TDrawing a block diagram is one thing,actually implementing the functionswith real components is quite another.

Although the circuit diagram in Fig-

ure 2 may look large and complex atfirst, its operation is relatively easy tounderstand thanks to the previousdescription of the block diagram. Let’stake the sub-circuits one by one.

PreselectorThe active element is a type BF961dual-gate MOSFET, T1, which guaran-tees minimum loading of the inductorsin the preselector. PIN diodes are usedto allow the outputs of a decimalcounter to switch the requisite induc-tors on and off. The counter, in turn, iscontrolled by the microprocessor. Forthe sake of repeatability, ready-mademiniature chokes from the E12 seriesare used in the preselector. Their Q fac-tors remain as high as possible thanks

to the small capacitive load presentedby the DG MOSFET. The preselectorhas six ranges:

1: 150 – 370 kHz2: 370 – 900 kHz3: 900 – 2200 kHz4: 2200 – 5400 kHz5: 5400 – 13200 kHz6: 13200 – 32000 kHz

The inductive part of the preselector isbrought to resonance by the capaci-tance formed by a pair of varicapdiodes, D14-D13. The varicap controlvoltage has a range from 0 to 9 V, andis supplied by the wiper of the prese-lector tuning control, P1.

The gain of the DG MOSFET is con-trolled in traditional fashion by meansof a direct voltage on gate 2. Althoughthe preselector already affords consid-erable suppression of unwanted fre-quencies, the MOSFET is followed by


C4

10n

C5

100n

L3

0mH82

D3

D4

R2

33

0Ω

C2

10n

C3

100n

L1

1mH5

D1

D2

R1

33

0Ω

L2

1mH5

C6

10n

C7

100n

L4

120µH

D5

D6

R3

33

0Ω

C8

10n

C9

100n

L5

18µH

D7

D8

R4

33

0Ω

C10

10n

C11

100n

L6

3µH3

D9

D10

R5

33

0Ω

C12

10n

C13

100n

L7

0µH68

D11

D12

R6

33

0Ω

C1

6p8

Q0 Q1 Q2 Q3 Q4 Q5

D14

BB112

D13

BB112

R7

1M

R8

100k

P1

50k

C14

100n

C15

100n

P2

50k

R9

10

0k

R10

330k

C16

100n

R12

68

0Ω

R67

47

k

R13

15

k

R15

68

k

R14

33

k

R16

18

0Ω

R17

3k

3C17

10p

C18

10p

C20

100n

C104

100n

C22

100n

C19

220p

C2315p

R11

33

k

C21

22p

R18

1k

L9

0µH33

L11

0µH22

L8

4µH7

L10

4µH7T1

BF961

C83

220p

T2

BF961

L12

L13

1

3

2

45M15AU

MB501-L

IC4

OUT

MC

IN

IN

4

6

2

5

1

8

L15

1

4

2

3 5

SFR455J

L16

1

4

2

3 5

A55GGP

R20

2K2

R19

2K2

C26

10n

C28

10n

L14

LMC4101

1

2

4

5

3

C25

100n

R23

5k

6

R24

5k

6

R68

3k

3

R44

27

0Ω

R47

56

0Ω

R45

33

k

R50

47

k

R51

56

Ω

R21

10k

R25

10k

R22

10k

R26

10k

R48

47k

R49

12k

R43

47k

C27

10n

C29

10n

D15

BA182

D17

BA182

D16

D18

C31

100n

C32

100n

C30

10n

T5

BF245C

C67

100n

C64

150p

D24

BB509

C66

100n

D25

BAT82

L21

MC33171

IC32

3

6

7

41

5

C72

10n

C68

5p6

T6

BFR91

R46

15k

C70

100n

C71

100n

C76

100n

C77

1n

C79

100n

C82

100n

C74

220n

C69

100n

C75

1n

C24

56p

C65

220n

R52

33

k

R552k

2

C73

1µ 16V

R54

82k

R53

47Ω

C78

1n

C100

10µ 63V

L23

100µH

C80

100p

C81

40p

X3

1MHz

SCLK

SDATA

SENABLE

C33

100n

C34

100n

C35

100n

R27

5k

6

C39

1n

C37

100n

C41 1n

C46

3n3

C38

100n

X1

44.545MHz

C40

4p7

TCA440

IFOUTRFIN

IFDEC

MULIN MULIN

IC1

MIXO IFINMIXO

IFIN

RFIN MOUT

OSC AGC

AGC

1415 1216

13

11

108

71

6

5

9

24 3

L17

0µH56

C42

22p

C36

100n

C44

4µ716V

C45

2µ216V

R29

39k

R28

8k

2

R30

12

k

R31

22

0k

R32

2k

2

R33

2k

2

R34

82

k

R36

47

kR35

82k

R37

22kR42

22

k

R59

10

k R60

47

k

R63

1k

D20

BAT85

D19

BAT

L18C47

47p

C48

100n

C50

10n

C51

100p

C52

2n2

T3

BF245C

C49

100p

L19

YMCS17105R2

D21

BAT85

D22BAT85

NE612

IC2OUTA

OUTB

INA

INB

OSC OSC

1

2 5

67

4

8

3

T4BS170

R38

22k

C58

100n

C60

10n

C61

22n

C62

3n3

C63

100n

C57

100n

C59

470p

C55

100n

R39

22

k

C54

100n

X2

CSB455A

R40

15k

R41

330k

L20

1mH

L22

1mH

P4

50k

D23

BB509

85

10k

P3

M1

0mA1. . .1mA5

T7BS170

LMC4101

C84

100n

R56

47

kC87

100n

T8BS170C85

100n

R57

47

kC88

100n

T9BS170C86

100n

R58

47

kC89

100n

C90

22n

C91

4n7

R61

56

0k

R62

3k9

T6

BC549C

C92

10n

C93

3n9

C94

1n

C98

100n

R64

12k

R65

47k

C95

10n

P5

50k

LM386-3

IC62

3

5

6

4

1

7

8 C99

220µ16V

8 Ω

LS1

1W

C97

1µ516V

C96

220µ 16V

78L05

IC7

78L09

IC8

C103

10µ63V

C101

10µ63V

C102

10µ63V

S +

9V

5V9V9V

9V

9V5V

9V

9V

12V

12V

9V

5V+12V

LS1

AM SSB FM

MIXER

3kHz

12kHz

WIDE

NARROW

VFO

*

SYNTHESIZER9V

WIDE

NARROW

UAM

UAM

USSB

UFM

SSB

USSB

UFM

2x

2x

980084 - 12

*

0...9V

0...2V8

0V

2V

1

2V1

2V9 1V8

9V

*

0V2

1V7

0V7

0V2

1V

0...8V

0V

0V

5V5 4V6

1V2

0V5

2V3

2V4

2V5

0V/5V

2V42V

0V2/4V8

9V

1V5

1V5

2V

1V5

2V 2V

8V

2V

0...

0V5

0...

0V6

*

0...0V3

SIGNAL

0V

0V

1V4

1V4

C56

1n

4V3

3V3

4V3

4V9

0VBFO

0V0V

0V 0V/5V 0V/5V5V/0V

0V

2V

3V5

1V

4

0V

1V4

1V4

6V

6V

zie tekst*see text*siehe Text*voir texte*

1st

C43

100n

5V8

C53

10p

AN

5V

0V

D1...D12 = BA479S

0µH56

PRESELECTOR RF-GAIN

4T

3T

4V8/0V2

MC145156-2

PDOUT

IC5DATA

FIN

CLK

SW2

SW1

OUT

RA1OSC

MC

10

11

12

13EN

15

14

19

IN

18

6

8

5

71

R66

1 Ω

BS170

D

G

S

BF245

G D

S

BF961

G2

G1

D

S

2

Figure 2. Practical circuit of the RF sec-tions of the general coverage receiver.Most of the functions defined in theblock diagram will be easy to find backin this schematic.

an additional low-pass filter with two‘notch’ sections, L9-C17 and L11-C18,for virtually complete suppression(–50 dB) of image frequencies and out-of band products.

1st mixer and synthesizerIn many up-market SW receivers, a

double-balanced type (DBM) isemployed as the first mixer to guaran-tee excellent large-signal behaviour.The main dis-advantages ofa passive DBMare the highlevel of the LO

signal (typically 7 dBm), and the inher-ent conversion loss of about –7 dB. Thepresent receiver employs a DG MOS-

FET in the firstmixer. Asopposed to aDBM, theM O S F E T


IC4b

SRG4

C1/

10

1D7

6

9

3

4

5

R

IC4a

SRG4

C1/

2

1D15

14

1

11

12

13

R

IC3b

SRG4

C1/

10

1D7

6

9

3

4

5

R

IC3a

SRG4

C1/

2

1D15

14

1

11

12

13

R

K1

K2

K4

K3

K5

S0

S1

S2

S3

S4

S5

S6

S7

S8

S9

S10

S11

S6

S1

S0

S5

S4

S8

S10

S9

S11

S7

S3

S2

S0

S1

S2

S3

S4

S5

S6

S7

S8

PIC16

IC1

OSC1

MCLR

OSC2

F84-04/P

RA4

RA1

RA0

RA2

RA3

RB0

RB1

RB2

RB3

RB4

RB5

RB6

RB7

17

18

13

12

11

10

16

14

15

1

3

9

8

7

6

2

4

5

R1

4k

7

R2

15

k

R4

15

k

R3

15

k

R8

22

k

R9

10

k

C2

100n

C1

27p

C3

100n

C4

100n

C5

100n

C14

100n

C9

100p

C8

100n

C10

100n

C12

100n

C13

100n

C6

100n

C7

100n

D1

1N4148

CTRDIV10/

IC2

CT=0

CT≥5

74HCT4017

DEC

14

13

15

12

11

10

4

9

6

5

1

7

3

2

& +

0

1

2

3

4

5

6

7

8

9

T1

BS170

R6

10

0Ω

500mW

R7

15

0Ω

D2S2

D3S3

D4S4

D5S5

D6S6

D7S7

D8S8

D9S9

D10S10

D11S11

D12S12

D13S13

S6'

S1'

S0'

S5'

S4'

S8'

S10'

S9'

S11'

S7'

S3'

S2'

KEYB'

10k

P1

S0'

S1'

S2'

S3'

S4'

S5'

S6'

S7'

S8'

S9'

S10'

S11'

KEYB'

ENCODERS1

2

3

1

+ M

– M

+ B

– B

M1 BACKLIGHT7812

IC5

78L05

IC6

D14

1N4001C11

470µ25V

IC3

16

8

IC4

16

8

IC2

16

8

R5

10k

SE

RC

LK

SE

RR

ES

A

B

PRESET

PCLK

KEYB

SCLK

SDATA

SENABLE

SCLK

SDATA

SENABLE

FM

SSB

AM

FM

SSB

AM

+15V

400mA

KEYB

Q0

Q1

Q2

Q3

Q4

Q5

Q0

Q1

Q2

Q3

Q4

Q5

LC

DIS

PL

AY

5V

5V

12V

5V12V

5V

5V

5V

5V

DEN

DE

N

IC3, IC4 = 4015

5V

980084 - 13

+12V

"7"

"3"

"6"

"9"

"#"

"2"

"5"

"8"

"0"

"1"

"4"

" "

3

Figure 3. The microprocessor controlcircuit is based on a PIC16F84. Tokeep receiver-internal interference toa minimum, the PIC is in ‘sleep’mode most of the time.


offers a conversion gain of about 10 dB,and it works fine at a relatively smallLO signal.

The combination of a synthesizer ICtype MC14156-2 (from Motorola) anda dual-modulus (÷128/÷129 or÷64/÷65) divider type MB501L (fromFujitsu) forms a phase-locked loop(PLL) whose step size equals the refer-ence frequency of 1 kHz, which isderived from quartz crystal X3 by anon-chip divider. The MC14156-2 is con-trolled by means of serial informationsupplied by the microprocessor. Theerror signal supplied by the synthe-sizer IC is filtered by a loop filter builtaround an opamp type MC33171 (IC3).Because the 1-kHz reference-frequencycomponent has to be minimized in thefilter, the PLL should allow for a rela-tively long lock time. Here, the largestfrequency change of the local oscillator(45.150 MHz to 77.000 MHz) takesabout 100 ms. Using the single-ended‘PDOUT’ terminal of the MC14156-2allows the loop filter to be kept simple.The MC33171 is used here because it iscapable of supplying a rail-to-railswing of the output voltage. This is amust if the VCO based on FET T5 is tocover the required frequency range(theoretically, 45.15 MHz to 77 MHz)without ‘dying’ as a result of a low var-icap control voltage. In practice, theVCO is slightly overdimensioned, cov-ering a frequency range of 37-85 MHzwith a control voltage of 0-9 V. TheVCO output signal is capacitively cou-pled to the first mixer (T2) as well as toa buffer stage around T6, which isdesigned to drive the ECL inputs ofthe MB501L divider chip.

IF amplifiers, AM/FM demodulatorsand SSB detectorReferring back to the block diagram,the good news is that all sub-circuitsbetween the first IF filter and the out-put of the last IF amplifier are con-tained in a single IC, the TCA440. Thisold faithful from Siemens contains apreamplifier, an oscillator, an IF ampli-fier, and an AGC with a dynamic rangeof no less than 100 dB (which is nomean requirement for SW listening).The two 455-kHz IF filters for narrow(3 kHz BW) and wide (12 kHz BW)reception are connected into and out ofthe TCA440 external circuitry by meansof PIN diodes and control signals sup-plied by the microprocessor. Other fil-ters than the Toko types indicated heremay be used as long as their inputimpedance is 2.2 kΩ, and the respec-tive 3-dB bandwidths are about 3 kHz(narrow) and 12 kHz (wide). TheTCA440 drives the S-meter directly viaits AGC output. Meters with differentsensitivities are accommodated withpreset P3.The injection signal for the secondmixer is supplied by the oscillator

inside the TCA440. This oscillator onlyneeds an external quartz crystal and acouple of passive parts to supply arock-steady 44.545 MHz signal.

The SSB detector is built around thefamiliar NE612 (or NE602), which con-tains a balanced mixer and an oscilla-tor. The latter is connected to an inex-pensive 455-kHz ceramic filter which is‘pulled’ by a varicap, D23. The result-ing deviation of about ±2 kHz is suffi-cient for USB and LSB reception(upper/lower sideband) if you turn theBFO control pot.

The FM demodulator is a classicratio detector with a FET amplifier infront of it. This detector has beendesigned to supply enough outputeven if an NBFM (narrow-band fre-quency modulation) signal is received.NBFM is commonly used in the 27-MHz (11-m) CB band.

The AM demodulator consists of asingle diode, D20, which also suppliesthe AGC drive signal.

The three tuneable inductors in thispart of the circuit are all 455-kHz,ready-made types from Toko. Theseunits contain internal tuning capaci-tors. Other 455-kHz transformers thanthe ones shown here may be used, aslong as the primary-to-secondary turnsratio is 20:1 (in case of L14 and L18),and the tap is exactly at the centre ofthe primary (in case of L19).

Audio signal sectionsThree BS170 FETs are used as analogueswitches, feeding either the FM, AM orSSB signal to filter/amplifier T10. Thecontrol signals at the gates of the FETsare, again, supplied by the micro-processor circuit. The audio bandfilteris designed for speech at radio com-munications quality, i.e., roll-off pointsare defined at 450 Hz and 3.3 kHz tokeep out most unwanted noise, and inthe case of SSB, neighbouring stations.The LM386 audio amplifier, finally,supplies about 1 watt into 8 ohms,which is good for a small external loud-speaker in your shack, or a pair of low-impedance headphones (preferred byveteran DXers).

M I C R O C O N T R O L L E RS E C T I O NThe schematic of the microcontrollersection in the receiver is given sepa-rately in Figure 3. This circuit also con-tains most of the power supply com-ponents.

The microcontroller used is thefamiliar PC16F84 from Microchip.Here, it executes a user program ofabout 1 kBytes from its on-chip ROM.The PIC controller is supplied ready-programmed by the Publishers.

The on-chip EEPROM is used tostore and retain frequencies. Becausethe processor clock does not have to beparticularly stable or accurate, the

cheapest clock option, an R-C network(R1-C1), is used. The processor runs atabout 4 MHz, however, it is only‘active’ when its action is required, forexample, when a key is pressed, or thesynthesizer has to be reloaded. To keepspurious signals to a minimum in thereceiver, the PIC will be ‘asleep’ mostof the time!

Three of the four shift registers type4015 expand the I/O functionality ofthe PIC into a 12-bit shift registerwhich is used to drive the keyboardand the LCD. The keyboard is not amatrix type. As indicated by the circuitdiagram, each switch has a separateconnection, while the other goes to a‘common’ rail. Pressing a key causes aninterrupt which serves both as a wake-up call and a service request for the‘sleeping’ processor. Turning the rotaryencoder also generates a hardwareinterrupt and causes the processor towake up. The encoder used here is aBourns type with 24 turns per full rota-tion. It enables the complete tuningrange of the receiver to be covered —just keep turning until the LCD showsthe desired frequency, and then care-fully adjust the preselector for bestreception. Alternatively, type thedesired start frequency into the key-pad, and tune from there. The rotaryencoder is connected directly to twoPIC I/O pins. Debouncing is effectedby hardware and software.

The remaining I/O pins of the PICare used to control the serial synthe-siser (RB5, RB6, RB7), and the preselec-tor, by way of decimal counter IC2(RB2, RB3).

The power supply is conventionallybased on 3-pin fixed voltage regulatorsfrom the 78 and 78L series. Three volt-ages are supplied: 12 V, two times 5 V,and 9 V. The latter and one of the 5-Vsupplies are part of the main receivercircuit discussed above (refer back toFigure 2). They obtain their input volt-age from the 12-V regulator on themicroprocessor board. The heaviestloads on the 12-V rail are obviously theaudio power amplifier IC, the S-meterlighting and the LCD backlight (ifused). The unstabilized input voltageshould be at least 15 V. An inexpensivepower mains adaptor may be used, butdo note that the receiver may draw upto 450 mA, so go for a relatively pow-erful adaptor.

(980084-1)

The construction, adjustment and operation ofthe receiver will be discussed in next month’sconcluding instalment.

EST & MEASUREMENT

RF decibel meterwideband with large measuring range

An RF decibelradio workshcexpensive, an

sensitive and/orovercome by a d

available: aenables an ac

structed. A few sn

Some parametersFrequency range 100 kHz -110 MHz with an error < 1 dB

100 kHz - 200 MHz with an error 2 dB

Decibel range 32 -117 dBp with an error at 10 MHz 1 dB

Scaling 10 mVdB-1

Input impedance 50 n

(or power) meter is an indispensable instrument in any)p. Unfortunately, accurate, wideband models are fairlyd home -constructed ones are generally not sufficientlyare very temperature -dependent. These drawbacks areevice from Analog Devices which has recently becomelow-cost DC -500 MHz, 92 dB logarithmic amplifier thatcurate, not too expensive RF decibel meter to be con -all modifications make the meter also suitable for low -

frequency measurements.I Design by P Bolch


An RF decibel (or power) meter is an indispensable instrument in anyradio workshop. Unfortunately, accurate, wideband models are fairlyexpensive, and home-constructed ones are generally not sufficiently

sensitive and/or are very temperature-dependent. These drawbacks areovercome by a device from Analog Devices which has recently become

available: a low-cost DC–500 MHz, 92 dB logarithmic amplifier thatenables an accurate, not too expensive RF decibel meter to be con-

structed. A few small modifications make the meter also suitable for low-frequency measurements.

26

Design by P. Bolch

RF decibel meterwideband with large measuring range

Some parametersFrequency range 100 kHz – 110 MHz with an error <1 dB

100 kHz – 200 MHz with an error ≤ 2 dB

Decibel range 32 – 117 dBµ with an error at 10 MHz ≤ 1 dB

Scaling 10 mV dB–1

Input impedance 50 Ω

TEST & MEASUREMENT

C I R C U I T D E S C R I P T I O NThe circuit diagram of the decibelmeter in Figure 1 stands out by its sim-plicity, which is due to the TypeAD8307 monolithic demodulating log-arithmic amplifier, IC1, from AnalogDevices.

The measurand (quantity to bemeasured) is applied to pin 8 (INP) ofIC1 via input socket K1 and capacitorC1. The capacitor ensures that no directvoltage can reach the IC. The secondinput of the IC, pin 1 (INM) is linked tothe earth line via capacitor C4. The val-ues of C1 and C4 are chosen to give alower limit of the frequency rangebelow 100 kHz.

Resistors R1 and R2 ensure that theinput impedance of the meter is theusual value in RF equipment of 50 Ω. Aparallel network is used to minimizeany parasitic properties of the resistors.It is recommended to useSMT (surface mount tech-nology) resistors.

Since the resistors arein parallel with the inputterminals, any direct volt-age present on the inputsignal will cause a poten-tial drop across them. If this causes aproblem, a coupling capacitor of about0.02 µF may be inserted between theinput socket and the resistors, but thiswill restrict the frequency range toabout 30 MHz.

The output of IC1 is essentially acurrent that causes a potential dropacross a 12.5 kΩ internal resistor whichis available at output pin 4. Series net-work R6-P1 is in parallel with the inter-nal resistance to modify the scale fac-tor, which is 25 mV dB–1 in the absenceof an external circuit.

Capacitor C5 averages the outputsignal to ensure a stable display. Itsvalue depends on the application: alarger capacitance gives a more stable,but slow, display; a smaller value is rec-ommended for fast sweeping.

Preset P2 permits parallel shifting ofthe characteristic to give an attenuationof up to 14 dB or an amplification of upto 26 dB between the input socket andpin 8 of IC1, provided that R5 = 0.Resistor R5 provides a narrowing ofthis preset range.

The purpose of resistor R4 is todecouple the output of IC1 from theremainder of the circuit and soenhance the response ratio of small sig-nals.

Owing to the high output imped-ance of IC1, buffer amplifier IC3 isessential to enable a low-impedanceload such as a moving-coil meter to belinked to the circuit.

Regulator IC2 ensures a stable sup-ply line for IC1. Low-pass filter R3-C2reduces any interference on the supplyline.

A small modification enables the cir-

cuit to be used as a low-frequency decibelmeter. Resistors R1 andR2, as well as capacitorsC1 and C4, are then not

used. Instead, pin 8 of IC1 is linked viaa parallel network of a 10 µF, 10 V tan-talum capacitor and a 4.7 kΩ resistor inseries with a 680 pF capacitor to theinput socket, while pin 1 is connectedto earth via an identical series-parallelnetwork. Also, capacitor C5 must be

replaced by a 1 µF, 10 V tantalumcapacitor (+ve terminal to pin 4).Finally, a 1 µF, 10 V tantalum capacitormust be fitted between pin 3 of IC1(+ve terminal) and earth. When thismodification is carried out, the meter isno longer usable as an RF decibelmeter, of course.

D I S P L A YThe display may be a digital multimeter,but, although this is accurate, it is not


CA3140

IC3

2

3

6

7

41

8

5

R3

10Ω0

R2

10

0Ω

R1

10

0Ω

R6

5k

62

R7

R5

R4

10Ω0

C1

10n

C2

100n

C5

100n

C4

10n

25k

P2

5k

P1

AD8307

IC1+IN

OUT–IN

OFS

INT

ENB

8

7

2

41

6

3

5

M1+

K1

IC2

78L05

C3

2µ2 10V

C6

10µ 63V

990008 - 11

*

*

zie tekst*siehe Text*see text*voir texte*

5V > 9V1

Figure 1. The circuitdiagram of the deci-bel meter is centredon the AD8307 fromAnalog Devices.

Figure 2. The decibel meter is best built on thisprinted-circuit board which is, however, notavailable ready made.

990008-1(C) ELEKTOR C3

C5

C6

IC1

IC2

IC3

P1

P2

R3

R4

R5R6

R7

T

M1 +

-0

+

990008-1

C1

C2

C4

R1

R2

990008-1(C) ELEKTOR2

Parts listResistors:R1, R2 = 100 Ω, SMDR3, R4 = 10.0 ΩR5, R7 = see textR6 = 5.62 kΩP1 = 5 kΩ (4.7 kΩ) multiturn upright

preset potentiometerP2 = 25 kΩ multiturn upright preset

potentiometer

Capacitors:C1, C4 = 0.01 µF, SMDC2 = 0.1 µF, SMD

C3 = 2.2 µF, 10 V, tantalumC5 = 0.1 µF. metallized polyesterC6 = 10 µF, 63 V, tantalum

Integrated circuits:IC1 = AD8307AN (Analog Devices)IC2 = 78LO5IC3 = CA3140E

Miscellaneous:K1 = 50 Ω BNC socket for board

mountingEnclosure

used. Keep all wiring as short as possi-ble, however.

If operation up to 30 MHz only isneeded, IC1 may be inserted in asocket, but for use at higher frequen-cies the circuit should be soldereddirectly on to the board. This is bestdone after all other components havebeen fitted and the board has beenchecked thoroughly. This measure is toprotect the AD8307, since this is not acheap component.

Since the meter is an RF unit, it isclear that it should be fitted in anearthed metal enclosure. The powersupply should, of course, not be fittedin the same enclosure. Another impor-tant aspect is that the 9–15 V supplyvoltage should be ‘clean’. It is advisableto use feedthrough capacitors at thepower line inputs and measurementoutput.

C A L I B R A T I O NThe meter circuit should be calibratedwith a suitable RF signal generator or,in an emergency, an AF signal genera-tor with calibrated attenuator.

Apply a signal at a frequency of10 MHz and a level of 60 dBµ (1 mVr.m.s.) to the input of the meter circuit.Using a digital multimeter, measure thevoltage at pin 3 of IC3, increase orreduce the output of the signal gener-ator by exactly 10 dB and turn P1 tocause a change in the multimeter read-ing of 100 mV. The absolute value ofthe output voltage is not significant.

Next, apply a signal at a level ofexactly 60 dBµ to pin 8 of IC1 and turnP2 until the meter indicates 600 mV.

If the requisite equipment is avail-able, the calibration process can berepeated at a number of frequenciesfor greater versatility of operation.

If a signal generator is not to hand,adjust P1 until the resistance betweenits wiper and earth is 1383 Ω measuredwith a digital multimeter. Finally,adjust P2 to obtain a voltage of 1.627 Vat pin 5 of IC1, again measured with adigital multimeter.

S O M E P R O P E R T I E SFigure 3 shows the transfer character-istic of the decibel meter at 10 MHz.The measurement range with an errorsmaller than 1 dB extends from about30 dBµ to around 115 dBµ. Over a largepart of this range, the error stays wellbelow 0.5 dB. The error rises rapidlyoutside the measurement range, whichis typical of the conversion process inIC1.

When the frequency is increased


900

1000

1100

1200

10 20 30 40 50 60 70 80 90 100 110 120

800

700

600

500

400

300

200

100

0

input voltage

outp

ut v

olta

ge

[mV

]

level meter AD8307 (f = 10 MHz)

990008 - 13c

3Figure 3. Transfercharacteristic of thedecibel meter at10 MHz.

Level (dBµ)

101520253035404550556065707580859095

100105110115120125

Level (dBm)

–97–92–87–82–77–72–67–62–57–52–47–42–37–32–27–22–17–12–7–2+3+8

+13+18

Uo(10 MHz) (mV)

281282285294312353397450497550596650695750795847895948994

10491090114311851218

Uo(110 MHz) (mV)

282283285294313356400450496544590641686737783833881933980

10331078113211781188

The output at 200 MHz for an input of 99 dBµ is 948 mVThe output at 300 MHz for an input of 100 dBµ is 942 mV

easily calibrated.A moving coil metering network

with series resistor R7 facilitates recog-nizing any drift such as encountered,for instance, during calibration, butdoes not make reading it easy.

Measurements with sweep fre-quencies can, of course, be displayedon an oscilloscope.

The decibel meter outputs a directvoltage that is directly proportional tothe input signal. The display is cali-brated in dBµ (decibel referred to1 microvolt). The scale factor is

100 mV dB–1, so that an input signal of100 dBµ results in an output voltage of1 V.

C O N S T R U C T I O NThe meter circuit is best built on theprinted-circuit board shown in Fig-ure 2, but this is not available readymade. As mentioned earlier, some ofthe components should be SMDs (sur-face mount devices) as specified in thecomponents list. If the circuit is con-structed on prototyping board, stan-dard components may, of course, be

(the well -calibrated prototype can beused up to 110MHz), the transfer char-acteristic shifts slightly downwards,but retains the linearity shown in Fig-ure 3. A number of characteristic valuesat 10 MHz and 110 MHz are given inTable 1.

If measurements over only a limitedrange are needed, the frequency -dependent slight shift of the transfercharacteristic may be negated duringthe calibration so that a slightly moreprecise meter reading is obtained.

Measurements carried out with theprototype at frequencies of 200 Mhzand 300 MHz at a stable input level of100 d13µ show that the circuit may beused without any problem at these fre-quencies.

[990008]

Figure 4. The finishedprototype decibelmeter.

AD8307The Type AD8307 monolithic logarithmic amplifier isintended for a number of applications, among which

Conversion of signal level to decibel form

Transmitter antenna power measurement

Receiver signal strength indication (RSSI)

Network and spectrum analysers (up to 120 dB)

Signal level determination down to 20 Hz

True decibel AC mode for multimeters

Its operation is based on the progressive compression(successive detection) technique, providing a dynamicrange of 92 dB to ± 3 dB law -conformance and 88 dB toa tight ± 1 dB error bound at all frequencies up to100 MHz.

The device is very stable and easy to use. It needs asupply voltage of 2.7-5.5 Vat 7.5 mA, corresponding to alow power consumption of 22.5 mW at 3 V A fast -actingCMOS-compatible control pin can disable the AD8307 toa standby current of not more than 150 µA.

Each of the cascaded amplifier/ limiter cells has a smallsignal gain of 14.3 dB with a -3 dB bandwidth of 900 MHz.

The input is fully differential at a moderately high imped-ance (1.1 k52 in parallel with 1.4 pF).

The device provides a basic dynamic range extendingfrom about -75 dBm (decibel referred to 1 mW) to around+ 17 dBm. A simple input -matching network can lower thisrange to -88 dBm to + 3 dBm. The logarithmic linearity istypically within 0.3 dB up to 100 MHz over the central por-tion of this range, and is degraded only slightly at500 MHz. There is no minimum frequency limit: theAD8307 may be used at audio frequencies down to DC.

The output is a voltage -scaled 25 mV dB -1, generated bya nominal current of 2 µA dB -1 through an internal 12.5 k52resistor. This voltage varies from 0.25 V at an input of-74 dBm (that is, the a.c. intercept is at -84 dBm, a 20 µVr.m.s. sinusoidal input), up to 2.5 V from an input of+ 16 dBm. This slope and intercept can be trimmed withexternal adjustments. For instance, with a 2.7 V supply, theoutput scaling may be lowered to 15 mVdB-1 to permit uti-lization of the full dynamic range.

The AD8307 has good supply insensitivity and temper-ature stability of the scaling parameters. The combinationof low cost, small size, low power consumption, highaccuracy and stability, large dynamic range, and a fre-quency range from DC to UHF make it useful in numerousapplications requiring the conversion of a signal to its deci-bel equivalent.

Further information on the Internet:www.analog.com/AD8307

Elektor Electronics 1/99 29

(the well-calibrated prototype can beused up to 110 MHz), the transfer char-acteristic shifts slightly downwards,but retains the linearity shown in Fig-ure 3. A number of characteristic valuesat 10 MHz and 110 MHz are given inTable 1.

If measurements over only a limitedrange are needed, the frequency-dependent slight shift of the transfercharacteristic may be negated duringthe calibration so that a slightly moreprecise meter reading is obtained.

Measurements carried out with theprototype at frequencies of 200 Mhzand 300 MHz at a stable input level of100 dBµ show that the circuit may beused without any problem at these fre-quencies.

[990008]


AD8307The Type AD8307 monolithic logarithmic amplifier isintended for a number of applications, among which

• Conversion of signal level to decibel form

• Transmitter antenna power measurement

• Receiver signal strength indication (RSSI)

• Network and spectrum analysers (up to 120 dB)

• Signal level determination down to 20 Hz

• True decibel AC mode for multimeters

Its operation is based on the progressive compression(successive detection) technique, providing a dynamicrange of 92 dB to ±3 dB law-conformance and 88 dB toa tight ±1 dB error bound at all frequencies up to100 MHz.

The device is very stable and easy to use. It needs asupply voltage of 2.7–5.5 V at 7.5 mA, corresponding to alow power consumption of 22.5 mW at 3 V. A fast-actingCMOS-compatible control pin can disable the AD8307 toa standby current of not more than 150 µA.

Each of the cascaded amplifier/ limiter cells has a smallsignal gain of 14.3 dB with a –3 dB bandwidth of 900 MHz.

The input is fully differential at a moderately high imped-ance (1.1 kΩ in parallel with 1.4 pF).

The device provides a basic dynamic range extendingfrom about –75 dBm (decibel referred to 1 mW) to around+17 dBm. A simple input-matching network can lower thisrange to –88 dBm to +3 dBm. The logarithmic linearity istypically within 0.3 dB up to 100 MHz over the central por-tion of this range, and is degraded only slightly at500 MHz. There is no minimum frequency limit: theAD8307 may be used at audio frequencies down to DC.

The output is a voltage-scaled 25 mV dB–1, generated bya nominal current of 2 µA dB–1 through an internal 12.5 kΩresistor. This voltage varies from 0.25 V at an input of–74 dBm (that is, the a.c. intercept is at –84 dBm, a 20 µVr.m.s. sinusoidal input), up to 2.5 V from an input of+16 dBm. This slope and intercept can be trimmed withexternal adjustments. For instance, with a 2.7 V supply, theoutput scaling may be lowered to 15 mV dB–1 to permit uti-lization of the full dynamic range.

The AD8307 has good supply insensitivity and temper-ature stability of the scaling parameters. The combinationof low cost, small size, low power consumption, highaccuracy and stability, large dynamic range, and a fre-quency range from DC to UHF make it useful in numerousapplications requiring the conversion of a signal to its deci-bel equivalent.

Further information on the Internet:www.analog.com/AD8307

Figure 4. The finishedprototype decibelmeter.

TEST & MEASUREMENT

conductance testerwith DIY sensor

Conductivity in the senseit is used in this article is

the ability of a substancesuch as water to conduct

electric current. It isexpressed in terms of

current per unit of appliedvoltage. It is the reciprocal

of resistivity. Conduc-tance is the reciprocal of

resistance and is mea-sured in siemens. It is

therefore the ratio of thecurrent through a sub-

stance to the potential dif-ference at its ends.Thetester described in this

article is intended forassessing the quality of

water, based on the acid-ity or alkalinity (pH), by

means of a measurementof the conductance of the

water.

Design by P Baer

INTRODUCTIONWater with a very high pH is not goodfor fish, plants or making tea or coffee.This is the reason that many aquariumowners, orchid growers, horticulturists,and many others use distilled or fil-tered water. Water filters are very pop-ular in domestic use, where the qualityof tap water is suspect. However, waterfilters themselves present a risk of ger-mination, requiring good attention tocleanliness (bottle needs thoroughwashing at least once a week).

A very environment -friendly wayof obtaining low -pH water is the use ofrainwater, but this depends heavily onthe area where the rainwater is col-lected. Such water may be tested forlow pH, that is, low conductivity, withthe present tester.

The pH of water is a logarithmicindex of the hydrogen -ion concentra-tion in the water. It is given by

pH= log10(1/[H+ 1)

where [H+] is the hydrogen -ion con-centration. A pH below 7 indicates acid-ity and one above 7, alkalinity, at 25 °C.

CONDUCTIVITY ANDCONDUCTANCEConductivity (or specific conductance),being the reciprocal of resistivity, ismeasured in the same way as resis-tance and expressed in S m-1 (siemensper metre); its symbol is u. At constanttemperature, the value of conductance,symbol G, of a substance depends onthe cross-sectional area, A, in m2 , thelength, 1, in m, and the conductivity, u,in S m-1:

G= o21/1. S (siemens)

This equation can be used with a solidas well as with a liquid substance.

The sensor used in the presenttester consists of two annular elec-trodes having a cross-sectional area of1 cm which are spaced 1 cm apart.These dimensions make the calculationof the conductance of the water beingtested straightforward.

Pure water, sold as distilled water, asused, for instance, in electric irons, lead -acid batteries and for horticultural pur-poses, has a conductivity of 1 x 10-3S m-1, so that the present meter would


I N T R O D U C T I O NWater with a very high pH is not goodfor fish, plants or making tea or coffee.This is the reason that many aquariumowners, orchid growers, horticulturists,and many others use distilled or fil-tered water. Water filters are very pop-ular in domestic use, where the qualityof tap water is suspect. However, waterfilters themselves present a risk of ger-mination, requiring good attention tocleanliness (bottle needs thoroughwashing at least once a week).

A very environment-friendly wayof obtaining low-pH water is the use ofrainwater, but this depends heavily onthe area where the rainwater is col-lected. Such water may be tested forlow pH, that is, low conductivity, withthe present tester.

The pH of water is a logarithmicindex of the hydrogen-ion concentra-tion in the water. It is given by

pH=log10(1/[H+])

where [H+] is the hydrogen-ion con-centration. A pH below 7 indicates acid-ity and one above 7, alkalinity, at 25 °C.

C O N D U C T I V I T Y A N DC O N D U C T A N C EConductivity (or specific conductance),being the reciprocal of resistivity, ismeasured in the same way as resis-tance and expressed in S m–1 (siemensper metre); its symbol is σ. At constanttemperature, the value of conductance,symbol G, of a substance depends onthe cross-sectional area, A, in m2 , thelength, l, in m, and the conductivity, σ,in S m–1:

G=σA/ l. S (siemens)

This equation can be used with a solidas well as with a liquid substance.

The sensor used in the presenttester consists of two annular elec-trodes having a cross-sectional area of1 cm which are spaced 1 cm apart.These dimensions make the calculationof the conductance of the water beingtested straightforward.

Pure water, sold as distilled water, asused, for instance, in electric irons, lead-acid batteries and for horticultural pur-poses, has a conductivity of 1×10–3

S m–1, so that the present meter would

Conductivity in the senseit is used in this article isthe ability of a substancesuch as water to conduct

electric current. It isexpressed in terms of

current per unit of appliedvoltage. It is the reciprocal

of resistivity. Conduc-tance is the reciprocal of

resistance and is mea-sured in siemens. It is

therefore the ratio of thecurrent through a sub-

stance to the potential dif-ference at its ends.Thetester described in this

article is intended forassessing the quality of

water, based on the acid-ity or alkalinity (pH), by

means of a measurementof the conductance of the

water.

30

Design by P. Baer

conductance testerwith DIY sensor

TEST & MEASUREMENT

1

9V

9V

P1

10k

R1

100n

31

7

38

39

21 20

2

19

LCD1 3.5 DIGIT

DP3 DP2 DP1

0

AC

, 's

OOF CI 0 D4 . D4 D4X

I

2P

0 01 I 01 01 BP

K

0

r0 / \ 0 / \ 0 P BAT

1/4-

io' t2 2 2 2 2 Z't fV. 2 2 8 m¢ (o o. . 0 0 . <2

22

31

7

9 10

15

11

24

29

16

30

23

27

25

26

13

13

14

4

9

5 24 25

2

23

7

22

6

17 18

2

19

3

20

4

21

5

37

30

E'd 0¢ O ,2 2 2 2 2 Z't

IN HI

TEST

IN LO

o o 0

tT W m 8

IC1

ICL7106

21'3 c'et (.D U. Al 13 CI m<V+

REF HI

REF LO

COMM

V-

28

40

6

C4

6-110µ

1C3 330n

63V

35

32

40 39 38

R9C7

100p

measure 10,uS (equivalent to a resis-tance of 100 k.(2). When the waterbecomes less pure, the resistancebetween the electrodes drops, and thevalue of conductance rises. Normal tapwater has a conductance of about1 mS, and sea water, 100 mS or more.

THE TESTERFrom the above, it is clear that thetester must be capable of measuringresistance or conductance. This seemssimple enough: take a constant currentsource, insert the sensor in a potentialdivider and apply the voltage acrossthe sensor via an analogue -to -digitalconverter (ADC) to a suitable display.

Unfortunately, the reality is not sosimple, since the resistance of a fluidmust be measured with an alternatinginstead of a direct current. This isbecause a direct current would causeelectrolysis which after a while woulddistort the sensor electrodes.

The tester, whose circuit diagram isshown in Figure 1, therefore uses a rec-tangular current. This enables the ADCto drive the liquid -crystal display(LCD) via pin 21, the backplane voltageterminal.

The backplane voltage is used toswitch, via transistor T2, constant -cur-rent sink T3 -D2 -R6, which is combinedwith constant -current source T4 -D3 -R7.Capacitor C4 is charged and dischargedin rhythm with the backplane voltage ata rate of 100 mA. Because of thisarrangement, the 100 µA current fromthe source is absorbed by the sink whenthe capacitor is being discharged. Con-sequently, an alternating current of± 100 µA flows through the tester, whichcauses a potential drop of ± 100 µV 52-1across the resistance, that is, the water.

34

100n

33

C9

470n

29 28 27

R10C10

220n

26

The test voltage is taken from acrossR2 -R3 -C3. During the negative half -period of the backplane voltage, tran-sistor T1 links the test voltage to earth.In essence, therefore, this PET operatesas a clocked synchronous rectifier with-out a threshold voltage and conse-quent non -linearity.

The direct test voltage so obtainedis directly proportional to the resis-tance of the water. It must, however, beinverted to provide a test display insiemens. In the present circuit this isachieved by applying the test voltageto the reference voltage input (REFHI)of IC1 and a constant voltage to thetest input pins (IN HI and IN LO). Thisresults in the display showing Uc/UT,instead of, as normal, U/Uref. (Uc is theconstant voltage across pins 30, 31, UT

R2

1M

220n

R3

T1

C141411

100n BF245A

Eri

BC557B

T3

elpR6

R5

Ean C4

980104 - 11

D3

R8

D2

Figure 1. Circuit diagram of theconductance meter.

is the test voltage, and U f is the refer-ence voltage). Preset P1 provides com-pensation for component and sensortolerances.

THE SENSORTo make the sensor, two rings of brassor other easily soldered metal, a 15 cmlength of RG58U coaxial cable, and aheavy-duty soldering iron are needed.The brass rings should have an inner

2

Figure 2. Parts required for con-structing the sensor.

solder braidhere

disc of brassor other easilysoldered metal

RG 58 U

cover withsilicon cement

d .10 mm

-I% A

solder conductorof coax cable here 06 mm

141 013 mm


101

980104-12


measure 10 µS (equivalent to a resis-tance of 100 kΩ). When the waterbecomes less pure, the resistancebetween the electrodes drops, and thevalue of conductance rises. Normal tapwater has a conductance of about1 mS, and sea water, 100 mS or more.

T H E T E S T E RFrom the above, it is clear that thetester must be capable of measuringresistance or conductance. This seemssimple enough: take a constant currentsource, insert the sensor in a potentialdivider and apply the voltage acrossthe sensor via an analogue-to-digitalconverter (ADC) to a suitable display.

Unfortunately, the reality is not sosimple, since the resistance of a fluidmust be measured with an alternatinginstead of a direct current. This isbecause a direct current would causeelectrolysis which after a while woulddistort the sensor electrodes.

The tester, whose circuit diagram isshown in Figure 1, therefore uses a rec-tangular current. This enables the ADCto drive the liquid-crystal display(LCD) via pin 21, the backplane voltageterminal.

The backplane voltage is used toswitch, via transistor T2, constant-cur-rent sink T3-D2-R6, which is combinedwith constant-current source T4-D3-R7.Capacitor C4 is charged and dischargedin rhythm with the backplane voltage ata rate of 100 mA. Because of thisarrangement, the 100 µA current fromthe source is absorbed by the sink whenthe capacitor is being discharged. Con-sequently, an alternating current of±100 µA flows through the tester, whichcauses a potential drop of ±100 µV Ω–1

across the resistance, that is, the water.


LCD1

BAT

OF

32

G3

31

F3 E3

10

D3

11C

329

B3

30

A3

27

G2

26

F2

13

E2

14

D2

15

C2

24

B2

25

A2

23

G1

22

F1

17

E1

18

D1

19

C1

20

B1

21

A1

40BP

39

3828

DP3

12

DP2

16

DP1

3.5 DIGIT

2P

K

X

9

1

3

2Y

8

10k

P1

R1

47

0k

R9

10

0k

R104

7k

R44M7

R3

1M

R7

20

k

R6

10

k

R8

22

k

ICL7106

REF HI

REF LO

IN LO

IN HI

OS

C 1

OS

C 2

OS

C 3

C R

EF

C R

EF

BU

FF

TEST

IC1

COMM

F 3

PO

L

INT

A/Z

A1

B1

D1

E1F1G1

12

A2

11

B2

10

C2

D2

14

E2

13

F2

25

G2

23A

316

B3

24

C3

15

D3

18

E3

1722

G3

19

AB

20

30

31

26

36

35

40 39 38 272829

37

3334

C1

BP

21

32

5428679

1V

V

3

C7

100p

C9

470n

C10

220n

C1

100n

C6

100n

C3

220n

C11

100n

C8

100n

C5330n

C4

330n

D3

D2

R5

100k

R2

1M

C2

10µ63V

T1

BF

D1

1N4148

T2

BC547B

T3

T4

BC557B

BT1

9V

BC547B

245A

980104 - 11

Probe

9V

9V

9V

37AC

1

Figure 1. Circuit diagram of theconductance meter.

980104 - 12

RG 58 U

solder braidhere

disc of brassor other easilysoldered metal



solder conductorof coax cable here

d = 10 mm

6 mm13 mm

2

Figure 2. Parts required for con-structing the sensor.

The test voltage is taken from acrossR2-R3-C3. During the negative half-period of the backplane voltage, tran-sistor T1 links the test voltage to earth.In essence, therefore, this FET operatesas a clocked synchronous rectifier with-out a threshold voltage and conse-quent non-linearity.

The direct test voltage so obtainedis directly proportional to the resis-tance of the water. It must, however, beinverted to provide a test display insiemens. In the present circuit this isachieved by applying the test voltageto the reference voltage input (REF.HI)of IC1 and a constant voltage to thetest input pins (IN HI and IN LO). Thisresults in the display showing UC/UT,instead of, as normal, U/Uref. (UC is theconstant voltage across pins 30, 31, UT

is the test voltage, and Uref is the refer-ence voltage). Preset P1 provides com-pensation for component and sensortolerances.

T H E S E N S O RTo make the sensor, two rings of brassor other easily soldered metal, a 15 cmlength of RG58U coaxial cable, and aheavy-duty soldering iron are needed.The brass rings should have an inner

3 l&r,,-)/er-t 11,\(

m n&kar-azt,v2,»»»: PAW.: ..14

diameter of 6 mm andan outer one of 13 mmto give them an effec-tive area of just under1 cm2. The inner diameter allows themto just fit over the inner conductor ofthe coaxial cable, whose outer insula-tion must be removed over a length ofabout 15 mm from one end and 10 mmof the exposed braid cut off. Theremaining 5 mm of braid must befolded back over the outer insulationof the cable. This ensures that the twobrass rings are about 1 cm apart (seeFigure 2). The inner core of the coaxialcable is then soldered to the outer brassring and the braid tothe inner ring. Theouter surfaces of therings (but not those fac-

Figure 4. Thpleted protboard.

C9

C6P1

Figure 3. The printed -circuit board for theconductance tester.

eotype

I

T3

R5

D1

in g each other) shouldthen be covered withsilicon cement.

CONSTRUCTIONThe remainder of the tester is best con-structed on the printed -circuit boardshown in Figure 3. Mind the polarityof the diodes and electrolytic capaci-tors. The IC should be soldered directlyto the board to allow the display to befitted directly above it.

Connect a standard potentiometeracross the test inputs and check that thedisplay shows corresponding conduc-tance values when the potentiometer is

turned from, say, 10k52to 1 k1.

When all is well, fitthe completed board

corn -

Parts listResistors:

= 470 kf2.R2, R3 = 1 MD.R4= 4.7 MD.R5, R9 = 100 kflR6= 10 kf2.R7 = 20 kf2.1:18 = 22 kf2.

Rlo = 47 kf2.P1 = 10 kf2 preset potentiometer

Capacitors:Ci, 06, 08, Cii = 0.1 [IF02 = 10 [IF, 63 V, radialC3, 010 = 0.22 [IFC4, C5 = 0.33 [IF07= 100 pFC9 = 0.47 [IF

Semiconductors:D1 = 1N4148D2, D3 = LED, green, 3 mm

= BF254A12, T3 = BC547B14 = BC557B

Integrated circuits:ICi = ICL7106CPL (Maxim)

Miscellaneous:LCD1 = 3.5 digit liquid -crystal dis-

play (note that ICi and LCD1 areavailable as a set)

BT1 = 9 V dry battery with clip1 off switch with on contactEnclosure as appropriatePCB Order no. 980104-1 (see Read-ers Services towards the end ofthis issue)

Conductivity at 20 eC

Silver 1.6 x 10-8 S m-1

Copper 1.7

Aluminium 2.8Tungsten 5.6Nickel 6.8Iron 10

Steel 18

Manganin 44

Carbon 3500

into a suitable enclosure in which acut-out for the display has been pro-vided. Connect the sensor to the probeterminals as shown in Figure 4. Do notforget an on/off switch.

FINALLY ...The tester has a range of 50 µS, whichcorresponds to a resistance of 20 k52and a maximum test voltage at IC1 of2 V This value will be displayed whenthe tester, or rather, the sensor, is dry.The upper limit of the test range(1999 µS) is set by the characteristics ofICi. Note also that the basic error of 5per cent increases slightly when thetest range is given an upper limit ofmore than 1000 µS. [980104]


diameter of 6 mm andan outer one of 13 mmto give them an effec-tive area of just under1 cm2. The inner diameter allows themto just fit over the inner conductor ofthe coaxial cable, whose outer insula-tion must be removed over a length ofabout 15 mm from one end and 10 mmof the exposed braid cut off. Theremaining 5 mm of braid must befolded back over the outer insulationof the cable. This ensures that the twobrass rings are about 1 cm apart (seeFigure 2). The inner core of the coaxialcable is then soldered to the outer brassring and the braid tothe inner ring. Theouter surfaces of therings (but not those fac-

ing each other) shouldthen be covered withsilicon cement.

C O N S T R U C T I O NThe remainder of the tester is best con-structed on the printed-circuit boardshown in Figure 3. Mind the polarityof the diodes and electrolytic capaci-tors. The IC should be soldered directlyto the board to allow the display to befitted directly above it.

Connect a standard potentiometeracross the test inputs and check that thedisplay shows corresponding conduc-tance values when the potentiometer is

turned from, say, 10 kΩto 1 kΩ.

When all is well, fitthe completed board

into a suitable enclosure in which acut-out for the display has been pro-vided. Connect the sensor to the probeterminals as shown in Figure 4. Do notforget an on/off switch.

F I N A L L Y …The tester has a range of 50 µS, whichcorresponds to a resistance of 20 kΩand a maximum test voltage at IC1 of2 V. This value will be displayed whenthe tester, or rather, the sensor, is dry.The upper limit of the test range(1999 µS) is set by the characteristics ofIC1. Note also that the basic error of 5per cent increases slightly when thetest range is given an upper limit ofmore than 1000 µS. [980104]

980104-1

(C) ELEKTOR C1

C2 C3

C4

C5

C6

C7C8

C9C10

C11

D1

D2

D3

FIX2

FIX

3

FIX4

H1

IC1

LCD1

P1

R1

R2

R3

R4

R5

R6

R7

R8

R9R10

T1

T2

T3

T4

1

1

+9V 0Probe

98

01

04

-1

98

01

04

-1(C

) ELE

KTO

R

Figure 3. The printed-circuit board for theconductance tester.

Figure 4. The com-pleted prototypeboard.

3Parts listResistors:R1 = 470 kΩR2, R3 = 1 MΩR4 = 4.7 MΩR5, R9 = 100 kΩR6 = 10 kΩR7 = 20 kΩR8 = 22 kΩR10 = 47 kΩP1 = 10 kΩ preset potentiometer

Capacitors:C1, C6, C8, C11 = 0.1 µFC2 = 10 µF, 63 V, radialC3, C10 = 0.22 µFC4, C5 = 0.33 µFC7 = 100 pFC9 = 0.47 µF

Semiconductors:D1 = 1N4148D2, D3 = LED, green, 3 mmT1 = BF254AT2, T3 = BC547BT4 = BC557B

Integrated circuits:IC1 = ICL7106CPL (Maxim)

Miscellaneous:LCD1 = 3.5 digit liquid-crystal dis-

play (note that IC1 and LCD1 areavailable as a set)

BT1 = 9 V dry battery with clip1 off switch with on contactEnclosure as appropriatePCB Order no. 980104-1 (see Read-

ers Services towards the end ofthis issue)

Conductivity at 20 °C

Silver 1.6 × 10–8 S m–1

Copper 1.7Aluminium 2.8Tungsten 5.6Nickel 6.8Iron 10Steel 18Manganin 44Carbon 3500

POWER SUPPLIES

DC -DCstep-up converter

no -iron converter for mobilecharging of low -power battery packs

The voltage step-up converter described inthis article is a transformerless designbased on just one integrated circuitand a handful of passive parts. Effi-

ciency is excellent given the sim-plicity of the circuit, which requiresno modifications for any input volt-

age between 6 V and 12 V, for outputvoltages of about 10 V and 22 V

respectively.

Design by W. Zeiller

One of the most frequently used appli-cations of voltage step-up converters isthat of a battery charger using the 12-V vehicle battery as its input powersource. After all, to charge a battery,you need a voltage which is alwaysgreater than the maximum voltagesupplied by the battery when fullycharged. So, charging a 12-V NiCd bat-tery pack as used in, say, a portablemobile radio or a laptop computerfrom the car battery calls for a circuitthat increases (`steps up') the 12-Vinput voltage to, say, 20 V or so whichmay be applied to the charger circuit.

Not so long ago, it was practicallyimpossible to design DC -DC step-upconverters without recourse to specialtransformer techniques using theinverter principle: use the input volt-age to power an oscillator which drivesa step-up transformer; next, rectify thehigh voltage at the secondary. Such cir-

cu its are typically bulky and not terri-bly efficient, although there are notice-able exceptions.

Today, most step-up converters aretailor-made switch -mode power sup-plies (SMPSUs) based on purpose -designed ICs. The design presentedhere is an exception in that it employsa low-cost audio power amplifier IC,the TDA2822M.

STEP UP THE VOLUMELooking at the circuit diagram in Fig-ure 1 you will not fail to note the sim-plicity of the circuit. Basically, theinputs and outputs of the two ampli-fiers in the TDA2822M are cross -cou-pled by capacitors C2 and C7 to causea (controlled) amount of oscillation. Infact, you are looking at a double AMV(astable multivibrator) acting as a pu sh-pull oscillator/charge pump driving aclassic diode -based voltage multiplier.


One of the most frequently used appli-cations of voltage step-up converters isthat of a battery charger using the 12-V vehicle battery as its input powersource. After all, to charge a battery,you need a voltage which is alwaysgreater than the maximum voltagesupplied by the battery when fullycharged. So, charging a 12-V NiCd bat-tery pack as used in, say, a portablemobile radio or a laptop computerfrom the car battery calls for a circuitthat increases (‘steps up’) the 12-Vinput voltage to, say, 20 V or so whichmay be applied to the charger circuit.

Not so long ago, it was practicallyimpossible to design DC-DC step-upconverters without recourse to specialtransformer techniques using theinverter principle: use the input volt-age to power an oscillator which drivesa step-up transformer; next, rectify thehigh voltage at the secondary. Such cir-

cuits are typically bulky and not terri-bly efficient, although there are notice-able exceptions.

Today, most step-up converters aretailor-made switch-mode power sup-plies (SMPSUs) based on purpose-designed ICs. The design presentedhere is an exception in that it employsa low-cost audio power amplifier IC,the TDA2822M.

S T E P U P T H E V O L U M ELooking at the circuit diagram in Fig-ure 1 you will not fail to note the sim-plicity of the circuit. Basically, theinputs and outputs of the two ampli-fiers in the TDA2822M are cross-cou-pled by capacitors C2 and C7 to causea (controlled) amount of oscillation. Infact, you are looking at a double AMV(astable multivibrator) acting as a push-pull oscillator/charge pump driving aclassic diode-based voltage multiplier.

The voltage step-up converter described inthis article is a transformerless designbased on just one integrated circuitand a handful of passive parts. Effi-

ciency is excellent given the sim-plicity of the circuit, which requiresno modifications for any input volt-

age between 6 V and 12 V, for outputvoltages of about 10 V and 22 V

respectively.

40

Design by W. Zeiller

DC-DCstep-up converter

no-iron converter for mobilecharging of low-power battery packs

POWER SUPPLIES

Simple as it may be, the circuit acts asa reasonably efficient voltage doubler(theoretically, that is).

Through their output capacitors (C4and C9) and associated diode pairs(D1-D2 and D3-D4), amplifiers IC1aand IC1b alternately contribute to theenergy (charge) built up in outputcapacitor C10. This energy is availablefor use by the load connected to theconverter output terminals.

Theoretically, the input voltage isdoubled, but there are derating factors.Firstly, the output transistors of theTDA2822M are not ideal devices andcause a small voltage loss. Add to thatthe voltage drop across the diodes andyou will appreciate that an input volt-age of 12 V produces an output voltageof just 22 V instead of the theoreticallyexpected 24 V. Unfortunately, the out-put voltage drops a little more whenthe converter is actually loaded, butthat will not be a problem in most bat-tery chargers thanks totheir internal regulatorcircuits (for constant cur-rent or constant voltage).

The oscillator operatesat a frequency of about2 kHz. This valuedepends to some extenton the actual supply volt-age and the load current.The Boucherot networks at the ampli-fier outputs, R3-C3 and R4-C8, maycome as a surprise here because theytypically occur in audio amplifierswere they serve to ‘straighten’ loud-speaker impedances. Here, the mainpurpose of the networks is to stabilizethe converter when the diodes areswitching.

C O N S T R U C T I O NThe circuit is best built on a printed cir-cuit board of which the copper tracklayout and component-mounting planare given in Figure 2. Constructionshould be a piece of cake, the boardbeing single-sided, and only common-or-garden components are used. Domake sure, however, that the followingparts are mounted the right wayaround on the board:- electrolytic capacitors C4, C9, C10,

C11;- diodes D1, D2, D3 and D4;- integrated circuit IC1.Having finished the solder work youshould subject the board to a thoroughvisual inspection, and correct any obvi-ous errors before you power up for thefirst time.

If difficult to obtain locally, the typeSB130 diodes may be replaced byalmost any other medium-powerSchottky diode capable of passing atleast 1 A. In the prototype, the well-known BYW29 was tried and found togive good results, too.

Finally, be sure to use the TDA2822M

only. Because of its 8-pin DIL enclosure,it is the only version that can be used onthis printed circuit board.

P E R F O R M A N C EThe maximum continuous output cur-rent that can be supplied will be about300 mA. The no-load current con-sumption of the converter is between6 and 8 mA. A prototype of the con-verter was put through its paces in ourdesign laboratory, with the followingresults:

Uin Iin Uout Iout Efficiency6 V 0.22 A 10 V 0.1 A 80%12 V 0.44 A 21.3 V 0.21A 85%

Not spectacular, but not bad either forsuch a simple design! (980073-1)


C1

22n

C6

22n

C5

47n

C2

22n

C3

100n

C8

100n

C7

22n

C4

470µ25V

C9

470µ25V

C11

1000µ16V

C12

100nR1

33k

R2

33k

R3

4Ω

7

R4

4Ω

7

D1

SB130

D3

SB130

D2

D4

C10

1000µ16V

12V12V

2x

2x

12V

980073 - 11

TDA2822MIC1

2

6

4

8

7

1

5

3

1

Figure 1. This transformerless DC-DCstep-up converter is based on a stereopower amplifier IC, the TDA2822M. Here,the two opamps are wired as a cross-coupled double AMV driving a traditionaldiode-capacitor voltage doubler. Switch-ing frequency is about 2 kHz.

Figure 2. Copper tracklayout and componentoverlay of the printed cir-cuit board designed forthe converter (board notavailable ready-made).

(C) ELEKTOR

980073-1

C1

C2

C3

C4

C5

C6

C7

C8

C9

C10

C11

C12

D1

D2

D3

D4

H1

H2

H3

H4

H6

IC1

R1

R2

R3

R498

00

73

-1

+ 00+

(C) ELEKTOR

98

00

73

-1

2

COMPONENTS LIST

Resistors:R1,R2 = 22kΩR3,R4 = 4Ω7

Capacitors:C1,C2,C7,C7 = 22nFC3,C8,C12 = 100nFC4,C9 = 470µF 25V radialC5 = 47nFC10 = 470µF 40V radialC11 = 1000µF 16V radial

Semiconductors:D1-D4 = SB130, BYR745 or BYW29IC1 = TDA2822M (SGS Thomson)

GENERAL INTEREST

RadioTextRadio Teletextnow also in Europe

Some European broadcasting organizationshave introduced, or are introducing, an

improved version of the Radio Teletext serviceinaugurated in the United Kingdom by a num-

ber of commercial radio stations in the late1980s. Like the UK Radio teletext service,

RadioText is separate from, but in addition to,RDS (Radio Data Services). The RadioText data

are carried on a 76 kHz sub -carrier (UK Radioteletext: 69 kHz), FM (FSK) modulated at

16 Kbit/s (UK: 5 Kbit/s) with the FM carrier devi-ated by ± 5.5 kHz (UK: ± 7.5 kHz). Sony has

designed a receiver, Textman, specially for thereception of the RadioText service.

By our Editorial Staff

INTRODUCTIONThe Radio Data System (RDS) is aEuropean system for broadcasting dig-ital data on VHF/FM transmitters. Thespectrum of the VHF/FM signal isshown in Figure 1. Receivers equip-pedwith RDS decoders decode the digitaldata which may include automatic tun-ing, station identification, service iden-tification, accurate time, a radio pagingservice, and much other information,.

The RDS specification was adoptedby the European Broadcasting Union(EBU) in 1984 and is now implementedon most European VHF/FM broadcasttransmitters.

Information at a data rate of1.1875 Kbit/s in groups of 104 bits issuperimposed on to a 57 kHz sub -car-rier locked to the 19 kHz stereo pilottone. The final multiplex (L+ R signal,19 kHz pilot tone, L-R signal on a sup-pressed 38 kHz sub -carrier, and theRDS signal) frequency modulates themain carrier.

Each of the groups of 104 bits intowhich the bit stream is divided is sub-divided into four blocks of 26 bits. Eachblock contains a 16 -bit informationword and a 10 -bit checkword for,


I N T R O D U C T I O NThe Radio Data System (RDS) is aEuropean system for broadcasting dig-ital data on VHF/FM transmitters. Thespectrum of the VHF/FM signal isshown in Figure 1. Receivers equip-pedwith RDS decoders decode the digitaldata which may include automatic tun-ing, station identification, service iden-tification, accurate time, a radio pagingservice, and much other information,.

The RDS specification was adoptedby the European Broadcasting Union(EBU) in 1984 and is now implementedon most European VHF/FM broadcasttransmitters.

Information at a data rate of1.1875 Kbit/s in groups of 104 bits issuperimposed on to a 57 kHz sub-car-rier locked to the 19 kHz stereo pilottone. The final multiplex (L+R signal,19 kHz pilot tone, L–R signal on a sup-pressed 38 kHz sub-carrier, and theRDS signal) frequency modulates themain carrier.

Each of the groups of 104 bits intowhich the bit stream is divided is sub-divided into four blocks of 26 bits. Eachblock contains a 16-bit informationword and a 10-bit checkword for,

Some European broadcasting organizationshave introduced, or are introducing, an

improved version of the Radio Teletext serviceinaugurated in the United Kingdom by a num-

ber of commercial radio stations in the late1980s. Like the UK Radio teletext service,

RadioText is separate from, but in addition to,RDS (Radio Data Services). The RadioText data

are carried on a 76 kHz sub-carrier (UK Radioteletext: 69 kHz), FM (FSK) modulated at

16 Kbit/s (UK: 5 Kbit/s) with the FM carrier devi-ated by ±5.5 kHz (UK: ±7.5 kHz). Sony has

designed a receiver, Textman, specially for thereception of the RadioText service.

42


RadioText Radio Teletextnow also in Europe

GENERAL INTEREST

among others, error detection and adegree of error correction.

DATA RADIO CHANNELIn the early 1990s, the American com-pany Digital DJ introduced a specialdata service, called Data Radio Chan-nel - DARC - for VHF/FM broadcasts.The system was developed by theNH K Science & Research Laboratoriesin Tokyo. Note that this is not based onthe American Subsidiary Communica-tions Authorization (SCA) system thatdates from the early 1980s and onwhich the UK Radio Teletext service isbased. The system has been acceptedby the International Telecommunica-tions Union (ITU) as a standard.

It is interesting to note that theDARC occupies a band of 16 Kbit/s,which is more than ten times as wideas the RDS channel (1.1875 Kbit/s).

The DARC is broadcast in the USAto support normal radio programmeswith Programme Associate Data (PAD).This service enables programme infor-mation, such as names of performers,cast lists, station identification, and soon, to be displayed on the liquid -crystaldisplay (LCD) of a suitable receiver or,by means of appropriate software, onthe monitor of a computer.

DARC TECHNOLOGY.In the DARC, data are carried on ahigh-speed sub -carrier of 76 kHz-seeFigure 1.

Information at a data rate of16 Kbit/s is superimposed on to the76 kHz sub -carrier which, being thethird harmonic of the 19 kHz stereopilot tone, is readily locked to this tone.The final multiplex,including the L+ R sig-nal, the 19 kHz pilottone, the L-R stereo dif-ference signal on a sup-pressed 38 kHz sub -car-rier, and the DARC sig-nal, frequencymodulates the main car-rier at ± 5.5 kHz.

The modulation fac-tor is 0.1. The level of thehigh-speed sub -carriermay be increased to-20 dB to reduce anycrosstalk from the maincarrier.

After error correc-tions and other neces-sary manipulations, thebandwidth of 16 Kbit/s isreduced to 8 Kbit/s, viawhich about 1000alphanumeric charactersper second can be trans-mitted.

The data superim-posed on to the high-speed sub -carrier is gen-erated via a computer

1

AmplitudeStereo pilot tone

'S' channelL - R L - R

RDS

t I t15 kHz 23 kHz 53 kHz

19 kHz 38 kHz 57 kHz

running Workbenchof Digital DJ. Afterencoding, the signal issent in accordancewith the U SEP emula-tion protocol to themain transmitters viaa microwave networkand a DARC encoder-see Figure 3.

1

RADIO TEXTMost data on the RadioText service arederived from the local television's tele-text service but, although this latter ser-vice contains thousands of pages,RadioText will be limited to a few hun-dred. These pages will be transmittedagain and again,repeated every fewminutes. The Textmanor other suitablereceiver stores all theinformation before itcan be seen on the dis-

Figure 1. Spectrum of aVHF/FM signal that con-tains the normal stereoaudio signals, the RDSdata channel, and theData Radio Channel-DARC.

HSS (high speed subcarrier)

I' 176 kHz 95 kHz

Frequency990016 - 11

play. This means thatthe waiting timeswhen leafing throughpages on TV Teletextare not encounteredwith RadioText.Because of the errorcorrection, pages can-

not be mutilated: a page is receivedeither in good order or not at all.

The resolution of the LCD on theTextman receiver is seven lines each ofnot more than 21 alphanumeric char-acters. Graphics, such as logos, mayalso be displayed. All information is, ofcourse, in black and white.

In Europe, as in the UK and USA,RadioText is primarilyof interest to commer-cial broadcasters, sincethe service is free (sofar) to the recipientand must, therefore,be paid for by adver-

I.

Figure 2. Typical broad-casting control roomwhere data can be pre-pared for disseminationvia the VHF/FM radionetwork. I


among others, error detection and adegree of error correction.

D A T A R A D I O C H A N N E LIn the early 1990s, the American com-pany Digital DJ introduced a specialdata service, called Data Radio Chan-nel – DARC – for VHF/FM broadcasts.The system was developed by theNHK Science & Research Laboratoriesin Tokyo. Note that this is not based onthe American Subsidiary Communica-tions Authorization (SCA) system thatdates from the early 1980s and onwhich the UK Radio Teletext service isbased. The system has been acceptedby the International Telecommunica-tions Union (ITU) as a standard.

It is interesting to note that theDARC occupies a band of 16 Kbit/s,which is more than ten times as wideas the RDS channel (1.1875 Kbit/s).

The DARC is broadcast in the USAto support normal radio programmeswith Programme Associate Data (PAD).This service enables programme infor-mation, such as names of performers,cast lists, station identification, and soon, to be displayed on the liquid-crystaldisplay (LCD) of a suitable receiver or,by means of appropriate software, onthe monitor of a computer.

D A R C T E C H N O L O G Y.In the DARC, data are carried on ahigh-speed sub-carrier of 76 kHz—seeFigure 1.

Information at a data rate of16 Kbit/s is superimposed on to the76 kHz sub-carrier which, being thethird harmonic of the 19 kHz stereopilot tone, is readily locked to this tone.The final multiplex,including the L+R sig-nal, the 19 kHz pilottone, the L–R stereo dif-ference signal on a sup-pressed 38 kHz sub-car-rier, and the DARC sig-nal, frequencymodulates the main car-rier at ±5.5 kHz.

The modulation fac-tor is 0.1. The level of thehigh-speed sub-carriermay be increased to–20 dB to reduce anycrosstalk from the maincarrier.

After error correc-tions and other neces-sary manipulations, thebandwidth of 16 Kbit/s isreduced to 8 Kbit/s, viawhich about 1000alphanumeric charactersper second can be trans-mitted.

The data superim-posed on to the high-speed sub-carrier is gen-erated via a computer

running Workbenchof Digital DJ. Afterencoding, the signal issent in accordancewith the USEP emula-tion protocol to themain transmitters viaa microwave networkand a DARC encoder—see Figure 3.

R A D I O T E X TMost data on the RadioText service arederived from the local television’s tele-text service but, although this latter ser-vice contains thousands of pages,RadioText will be limited to a few hun-dred. These pages will be transmittedagain and again,repeated every fewminutes. The Textmanor other suitablereceiver stores all theinformation before itcan be seen on the dis-

play. This means thatthe waiting timeswhen leafing throughpages on TV Teletextare not encounteredwith RadioText.Because of the errorcorrection, pages can-

not be mutilated: a page is receivedeither in good order or not at all.

The resolution of the LCD on theTextman receiver is seven lines each ofnot more than 21 alphanumeric char-acters. Graphics, such as logos, mayalso be displayed. All information is, ofcourse, in black and white.

In Europe, as in the UK and USA,RadioText is primarilyof interest to commer-cial broadcasters, sincethe service is free (sofar) to the recipientand must, therefore,be paid for by adver-


Amplitude

Frequency

Stereo pilot tone

L - R L - R'S' channel

RDS

15 kHz19 kHz 38 kHz 57 kHz

23 kHz 53 kHz 76 kHz 95 kHz

990016 - 11

HSS (high speed subcarrier)

1

Figure 1. Spectrum of aVHF/FM signal that con-tains the normal stereoaudio signals, the RDSdata channel, and theData Radio Channel—DARC.

Figure 2. Typical broad-casting control roomwhere data can be pre-pared for disseminationvia the VHF/FM radionetwork.

3

tisers.

WorkbenchDigital DJ

USEP

Uni-directional

However, theTextman is already pro-vided with facilities that enable ser-vices to be paid for before they can bedisplayed.

The RadioText service may eventu-ally also be used for radio paging, butat present there are no definite plansfor this. Nevertheless, each Textmanreceiver is already pro-vided with a uniquecode, which can beaccessed via the menus

Figure 4.SRF-DR2

DARCEncoder

DARCEncoder

Figure 3. Diagram of abroadcasting networkthat transmits Radio -Text data.

Transmitter

Transmitter

Transmitter

990016 - 12

of the receiver.

TEXTMAN SRF-DR2000The only receiver currently commer-cially available for the reception ofRadioText is the Sony TypeSRF-DR2000 Textman. The ICs used inthe receiver are Sony designs, whichare not (yet) available to other OEDs.

When they become,the RadioText facilitymay also be built intomain s -operated tuners

The Type000 Textman

receiver from Sony.

and car radios.The Sony receiver looks like the

familar Walkman TM with controls forfive preset transmitter. It has a blackand white display with associatedmenu controls. Its internal memoryenables important messages to bestored for some time.

A drawback of the receiver is thatan earpiece has to be used since thelead of this forms the antenna withoutwhich, of course, the receiver worksbadly or not at all.

The receiver may be used all overEurope and, with a suitable configura-tion menu, in North America.

There is an automatic power -offfacility which ensures that the batteriesare not exhaused when the user for-gets to switch off the receiver.

An FM data control on the frontpanel enables the FM data decoder tobe disabled when no data services areused. If the receiver is used as a stan-dard portable receiver, the datadecoder is switched off automaticallyafter twenty minutes.

FINALLYIf the RadioText information is dissem-inated via low -power local transmitter,the system may be used to give driversparking information in car parks, ordraw the attention of customers in adepartment store to special offers, andso. At music festivals, visitors may beadvised via radio of the names of theperformers and, indeed, what is beingperformed. [990016]


tisers. However, theTextman is already pro-vided with facilities that enable ser-vices to be paid for before they can bedisplayed.

The RadioText service may eventu-ally also be used for radio paging, butat present there are no definite plansfor this. Nevertheless, each Textmanreceiver is already pro-vided with a uniquecode, which can beaccessed via the menus

of the receiver.

T E X T M A N S R F - D R 2 0 0 0The only receiver currently commer-cially available for the reception ofRadioText is the Sony TypeSRF-DR2000 Textman. The ICs used inthe receiver are Sony designs, whichare not (yet) available to other OEDs.

When they become,the RadioText facilitymay also be built intomains-operated tuners

and car radios.The Sony receiver looks like the

familar Walkman™ with controls forfive preset transmitter. It has a blackand white display with associatedmenu controls. Its internal memoryenables important messages to bestored for some time.

A drawback of the receiver is thatan earpiece has to be used since thelead of this forms the antenna withoutwhich, of course, the receiver worksbadly or not at all.

The receiver may be used all overEurope and, with a suitable configura-tion menu, in North America.

There is an automatic power-offfacility which ensures that the batteriesare not exhaused when the user for-gets to switch off the receiver.

An FM data control on the frontpanel enables the FM data decoder tobe disabled when no data services areused. If the receiver is used as a stan-dard portable receiver, the datadecoder is switched off automaticallyafter twenty minutes.

F I N A L L YIf the RadioText information is dissem-inated via low-power local transmitter,the system may be used to give driversparking information in car parks, ordraw the attention of customers in adepartment store to special offers, andso. At music festivals, visitors may beadvised via radio of the names of theperformers and, indeed, what is beingperformed. [990016]

Microwavenetwork

WorkbenchDigital DJ

USEP

USEP

Uni-directional

DARCEncoder

Transmitter

DARCEncoder

Transmitter

990016 - 12

DARCEncoder

Transmitter

3

Figure 3. Diagram of abroadcasting networkthat transmits Radio-Text data.

Figure 4. The TypeSRF-DR2000 Textmanreceiver from Sony.

GENERAL INTEREST

The Iridium NetworkMobile telephony via satellite

Proliferation of mobilephones is already a fact,

but worldwide mobile tele-phone communication hasbecome possible only with

the advent of the IridiumNetwork. The network is a

joint venture between USorganizations Motorola

and Lockheed Martin Cor-poration, and the RussianKrunichev State Research

Space Centre. The firstsatellites were launchedtowards the end of lastyear; when fully opera-tional, the network will

embrace 66 satellites invarious orbits around theearth at a height of some

780 kilometres (470miles). This fairly low orbitis dictated by the relatively

low power of mobilephones.


INTRODUCTIONIridum is the first truly worldwidefunctioning satellite telephone system.The name 'Iridium' was chosenbecause the original concept of thenetwork envisaged 77 satellites andiridium, a metallic element of the plat-inum family, is number 77 in the Peri-odic System. Although the final net-work will consist of 66 satellites (plussix spare), the name was retained,probably because the element at num-ber 66 of the Periodic System, dyspro-sium, a member of the rare-earthgroup, is a rather more difficult name

to remember.Basically, the system works as fol-

lows. A call from a mobile phone istransmitted to the nearest Iridium satel-lite, from where it is relayed to anothersatellite in the system which is nearestthe call's final destination. From this,the signal is beamed down to anothercellular phone or a traditional line tele-phone.

Not many people may have noticedit, but the first Iridium satellite wastaken into operational use a couple ofmonths ago. With this, a concept that atits birth n years ago was dubbed `sci-


I N T R O D U C T I O NIridum is the first truly worldwidefunctioning satellite telephone system.The name ‘Iridium’ was chosenbecause the original concept of thenetwork envisaged 77 satellites andiridium, a metallic element of the plat-inum family, is number 77 in the Peri-odic System. Although the final net-work will consist of 66 satellites (plussix spare), the name was retained,probably because the element at num-ber 66 of the Periodic System, dyspro-sium, a member of the rare-earthgroup, is a rather more difficult name

to remember.Basically, the system works as fol-

lows. A call from a mobile phone istransmitted to the nearest Iridium satel-lite, from where it is relayed to anothersatellite in the system which is nearestthe call’s final destination. From this,the signal is beamed down to anothercellular phone or a traditional line tele-phone.

Not many people may have noticedit, but the first Iridium satellite wastaken into operational use a couple ofmonths ago. With this, a concept that atits birth 11 years ago was dubbed ‘sci-

Proliferation of mobilephones is already a fact,

but worldwide mobile tele-phone communication hasbecome possible only with

the advent of the IridiumNetwork. The network is a

joint venture between USorganizations Motorola

and Lockheed Martin Cor-poration, and the RussianKrunichev State Research

Space Centre. The firstsatellites were launched

towards the end of lastyear; when fully opera-tional, the network will

embrace 66 satellites invarious orbits around theearth at a height of some

780 kilometres (470miles). This fairly low orbitis dictated by the relatively

low power of mobilephones.

46


The Iridium Network Mobile telephony via satellite

GENERAL INTEREST

ence fiction' by many, became reality.Many aspects of the network are

new (at least as far as non-military useis concerned).

Instead of some large geostationarysatellites, it uses a fairly large num-ber of smaller ones in low orbits.All satellites are in communicationwith each other and so form a trulyworldwide network.Mobile phones and pagers any-where in the world can be accesseddirectly by the relevant satellite,which means that the system isindependent of terrestrial infra-stru ctures.The system is compatible with allcurrent terrestrial mobile telephonenetworks, such as GSM (Global Sys-tem for Mobile Communication) inEurope, IS -95 in the USA, PDC inJapan, and others. This means thatfrom now on a GSM user can com-municate with an IS -95 user (whichuntil recently was impossible).An Iridium subscriber can be con-tacted from anywhere in the worldon just one number - which may behis/her current mobile phone num-ber.

LEO/MEO VS GEOUntil Iridium became operational, civilsatellite communications were con-ducted almost exclusively by geosta-tionary (GEO) satellites about 36 000km (almost 22 000 miles) above theearth. Exceptions were, for instance,the television satellites in the formerSoviet Union which, to cover theextreme north of the country, were inhighly elliptical orbit.

The Iridium Network is the firstcivil communication system that usessatellites in low -earth orbit (LEO) andmedium -earth orbit (MEO), which liebetween 700 km (425 miles) and 10 000km (about 6 000 miles).

The difficulty with using high -orbitsatellites is not so much the requisitehigh (transmitter) power or more sen-sitive receivers in the mobile phones(which would become much larger),but rather the signal delay, particularlyin the case of digital equipment. Theonly satellite telephone service prior toIridium was via the geostationaryInmarsat satellite which uses equip-ment the size of a briefcase andrequires the antenna to be directed atthe satellite. As Inmarsat will continuetheir GEO service, they have indicatedthat the user equipment will be drasti-cally reduced in size.

Other planned systems - seeoverview in Table 1- will, however, useLEO and MEO satellites.

The LEO satellites in the IridiumNetwork orbit the earth in about 100minutes. Motorola's future system,Celestri, is planned to use LEO as wellas MEO satellites. This system will pro-

vide video, interactivemultimedia, and dataservices.

Another system of thefuture, Microsoft'sTeledesic, is planned tobecome an Internet satel-lite network.

A quite differentapproach is that ofEllipso, which isintended for communi-cations over the northernhemisphere, supple-menting terrestrial ser-vices. This network willuse highly -elliptical orbit(HEO) satellites, theheight of whose orbitwill vary from 500 km(300 miles) to 8 000 km(almost 5 000 miles).

THE IRIDUMNETWORKThe Iridium systemembraces 66 operationaland six spare satellites. Itis planned to have a life of5-8 years. To enhance thereliability of the system, anumber of additionalsatellites have been putinto orbit, which enable anumber of control centres to monitorthe movements and working of theoperational satellites in their orbits. Themain monitor centre in Landsdown,Virginia, USA, is supported by fixedmonitor stations in North Canada andHawaii, and a transportable one in Ice-land.

Each satellite is in contact in fourdirections with other satlelites in thesystem in the Ka -band (18-30 GHz).

The satellites are linked to the ter-restrial networks by twelve gateways(earth stations) on four continents.Three or four 3 -metre diameter rotaryparabolic antennas track the high-speed (29 000 km/h - 17 500 mph) satel-lites and transfer voice and data infor-mation. Communication between thecontrol satellites and the gateways is inthe Ka band (down -link 19.4-19.6 GHz;

1.F.igure 1. The 66 opera-tional satellites in theIridium Network usesix different orbits toprovide worldwidecommunications.

up -link29.1-29.3 Ghz).

The gateways passthe information tothe relevant terres-trial network (and

convert the used protocol if necessary)and also manage and store any infor-mation for relay to Iridium pagers. TheEuropean gateway is at Fucino in Italy.

Communication between the oper-ational satellites and the providers ofthe mobile phone and pager services isin the L -band: 1616-1625.5 MHz. Thesame frequencies are used for the upand down links, which allows thesame components to be used for trans-mitter and receiver.

As an aside, the use of L -banddownlink frequencies is likely to causeproblems for radio astronomers, sincethey are very close to one of the spec-tral lines (1612 MHz - the others are at1665, 1667, and 1720 MHz)) of thehydroxyl molecule (HO), which is oneof the most important frequencies forradio astronomers.

Name of system Operator Satellites Planned forInmarsat Inmarsat 12 GEO Late 1999

Globalstar Global Star 64 LEO Late 1999ICO ICO Global 12-15 ME0 2000

CommunicationsSpaceway Hughes 12-20 GEO 2000

CommunicationsCelestri Motorola 63 LEO/GEO 2001

Odyssey TRW/Teleglobe 12 MEO 2001

Skybridge Alcatel Alsthom 64 LEO 2001

Teledesic Microsoft 288 LEO 2002Ellipso Mobile 17 HEO 2002

Communications


ence fiction’ by many, became reality.Many aspects of the network are

new (at least as far as non-military useis concerned).• Instead of some large geostationary

satellites, it uses a fairly large num-ber of smaller ones in low orbits.

• All satellites are in communicationwith each other and so form a trulyworldwide network.

• Mobile phones and pagers any-where in the world can be accesseddirectly by the relevant satellite,which means that the system isindependent of terrestrial infra-structures.

• The system is compatible with allcurrent terrestrial mobile telephonenetworks, such as GSM (Global Sys-tem for Mobile Communication) inEurope, IS-95 in the USA, PDC inJapan, and others. This means thatfrom now on a GSM user can com-municate with an IS-95 user (whichuntil recently was impossible).

• An Iridium subscriber can be con-tacted from anywhere in the worldon just one number – which may behis/her current mobile phone num-ber.

L E O / M E O V S G E OUntil Iridium became operational, civilsatellite communications were con-ducted almost exclusively by geosta-tionary (GEO) satellites about 36 000km (almost 22 000 miles) above theearth. Exceptions were, for instance,the television satellites in the formerSoviet Union which, to cover theextreme north of the country, were inhighly elliptical orbit.

The Iridium Network is the firstcivil communication system that usessatellites in low-earth orbit (LEO) andmedium-earth orbit (MEO), which liebetween 700 km (425 miles) and 10 000km (about 6 000 miles).

The difficulty with using high-orbitsatellites is not so much the requisitehigh (transmitter) power or more sen-sitive receivers in the mobile phones(which would become much larger),but rather the signal delay, particularlyin the case of digital equipment. Theonly satellite telephone service prior toIridium was via the geostationaryInmarsat satellite which uses equip-ment the size of a briefcase andrequires the antenna to be directed atthe satellite. As Inmarsat will continuetheir GEO service, they have indicatedthat the user equipment will be drasti-cally reduced in size.

Other planned systems – seeoverview in Table 1 – will, however, useLEO and MEO satellites.

The LEO satellites in the IridiumNetwork orbit the earth in about 100minutes. Motorola’s future system,Celestri, is planned to use LEO as wellas MEO satellites. This system will pro-

vide video, interactivemultimedia, and dataservices.

Another system of thefuture, Microsoft’sTeledesic, is planned tobecome an Internet satel-lite network.

A quite differentapproach is that ofEllipso, which isintended for communi-cations over the northernhemisphere, supple-menting terrestrial ser-vices. This network willuse highly-elliptical orbit(HEO) satellites, theheight of whose orbitwill vary from 500 km(300 miles) to 8 000 km(almost 5 000 miles).

T H E I R I D U MN E T W O R KThe Iridium systemembraces 66 operationaland six spare satellites. Itis planned to have a life of5–8 years. To enhance thereliability of the system, anumber of additionalsatellites have been putinto orbit, which enable anumber of control centres to monitorthe movements and working of theoperational satellites in their orbits. Themain monitor centre in Landsdown,Virginia, USA, is supported by fixedmonitor stations in North Canada andHawaii, and a transportable one in Ice-land.

Each satellite is in contact in fourdirections with other satlelites in thesystem in the Ka-band (18–30 GHz).

The satellites are linked to the ter-restrial networks by twelve gateways(earth stations) on four continents.Three or four 3-metre diameter rotaryparabolic antennas track the high-speed (29 000 km/h – 17 500 mph) satel-lites and transfer voice and data infor-mation. Communication between thecontrol satellites and the gateways is inthe Ka band (down-link 19.4–19.6 GHz;

u p - l i n k29.1–29.3 Ghz).

The gateways passthe information tothe relevant terres-trial network (and

convert the used protocol if necessary)and also manage and store any infor-mation for relay to Iridium pagers. TheEuropean gateway is at Fucino in Italy.

Communication between the oper-ational satellites and the providers ofthe mobile phone and pager services isin the L-band: 1616–1625.5 MHz. Thesame frequencies are used for the upand down links, which allows thesame components to be used for trans-mitter and receiver.

As an aside, the use of L-banddownlink frequencies is likely to causeproblems for radio astronomers, sincethey are very close to one of the spec-tral lines (1612 MHz – the others are at1665, 1667, and 1720 MHz)) of thehydroxyl molecule (HO), which is oneof the most important frequencies forradio astronomers.


Figure 1. The 66 opera-tional satellites in theIridium Network usesix different orbits toprovide worldwidecommunications.

Name of systemInmarsat

GlobalstarICO

Spaceway

CelestriOdyssey

SkybridgeTeledesic

Ellipso

OperatorInmarsat

Global StarICO Global

CommunicationsHughes

CommunicationsMotorola

TRW/TeleglobeAlcatel Alsthom

MicrosoftMobile

Communications

Satellites12 GEO64 LEO

12–15 MEO

12–20 GEO

63 LEO/GEO12 MEO64 LEO

288 LEO17 HEO

Planned forLate 1999Late 1999

2000

2000

20012001200120022002

MOBILE PHONESAND PAGERSThe mobile phones and pagers aremanufactured by Motorola andKyocera. The mobile phones are verysimilar to those used on GSM networksand are able to operate not only withthe Iridium standard, but also on anumber of other standards, such asGSM and IS -95.

The two designs are quite dissimi-lar. The Motorola phone is primarilyintended for use in the Iridium system,which can be expanded by an optionalcassette (which can be slotted into thephone) for use in terrestrial mobilephone services. Each different stan-dard requires a different cassette, forinstance, in Europe, a 900 MHz or1800 MHz GSM model; in the USA,Africa and Russia, an 800 MHz IS -95

version.The phone weighs

about 450 grams. In satel-lite service, its averagetransmit power is645 mW, but there is a linkmargin of about 15.5 dB.

The current price, including one cas-sette, is about $US3,000.

The Kyocera design is quite differ-ent, almost the opposite of theMotorola phone. Its handset, whichweighs only 80 grams, is intended pri-marily for use in terrestrial mobilephone systems. If communication viasatellite is required, the handset isinserted into a special 'docking station':the combination weighs 450 grams.When the user travels from an areawith a different service standard, a sec-ond handset is needed. The dockingstation and batteries can be used irre-spective of which standard prevails.

Figure 2. Assembly of the nearly 700 kg weighing satelliteat the Motorola-SATCOM plant in Chandler, Arizona. Notethe triangular cross-section of the carrier structure whichis made of reinforced man-made fibre.

PAGING VIA IRIDIUMThe Irridium Network is not only thefirst global mobile telephone service,but it also offers the first paging servicethat ensures accessibility anywhere inthe world without a mobile phone orwith the mobile phone switched off.The message sender need not knowwhere the recipient is as this is auto-matically established by the Iridiumsystem (provided, of course, that thepager is switched on). The Iridium Pag-ing Service allows alphanumeric mes-

Figure 3. TheKyocera Iridiumphone cosists of ahandset intended forterrestrial serviceswhich is transformedinto a satellite mobilephone by inserting itinto a 'docking sta-tion' with L -bandantenna.


M O B I L E P H O N E SA N D P A G E R SThe mobile phones and pagers aremanufactured by Motorola andKyocera. The mobile phones are verysimilar to those used on GSM networksand are able to operate not only withthe Iridium standard, but also on anumber of other standards, such asGSM and IS-95.

The two designs are quite dissimi-lar. The Motorola phone is primarilyintended for use in the Iridium system,which can be expanded by an optionalcassette (which can be slotted into thephone) for use in terrestrial mobilephone services. Each different stan-dard requires a different cassette, forinstance, in Europe, a 900 MHz or1800 MHz GSM model; in the USA,Africa and Russia, an 800 MHz IS-95

version. The phone weighs

about 450 grams. In satel-lite service, its averagetransmit power is645 mW, but there is a linkmargin of about 15.5 dB.

The current price, including one cas-sette, is about $US3,000.

The Kyocera design is quite differ-ent, almost the opposite of theMotorola phone. Its handset, whichweighs only 80 grams, is intended pri-marily for use in terrestrial mobilephone systems. If communication viasatellite is required, the handset isinserted into a special ‘docking station’:the combination weighs 450 grams.When the user travels from an areawith a different service standard, a sec-ond handset is needed. The dockingstation and batteries can be used irre-spective of which standard prevails.

P A G I N G V I A I R I D I U MThe Irridium Network is not only thefirst global mobile telephone service,but it also offers the first paging servicethat ensures accessibility anywhere inthe world without a mobile phone orwith the mobile phone switched off.The message sender need not knowwhere the recipient is as this is auto-matically established by the Iridiumsystem (provided, of course, that thepager is switched on). The Iridium Pag-ing Service allows alphanumeric mes-


Figure 2. Assembly of the nearly 700 kg weighing satelliteat the Motorola-SATCOM plant in Chandler, Arizona. Notethe triangular cross-section of the carrier structure whichis made of reinforced man-made fibre.

Figure 3. TheKyocera Iridiumphone cosists of ahandset intended forterrestrial serviceswhich is transformedinto a satellite mobilephone by inserting itinto a ‘docking sta-tion’ with L-bandantenna.

sages of up to 200 characters as well asnumerical ones of up to 20 ciphers. Thepager specifications embrace interna-tional graphic characters and LED dis-plays. There is an optional possibilityof sending a message to groups ofaddresses. The batteries last for about30 days.

OPERATORS ANDPROVIDERSAlthough the design of the equipmentwas carried out by Motorola and thatof the satellites by Lockheed Martin,the Iridium Service is operated by Irid-ium LLC. This is an international con-sortium of communication companiesand other industrial organizations,including Motorola, Lockheed Martin,British Aerospace, INMARSAT, andothers, which have invested in thedevelopment and marketing of theIridium service.

The consortium has representativesin 15 countries all over the world. Atthe time of writing, it had concludedabout 300 agreements with serviceproviders and roaming partners,which together represent more than100 million customers in 122 countries.Further agreements in more than 100countries are confidentally expectedwithin the first few months of this year.

Users of current mobile phones can

obtain information on the Irid-ium service, and possibly rele-vant handsets, from their cur-rent provider. Billing for theIridium service is through theseproviders. Also, they retain theircurrent mobile phone number.People who do not want to useterrestrial mobile phone servicescan take out a satellite -onlymobile phone contract, and thisservice issues numbers startingwith the international code8816.

[990003]

Figure 4. Iridiumpager and mobilephone fromMotorola. Thephone has an inter-changeable fre-quency module fora number of mobilephone standardsand needs an addi-tional L -bandantenna unit foroperation in theIridium Network.

C.) MOTOROLA

08:CONTRACT HAS BEE'APPROVED! THE NUTMEETING IS WECHEDGYMORN IHG

911111111111111111P"'"1-r-4

The Iridium Satellites

Each of the 66 operational and fourteen spare and controlsatellites built so far weighs 689 kilograms (just over 1500pounds) and is 4.5 metres (just under 15 ft) long. Its poweris provided by two wing -shaped panels almost 6 metres(20 ft) long that are fitted with gallium -arsenide solar cells.

The triangular carrier structure is made from reinforcedman-made fibre.

The solar generator charges a 22 -cell nickel -hydrogenbattery, which becomes the power supply when the satel-lite is in the shadow of the earth (about half the 100 minuteorbiting time).

The average transmitter power is 660 W The centralelectronics circuits are based on ASICs (Application Spe-cific ICs). The computing power of the multi -processor net-work is about 100 mips (million instructions per second).At full load, up to 11,000 telephone conversations can behandled simultaneously.

The RF stages of the transmitter and receiver are based

on MMICs (Monolithic Microwave ICs), while the amplifiersin the receiver use low -noise GAS-FETs.

Communication in the Ka -band to other Iridium satel-lites is carried out by four antennas (two North -South andtwo East-West) on frequencies between 23.18 GHz and23.38 GHz.

Direct communication with mobile phones and pagerstakes place in the L -band between 1616 and 1626.5 MHz viathree main antennas and 16 spotbeams.

Communications with the gateways is via four movableantennas in the Ka -band (downlink 19.4-19.6 Ghz anduplink 29.1-29.3 GHz).

All antennas consist of flat panels and use state-of-the-art technology.

As mentioned earlier, the 80 satellites were designedand manufactured by Lockheed Martin. This company hasto date built more LEO satellites than any other organiza-tion in the western world. In spite of all this experience,there were a number of failures, and these set back theoverall program by about six months.

Most of the satellites are launched in groups of five by aBoeing Delta II Rocket from the Vandenberg launch pad inCalifornia.

Some satellites were launched in groups of seven by aRussian Proton Rocket from the Baikonur Kosmodrom sitein Kazakhstan.

Some satellites will be launched by China, but the GreatWalls Industries Corporation's 'Long March 2C/SD' rock-ets can accommodate only two satellites at a time.

The spare satellites will be launched by Delta II andLong March 2C/SD rockets. At the end of last year, seven ofthe 14 spare and control satellites were already in orbit.


sages of up to 200 characters as well asnumerical ones of up to 20 ciphers. Thepager specifications embrace interna-tional graphic characters and LED dis-plays. There is an optional possibilityof sending a message to groups ofaddresses. The batteries last for about30 days.

O P E R A T O R S A N DP R O V I D E R SAlthough the design of the equipmentwas carried out by Motorola and thatof the satellites by Lockheed Martin,the Iridium Service is operated by Irid-ium LLC. This is an international con-sortium of communication companiesand other industrial organizations,including Motorola, Lockheed Martin,British Aerospace, INMARSAT, andothers, which have invested in thedevelopment and marketing of theIridium service.

The consortium has representativesin 15 countries all over the world. Atthe time of writing, it had concludedabout 300 agreements with serviceproviders and roaming partners,which together represent more than100 million customers in 122 countries.Further agreements in more than 100countries are confidentally expectedwithin the first few months of this year.

Users of current mobile phones can

obtain information on the Irid-ium service, and possibly rele-vant handsets, from their cur-rent provider. Billing for theIridium service is through theseproviders. Also, they retain theircurrent mobile phone number.People who do not want to useterrestrial mobile phone servicescan take out a satellite-onlymobile phone contract, and thisservice issues numbers startingwith the international code8816.

[990003]


Each of the 66 operational and fourteen spare and controlsatellites built so far weighs 689 kilograms (just over 1500pounds) and is 4.5 metres (just under 15 ft) long. Its poweris provided by two wing-shaped panels almost 6 metres(20 ft) long that are fitted with gallium-arsenide solar cells.

The triangular carrier structure is made from reinforcedman-made fibre.

The solar generator charges a 22-cell nickel-hydrogenbattery, which becomes the power supply when the satel-lite is in the shadow of the earth (about half the 100 minuteorbiting time).

The average transmitter power is 660 W. The centralelectronics circuits are based on ASICs (Application Spe-cific ICs). The computing power of the multi-processor net-work is about 100 mips (million instructions per second).At full load, up to 11,000 telephone conversations can behandled simultaneously.

The RF stages of the transmitter and receiver are based

on MMICs (Monolithic Microwave ICs), while the amplifiersin the receiver use low-noise GAS-FETs.

Communication in the Ka-band to other Iridium satel-lites is carried out by four antennas (two North-South andtwo East-West) on frequencies between 23.18 GHz and23.38 GHz.

Direct communication with mobile phones and pagerstakes place in the L-band between 1616 and 1626.5 MHz viathree main antennas and 16 spotbeams.

Communications with the gateways is via four movableantennas in the Ka-band (downlink 19.4–19.6 Ghz anduplink 29.1–29.3 GHz).

All antennas consist of flat panels and use state-of-the-art technology.

As mentioned earlier, the 80 satellites were designedand manufactured by Lockheed Martin. This company hasto date built more LEO satellites than any other organiza-tion in the western world. In spite of all this experience,there were a number of failures, and these set back theoverall program by about six months.

Most of the satellites are launched in groups of five by aBoeing Delta II Rocket from the Vandenberg launch pad inCalifornia.

Some satellites were launched in groups of seven by aRussian Proton Rocket from the Baikonur Kosmodrom sitein Kazakhstan.

Some satellites will be launched by China, but the GreatWalls Industries Corporation’s ‘Long March 2C/SD’ rock-ets can accommodate only two satellites at a time.

The spare satellites will be launched by Delta II andLong March 2C/SD rockets. At the end of last year, seven ofthe 14 spare and control satellites were already in orbit.

The Iridium Satellites

Figure 4. Iridiumpager and mobilephone fromMotorola. Thephone has an inter-changeable fre-quency module fora number of mobilephone standardsand needs an addi-tional L-bandantenna unit foroperation in theIridium Network.

TEST & MEASUREMENT

multiburstgenerator

video test signal from two ICs

In many instances,suspect television

receivers and moni-tors can be tested

with a simple test pat-tern generator as

described in this arti-cle. The circuit pro-

posed in this, whichconsists of only a

handful of compo-nents, generates a

series of test bars, theso-called multiburst.

Design by W. den Hollander

INTRODUCTIONAlthough modern television receiversand monitors are highly immune tonoise and interference, there willalways be a need for a test pattern gen-erator. The compact circuit presentedenables many of the basic functions ofa television receiver or monitor to bechecked. Owing to its simplicity, it pro-vides only a black -and -white pattern,of course.

DESIGNCONSIDERATIONSThe simple test pattern of grey bars,called a multiburst, generated by thecircuit consists of a wave train contain-

ing different frequencies at a level of(average value 650 mV).420 mVP

-pThe multiburst enables the frequencyresponse, and certain other propertiesof the system on test to be measured. Ina properly adjusted receiver or moni-tor, all components, with the possibleexception of the 5.8MHz burst, may bepresent. The reference bar and the 2Tpulse are repeated in line 330, so thepulse repetition rate (ppr)of theseimportant test signals is 50 Hz.

The multiburst also enables thevideo amplifiers, sync(hronization) se-parator and output amplifiers to betested.

Disregarding the power supply, the


I N T R O D U C T I O NAlthough modern television receiversand monitors are highly immune tonoise and interference, there willalways be a need for a test pattern gen-erator. The compact circuit presentedenables many of the basic functions ofa television receiver or monitor to bechecked. Owing to its simplicity, it pro-vides only a black-and-white pattern,of course.

D E S I G NC O N S I D E R A T I O N SThe simple test pattern of grey bars,called a multiburst, generated by thecircuit consists of a wave train contain-

ing different frequencies at a level of420 mVp-p (average value 650 mV).The multiburst enables the frequencyresponse, and certain other propertiesof the system on test to be measured. Ina properly adjusted receiver or moni-tor, all components, with the possibleexception of the 5.8 MHz burst, may bepresent. The reference bar and the 2Tpulse are repeated in line 330, so thepulse repetition rate (ppr)of theseimportant test signals is 50 Hz.

The multiburst also enables thevideo amplifiers, sync(hronization) se-parator and output amplifiers to betested.

Disregarding the power supply, the

In many instances, suspect television

receivers and moni-tors can be tested

with a simple test pat-tern generator as

described in this arti-cle. The circuit pro-

posed in this, whichconsists of only a

handful of compo-nents, generates a

series of test bars, theso-called multiburst.

52

Design by W. den Hollander

multiburstgenerator

video test signal from two ICs

TEST & MEASUREMENT

generator is based on two ICs: an in-system programmable logic assembly)and a VCO (voltage-controlled oscilla-tor). The circuit diagram is shown inFigure 1.

C I R C U I T D E S C R I P T I O NThe in-system programmable IC (IC1)contains 36 macrocells and a total of800 gates. In the present application, itis used to generate the sync(hroniza-tion) signal, CS, the blanking signal,CB, and the control voltage for theVCO.

The sync signal, which meets inter-national standards, is available at pin 7of IC1, and from there applied to out-put amplifier T1 via resistor R14.

The blanking signal, manifested bya logic low level at pin 14 of IC1, is usedto disable the oscillator in IC2 duringthe sync pulse.

As shown by the timing diagram inFigure 2, a pulse appears at each ofoutputs D0–D7 during the line period(64 µs). These pulses are converted intoa staircase consisting of direct voltagesteps. The staircase signal is used tocontrol the VCO in IC2 via pin 9, result-ing in a frequency-modulated outputsignal – the multiburst – at pin 4 of IC2.

The central frequency of the VCO isdetermined by the values of R12 andC3 and the setting of P1. With values asspecified and P1 adjusted for a centralfrequency of 1 MHz, the output fre-quency rises in eight steps to a value of10.5 MHz.

The d.c. operating point of outputamplifier T1 is determined by the val-ues of R15 and R16. The video and syncsignals are superimposed on this viaR13, R14 and C4. The amplifier isdesigned to have an output impedanceof 75 Ω at an output voltage of 1 Vp-p.

The important signals in the circuitare summarized in Figure 3. In this dia-gram, the sync signal, on to which theoutput signal of the VCO is superim-posed, is on the top line. The blankingsignal is on the second line and thestaircase control voltage for the VCOon the third.

The output signal of the generatoris on the bottom line. This showsclearly that at the onset of each hori-zontal line, that is, every 64 µs, theoscillator is disabled.

Note that the irregularities on thescreen images are caused by the sam-pling process of the oscilloscope used.

The circuit is pow-ered by a simple sup-ply whose output isstabilized at 5 V by regulator IC3.Diode D2 prevents damage throughinadvertent incorrect polarity of theinput voltage. Diode D1 is a supply-onindicator. The input voltage may bederived from a simple mains adaptorwith an output of 9–12 V.

C O N S T R U C T I O NThe generator is conveniently built onthe printed-circuit board shown in Fig-

ure 4 (available ready-made – see ReadersServices towards the

end of this issue).IC1 should be placed in a suitable

socket: mind the polarity. Do not over-look the wire bridge alongside R10.

T E S T I N GWhen the generator has been built andinspected thoroughly, switch on themains supply to the adaptor. Connectan oscilloscope to the output andadjust P1 so that the lowest frequency


R2

2k

2

R11

1k

2

R15

2k

2

R16

1k

8

R17

10

0Ω

R12

1k

5

R1

10M

R18

75Ω

R13

2k2

R9560Ω

R81k

R71k5

R10270Ω

R62k2

R53k0

R43k9

R34k7

X1

10MHzC1

22p

C2

22p

C6

100n

C8

100n

C9

100n

C5

10p

C11

330n

C7

100n

C4

8p2

R14

820Ω

74HCT4046

VCOIN VCOUT

IC2

DEMO

SIN

CIN

INH

ZEN

PLLP1

P2

PP

12 10

15

13

11

14

CX

CX

R1

R2

16

6

9

5

4

1

2

7

3

Φ

8

C3

1n

XC9536-15-PC44

IC1

CLKI CLK

43

29D0

28D1

26D2

25D3

22D4

20D5

19D6

18D7

14CB

CS

32

33

34

35

36

37

38

39

40

41

21

10

11

12

13

15

16

17

23

24

27

30

31

42

44

5

7

1

2

3

4

6

8

9

500ΩP1

T1

BC547B

C12

220µ25V

C10

10µ63V

R19

2k

7

7805

IC3D2

1N4001

D1

> 8V 5V

5V

5V

980095 - 11

630

CS

CB

D0

D1

D2

D3

D4

D5

D6

D7

640 650 660 670 680 690 700 710 720 730

630 640 650 660 670 680 690 700 710 720 730

980095 - 12

Figure 1. Circuit dia-gram of the multiburstgenerator.

Figure 2. The in-systemprogrammable IC gen-erates a number of dig-ital signals that are thebasis of the output testsignal.

1

2

15 -Sep -9815:23:57

Main Menu 111E!!!!!!7" iniurrMEMOOMMMEN511.r7=11ri!INIMMOMMIONNMONMPEaMONIOEMPAMMEMMEW

nT171-71'

MMEMEMEMME

Parts listResistors:

= 10 Mf2R2, R6, R13, R15 = 2.2 k52R3 = 4.7 k52R4 = 3.9 k52R5 = 3.00/3.01 k52R7, Ri2 = 1.5 k52R8 = 1 kf2

560Rio= 27052R11 = 1.2 k52R14 = 820 52R16 = 1.8 k52R17= 1 00 CIR18= 7552R19 = 2.7 k52Pi = 500 52 preset

Capacitors:Di, C2 = 22 pF03 = 0.001 [IF04 = 8.2 pF05= 1 0 pFC6-09 = 0.1 [IF, ceramic010 = 10 [IF, 63 V, radialCii = 0.33 [IF

220 [IF, 25 V, radial012=

Semiconductors:Di = LED, high efficiency, redD2= 1N4001

= BC547B

Integrated circuits:ICi = XC9536-15-PC44 (Xilinx); avail-able ready programmed underOrder no. 986520-1*

102 = 74HCT4046103 = 7805

Miscellaneous:X1 = crystal, 10 MHzPCB Order no. 980095-C*Diskette with Jedec source file: Order

no. 986029-1*

*(see Readers Services towards theend of this issue)

TV EXT

Line 36( se)

4

Mem C10 ms 2 V

Mem D10 ms 2 V

Chan 110 fAS 2 VChan 210 ms 5 V

CH1 .2 V10=CH2 .5 V10=

T/div 1.0 is980095 -14

Figure 3. Oscillogram ofthe signals at four loca-tions in the generator.

000001 R12 10

980095-1 I-°

is 1 MHz.The generator is intended for test-

ing television receivers or monitorsthat have a field frequency of 50 Hzand 625 lines per picture. If differentfrequencies are desired, the value ofresistors R3-R10 should be altered asrequired.

Note that, in contrast to proprietarymultiburst generators, the present onedoes not provide a sinusoidal output.

[980095]

Figure 4. Printed -circuitboard for the multiburstgenerator.

(C) ELEKTOR 0980095-1


is 1 MHz.The generator is intended for test-

ing television receivers or monitorsthat have a field frequency of 50 Hzand 625 lines per picture. If differentfrequencies are desired, the value ofresistors R3–R10 should be altered asrequired.

Note that, in contrast to proprietarymultiburst generators, the present onedoes not provide a sinusoidal output.

[980095]

980095 - 14

980095-1(C) ELEKTOR

C1C2

C3

C4

C5

C6

C7

C8

C9

C10

C11

C12

D1

D2 H1

H2

H3

H4

IC1

IC2

IC3

OUT1

P1

R1

R2

R3R4R5

R6R7R8R9

R10

R11

R12

R13R14

R15

R16

R17

R18

R19

T1

X1

980095-1

0+

T 980095-1(C) ELEKTOR

Parts listResistors:R1 = 10 MΩR2, R6, R13, R15 = 2.2 kΩR3 = 4.7 kΩR4 = 3.9 kΩR5 = 3.00/3.01 kΩR7, R12 = 1.5 kΩR8 = 1 kΩR9 = 560 ΩR10 = 270 ΩR11 = 1.2 kΩR14 = 820 ΩR16 = 1.8 kΩR17 = 100 ΩR18 = 75 ΩR19 = 2.7 kΩP1 = 500 Ω preset

Capacitors:C1, C2 = 22 pFC3 = 0.001 µFC4 = 8.2 pFC5 = 10 pFC6–C9 = 0.1 µF, ceramicC10 = 10 µF, 63 V, radialC11 = 0.33 µFC12 = 220 µF, 25 V, radial

Semiconductors:D1 = LED, high efficiency, redD2 = 1N4001T1 = BC547B

Integrated circuits:IC1 = XC9536-15-PC44 (Xilinx); avail-

able ready programmed underOrder no. 986520-1*

IC2 = 74HCT4046IC3 = 7805

Miscellaneous:X1 = crystal, 10 MHzPCB Order no. 980095-C*Diskette with Jedec source file: Order

no. 986029-1*

*(see Readers Services towards theend of this issue)

3

Figure 3. Oscillogram ofthe signals at four loca-tions in the generator.

Figure 4. Printed-circuitboard for the multiburstgenerator.

4

APPLICATION NOTEThe content of this note is based on information received from manufacturers in theelectrical and electronics industries or their representatives and does not imply prac-tical experience by Elektor Electronics or its consultants.

fluorescent lampstarter

2 -chip set from SGS-Thomson

ST Electronics (formerly SGS-Thomson) haveearlier this year brought out a 2 -chip starter,the EFS (Starlight) Kit, which, in conjunction

with four additional passive components, canbe used as a direct replacement of the glow

switch starter. The kit comprises a driver andan ASDTM, which contains a power switch, and

a power supply for the driver. The driver isprogrammed to ensure efficient ignition of the

fluorescent tube.

An SOS -Thomson ApplicationExpanded by G. Kleine

INTRODUCTION

A low pressure fluorescent lamp nor-mally consists of a glass tube, 38 mm indiameter and between 600 mm and2400 mm long. The tube is filled with anoble gas at a pressure of not morethan 4 per cent of atmospheric pressureand also contains a drop of liquid mer-cury. The interior surface of the tube iscoated with phosphor which convertsthe ultraviolet light produced by thedischarge into visible light.

At each end of the tube are elec-trodes that serve alternately as cathodeand anode since these lamps are usedin a.c. circuits.

The cathodes of a hot -cathode fluo-rescent lamp consist of a coiled filamentcoated with barium -oxide that is held


I N T R O D U C T I O N

A low pressure fluorescent lamp nor-mally consists of a glass tube, 38 mm indiameter and between 600 mm and2400 mm long. The tube is filled with anoble gas at a pressure of not morethan 4 per cent of atmospheric pressureand also contains a drop of liquid mer-cury. The interior surface of the tube iscoated with phosphor which convertsthe ultraviolet light produced by thedischarge into visible light.

At each end of the tube are elec-trodes that serve alternately as cathodeand anode since these lamps are usedin a.c. circuits.

The cathodes of a hot-cathode fluo-rescent lamp consist of a coiled filamentcoated with barium-oxide that is held

56

An SGS-Thomson ApplicationExpanded by G. Kleine

fluorescent lampstarter

2-chip set from SGS-Thomson

The content of this note is based on information received from manufacturers in theelectrical and electronics industries or their representatives and does not imply prac-tical experience by Elektor Electronics or its consultants.APPLICATION NOTE

ST Electronics (formerly SGS-Thomson) haveearlier this year brought out a 2-chip starter,the EFS (Starlight) Kit, which, in conjunction

with four additional passive components, canbe used as a direct replacement of the glow

switch starter. The kit comprises a driver andan ASD™, which contains a power switch, and

a power supply for the driver. The driver isprogrammed to ensure efficient ignition of the

fluorescent tube.

in place by nickel support wires.Anodes in the form of metal strips

or wires attached to the support wiresare used in high-loading tubes, but inlow-loading tubes the support wiresact as anodes.

The noble gas serves to reduce therate of evaporation of the barium from

the cathode and also to reduce thestarting voltage.

The tube cathodes are heated bypassing a current through each fila-ment, which causes a large number ofelectrons to be emitted by the oxidecoating. Assuming sufficient voltage isapplied across the tube, a glow dis-charge is set up through the noble gaswhich excites or ionizes the mercuryatoms throughout the tube and estab-lishes a mercury-arc discharge.

A fluorescent tube requires a ballast(usually a choke) to limit the currentthrough the tube, a starter for preheat-ing the cathode filament. In manycases, this is complemented by apower-factor-correcting capacitor.

As the arc current increases, thevoltage drop across the choke rises, sothat the potential across the tube dropsuntil a balance is reached.

The simplest and most frequentlyused starting device is the starterswitch. When this is closed, a currentflows through the ballast and the twocathodes in series, whereupon, as soonas the cathodes reach the emissiontemperature, local ionization is set upand the ends of the tube begin to glow.The starter switch is then openedrapidly, and the resulting change incurrent causes a large counter-e.m.f.across the coil, which is applied acrossthe tube and is sufficient to cause thearc to strike.

The operations just described are

illustrated in Figure 1.Automatic starter switches permit a

heating current to flow for a predeter-mined time before they open and pro-duce the striking pulse. The traditionalstarting switch is a glow type, in whichthe switch contacts are mounted onbimetallic strips that bend towardseach other when they are heated.

In many cases, bimetal startershave been, or are being, replaced byelectronic starters. Current electronicstarters for fluorescent lamps use aunidirectional switch, like a MOStransistor or GTO (Gate Turn Off)thyristor. Since a starter is a bidirec-tional switch it is necessary to use arectifier bridge and, in addition, 2 or 3diodes in series with the GTO to getthe necessary switch-off effect. Thewhole switch is controlled with ananalogue timer built around a smallSCR (Silicon Controlled Rectifier). Theswitch in the new kit from ST Elec-tronics is a bidirectional type, whichsubstantially reduces the number of


240V

ballast

La

approx. 60V

50Hz

240V

ballast

La

50Hz

240V

ballast

La

approx. 1kV

50Hz

240V

ballast

approx. 80V

50Hz

I glow

La

1

Figure 1. Operation ofa fluorescent tubewith traditional starterswitch.

14

8 980101 - 12a

1

7

SO14

lamp statusoff/ignited

zerocrossing

over-current

under-voltage

preheatcounter

currentsource

counter8 ignitionattempts

2 PreheatSelect

G2

G1

13

14

11 3 5 6 7

9

8 10

12

4

1TubeState

Shunt

VCC

NCC

GND3, 5, 6, 7, 8, 10 = GND

!

980101 - 12b

powersupply

driverON/OFF

4

2VCC

1G1

5G2

GND

TUBE

2VCC

1G1

3Tube

4GND5G2

Pentawatt HV

3

1.3

kV C

lam

p

( TO220)

2

Figure 2. Internalblock diagram of (a)the driver and (b)the ASD™ in theStarlight Kit.

a

b

3

4

240V',50Hz

ballast

*see tekst

Figure 3. Basic appli-cation diagram of theEFS (Starlight) starter.

Figure 4. Track layoutof the PCB for theStarlight Kit. (200%)

L114w

Figure 5. Componentlayout of the PCB forthe Starlight Kit. (200%)

980101 - 13

additional components and the powerlosses (only one forward voltageinstead of 5 or 8 depending on theunidirectional version).

BASIC APPLICATIONA block diagram of the driver is shownin Figure 2a.

The bidirectional power switch per-forms two functions:

preheating the tube;igniting the tube.

During the preheating period, theASDTM conducts hard. The tube isshort-circuited by the starter, and thecurrent flows through its filaments. Inthis condition, the tube cannot light,but the temperature of the filamentsrises steadily.

At the end of the preheating period,the filaments of the tube are hotenough to emit electrons into the gasand to permit the gas igniting. TheASD then switches off the preheatingcurrent. At that moment, the ballast isequivalent to a current generatorwhere 1= 1so \-(1so = switch -off current).-

While the ASD switches off, the startervoltage increases. The amplitude ofthis high -voltage spike is clamped bythe ASD (U --- 1350 V). Since thepower switch and the tube are in par -

Recommended components:

= 150 kf2, 1/2 W, SMDR2 = 1 kf2 or 2 kf2 (see text) 0.5 WSMD

R3 = 0.39 f2, 1/4 WAll resistors 5%

Capacitors:C= 22 µF, 16 V, 10%

Integrated circuits:See Table

allel, the striking pulse is applieddirectly to the tube. The electromag-netic energy of the ballast is then dis-charged through the tube and theASDTM

The ASD contains an auxiliarypower supply for the driver, a powerclamping device (1350 V) and a bidi-rectional switch (Figure 2b).

A fluorescent tube requires a mini-mum level of energy to be ignited, butthis energy depends largely on thetemperature of the tube. The lower thistemperature, the less energy, and themore difficult it is for the tube to ignite.The energy, E, stored in the ballast, L,is directly proportional to the /so:

E= Lls0212.

This means that the required energy,and thus the /so, is a maximum at theminimum temperature. Results basedon experiments show that it is neces-sary to switch off a current of 350 mAto strike a 58 W tube at -20 °C (with thevoltage clamped at 1300 V).

So, the best way for a tube to strikeindependently of temperature is tokeep the /50 at a maximum over thewhole temperature range. In practice,the /50 is maintained at 350 mA at tem-peratures below -10 °C and thenallowed to drop gradually to about180 mA at + 75 °C.

AUXILIARY POWERSUPPLYTo reduce the number of components,an auxiliary power supply is incorpo-rated in the ASD, which worksdirectly from the mains. A capacitor atthe output of the supply is charged atthe beginning and end of each posi-tive half -cycle of the mains voltage.During the preheat period, a part ofthe current flowing through the ASD isused to charge the output capacitor, sothere is no line current distortion.When the tube is ignited, the capaci-tor is periodically recharged by theASD to allow monitoring of the tubeby the starter kit.

FUNCTIONALDESCRIPTIONA functional diagram is shown in Fig-ure 3. At switch -on, an integratedUnder Voltage Lock Out (UVLO) func-tion resets the driver as long as thesupply voltage stays below a safe level.

The ignition sequence begins withthe preheat period. Two differentperiod lengths are available:

if pin 2 is connected to ground, thepreheat period is 1.5 s;if pin 2 is connected to Vcc, the pre-heat period is 2.5 s.


additional components and the powerlosses (only one forward voltageinstead of 5 or 8 depending on theunidirectional version).

B A S I C A P P L I C A T I O N

A block diagram of the driver is shownin Figure 2a.

The bidirectional power switch per-forms two functions:

• preheating the tube;• igniting the tube.

During the preheating period, theASD™ conducts hard. The tube isshort-circuited by the starter, and thecurrent flows through its filaments. Inthis condition, the tube cannot light,but the temperature of the filamentsrises steadily.

At the end of the preheating period,the filaments of the tube are hotenough to emit electrons into the gasand to permit the gas igniting. TheASD then switches off the preheatingcurrent. At that moment, the ballast isequivalent to a current generatorwhere I = ISO (ISO = switch-off current).While the ASD switches off, the startervoltage increases. The amplitude ofthis high-voltage spike is clamped bythe ASD (UBR ≈ 1350 V). Since thepower switch and the tube are in par-

allel, the striking pulse is applieddirectly to the tube. The electromag-netic energy of the ballast is then dis-charged through the tube and theASD™.

The ASD contains an auxiliarypower supply for the driver, a powerclamping device (1350 V) and a bidi-rectional switch (Figure 2b).

A fluorescent tube requires a mini-mum level of energy to be ignited, butthis energy depends largely on thetemperature of the tube. The lower thistemperature, the less energy, and themore difficult it is for the tube to ignite.The energy, E, stored in the ballast, L,is directly proportional to the ISO:

E = LISO2 / 2.

This means that the required energy,and thus the ISO, is a maximum at theminimum temperature. Results basedon experiments show that it is neces-sary to switch off a current of 350 mAto strike a 58 W tube at –20 °C (with thevoltage clamped at 1300 V).

So, the best way for a tube to strikeindependently of temperature is tokeep the ISO at a maximum over thewhole temperature range. In practice,the ISO is maintained at 350 mA at tem-peratures below –10 °C and thenallowed to drop gradually to about180 mA at +75 °C.

A U X I L I A R Y P O W E RS U P P L YTo reduce the number of components,an auxiliary power supply is incorpo-rated in the ASD, which worksdirectly from the mains. A capacitor atthe output of the supply is charged atthe beginning and end of each posi-tive half-cycle of the mains voltage.During the preheat period, a part ofthe current flowing through the ASD isused to charge the output capacitor, sothere is no line current distortion.When the tube is ignited, the capaci-tor is periodically recharged by theASD to allow monitoring of the tubeby the starter kit.

F U N C T I O N A LD E S C R I P T I O NA functional diagram is shown in Fig-ure 3. At switch-on, an integratedUnder Voltage Lock Out (UVLO) func-tion resets the driver as long as thesupply voltage stays below a safe level.

The ignition sequence begins withthe preheat period. Two differentperiod lengths are available:

• if pin 2 is connected to ground, thepreheat period is 1.5 s;

• if pin 2 is connected to VCC, the pre-heat period is 2.5 s.


240V

ballast

La

50HzDRIVER

U2ASD

U1

*

*see tekst

18...70W 9

114

13

12

1

5

2

4

3

24 3,5,6,7,8,10,11

R3

1/4W

Br1

Br2

C122µ16V

14V

R1a47k

R1c

R1b

56k

47k

R21k

980101 - 13

3

Figure 3. Basic appli-cation diagram of theEFS (Starlight) starter.

Figure 4. Track layoutof the PCB for theStarlight Kit. (200%)

Figure 5. Componentlayout of the PCB forthe Starlight Kit. (200%)

4

5

Recommended components:

R1 = 150 kΩ, 1/2 W, SMDR2 = 1 kΩ or 2 kΩ (see text) 0.5 W

SMDR3 = 0.39 Ω, 1/4 WAll resistors 5%

Capacitors:C = 22 µF, 16 V, 10%

Integrated circuits:See Table

During the preheat period, the drivermaintains the ASD in the full on state,making the starter equivalent to aclosed switch.

At the end of the preheat period,the starter ignites the fluorescent tube.To this end, the driver continuouslymonitors the current through thestarter. When the current reaches theswitch-off level (350 mA), the driverturns off the ASD. This causes a high-voltage pulse to be induced across thetube, which is limited by the ASDbreakdown voltage of 1350 V.

The driver senses the state of thetube (on or off). When the tube fails tostrike and remains off for eight periodsof the mains voltage, a new preheatperiod, shorter than the first, is begun,followed by a new ignition attempt.The driver will try to fire the tube eighttimes. If none of these attempts is suc-cessful, the driver is set to the standbymode, and the whole starter stops untilthe mains is switched off and then onagain.

Once the tube is ignited, the driverstays in the standby mode, but contin-ues to monitor the state of the tube.The ASD applies a short high-voltagepulse of 1 mJ at the beginning of eachpositive mains half-cycle. If the mainsvoltage drops briefly, the tube turns offmomentarily, but the short pulse is suf-ficient to sustain the arc in the tube, soavoiding the need for a new ignitionsequence.

During normal operation of thetube, the short pulse is masked by thetube conduction. If the mains is inter-rupted for some time, the lamp isswitched off completely and a newignition sequence is started as soon asthe mains voltage is restored.

C O N S T R U C T I O NThe complete starter may be built onthe printed-circuit board shown in Fig-ures 4 and 5. This fits nicely in the caseof a defect glow type starter.

The components are SMD (surface-mount device) types. Start the solder-ing with the resistors, then the capacitor,and finally the two ICs.

The preheat time must then be set:pin 2 is the preheat time select pin. Toselect a 1.5 s preheat time, drill and cutthe VCC to pin 2 at the metallized hole.To select a 2.5 s preheat time, drill andcut the GND to pin 2. Either of these

actions must be carried out to avoid asupply short circuit.

F I N A L L Y …When the board has been fitted in aglow type starter case, insert it into therelevant connector on the tube holder.When the lamp is switched on, itshould ignite at once. If it does not,lenghten the preheat time.

If increasing the preheat time doesnot help or if instead of the eightattempts at igniting the tube there isonly one, the tube may have an inte-gral series compensation capacitor asshown in Figure 6. In such a case, thereis only one attempt at igniting the tube,but this should be sufficient.

There is an arrangement for twospecial tubes in series, the so-calledtwin tube configuration, as shown inFigure 7. This uses two starters, butonly one ballast. The ST Electronicsstarter can be used in this arrangementif the value of R2 is increased to 2 kΩ.If in this configuration the tubes do notignite, turn one of the starters 180° inits socket.

[980101]


240V

ballastC

La

980101 - 14

50Hzelektronic

starter

240V

ballast

La1

980101 - 15

50Hz

elektronicstarter 1

La2

elektronicstarter 2

6

7

Figure 6. Some fluo-rescent tubes have anintegral series com-pensation capacitor.

Figure 7. Twin tubeconfiguration.

EFS Starlight Kit EFS1 EFS2 EFS3

Tube rating (W) 36–58 18–70 18–125

Umgebungstemperatur [°C] 5 to 75 –20 to 75 –40 to 85

Twin tube serial connection Yes Yes Yes

Driver (U2) EFS1-A EFS2-A EFS3-A

ASD™ (U1) EFS1-1 EFS2-1 EFS3-1

WHEN ELECTRONICS WAS YOUNG (1)At the threshold of a new millennium:a look back over centuries past

Standing at the threshold of a new millennium, it is asobering thought to reflect that static electricity hasbeen known for almost 2500 years: the Greeks discov-ered that friction on amber (Greek: e)erpov - elec-tron) by fur gave rise to attractive forces. However, theword electron was not used until 1897 when J JThomson discovered the electron as we know it.

Electrical technology and its offspring electronicshave had, and still have, a tremendous impacthuman society. Today, it would be difficult to imaginea world without electric lights, radio, television, tele-phone, cars (no ignition), lifts, aircraft. afxourse,without electricity there would still have -15-ea trains,(mechanical) computers, -and other appliances. Thedevelopment of electrical technology and, later, elec-tronics may be compared with that of hieroglyphs,cuneiform, and printing.

Static electricity was not studied for many centuriesafter its discovery until the early 1700s,- when-Gr-ay-distinguished between conductors and insulators and_Fry found -that electricity can be positive and negative.

It is difficult to say -when-the history of electronics

started. Although the work of Gray and Fry was im-portant, as was the invention of the first capacitor, orLeyden jar, in 1745, by von Kleist (and, almost a yearlater, by Muschenbroek at the University of Leyden inthe Netherlands), but these were still concerned withstatic electricity. Electronics as we know it is the sci-ence and technology of the conduction of electricity,that is, the movement of electrons, in a gas, vacuum,or semiconductor. The conduction of electricity inother materials or substances is really the domain ofelectrical technology.

The study of electricity started so ably by 18th cerj-_tury pioneers like Gray, Fry, Muschenbroek, pncl vonKleist,fe-din 1800 to the invention of the dry batteryby Alessandro Volta, followed in 1803 by Ritter's in-vention of the rechargeable battery. Since the supplyof electric power provided by such batteries is impor-tant to electronics, the year 1800 seems a good start-ing paint fill u Short overview of the history of elec-&mks-although, strictly speaking, almost all of thathistory really belongs to 'our' century: the 20th

Over the coming months, we shall have a look at anumber of milestones in the history of electronics andthe grand women -ho made it all possible.

[995008]


At the threshold of a new millennium: a look back over centuries past

Standing at the threshold of a new millennium, it is asobering thought to reflect that static electricity hasbeen known for almost 2500 years: the Greeks discov-ered that friction on amber (Greek: ελεχτρον – elec-tron) by fur gave rise to attractive forces. However, theword electron was not used until 1897 when J JThomson discovered the electron as we know it.

Electrical technology and its offspring electronicshave had, and still have, a tremendous impact onhuman society. Today, it would be difficult to imaginea world without electric lights, radio, television, tele-phone, cars (no ignition), lifts, aircraft. Of course,without electricity there would still have been trains,(mechanical) computers, and other appliances. Thedevelopment of electrical technology and, later, elec-tronics may be compared with that of hieroglyphs,cuneiform, and printing.

Static electricity was not studied for many centuriesafter its discovery until the early 1700s, when Graydistinguished between conductors and insulators andFry found that electricity can be positive and negative.

It is difficult to say when the history of electronics

started. Although the work of Gray and Fry was im-portant, as was the invention of the first capacitor, orLeyden jar, in 1745, by von Kleist (and, almost a yearlater, by Muschenbroek at the University of Leyden inthe Netherlands), but these were still concerned withstatic electricity. Electronics as we know it is the sci-ence and technology of the conduction of electricity,that is, the movement of electrons, in a gas, vacuum,or semiconductor. The conduction of electricity inother materials or substances is really the domain ofelectrical technology.

The study of electricity started so ably by 18th cen-tury pioneers like Gray, Fry, Muschenbroek, and vonKleist, led in 1800 to the invention of the dry batteryby Alessandro Volta, followed in 1803 by Ritter’s in-vention of the rechargeable battery. Since the supplyof electric power provided by such batteries is impor-tant to electronics, the year 1800 seems a good start-ing point for a short overview of the history of elec-tronics although, strictly speaking, almost all of thathistory really belongs to ‘our’ century: the 20th.

Over the coming months, we shall have a look at anumber of milestones in the history of electronics andthe men and women who made it all possible.

[995008]


WHEN ELECTRONICS WAS YOUNG (1)

IN C

8 1 IN

2 7 1 NC

TOP VIEW sw E

6 1 MG

OUT E 4 5 GND

SPJ001 11

INM E COM E OFS E

OUTS

AD8307 TOP VIEW

(Not to Scale)

INP

VPS

E ENB

E INT

- 16

SUPPLY 02:

7.5m

AD8307

BANDGAP REFERENCE AND BIASING

SIX 14.3dB SOOMHz AMPLIFIER STAGES

-INPUT 0. NINE DETECTOR CELLS

SPACED 14.3dB COMMON® COM

ENBO ENABLE

INTO INT. ADJ

MIRROR

2µA OUT MB

2 -C.) OUTPUT

12.51A

INPUT -OFFSET COMPENSATION LOOP

COM OFSW

OFS. ADJ.

599001 -14

'10

INPUT FREQUENCY 10121112

INPUT FREQUENCY MOM

INPUT FREQUENCY ROOMY -

INPUT FREQUENCY 50016,12

-60 -00 -20 AMPLITUDE - clBm

20

SW MC Divide Ratio

H H 1/64

H L 1/65

L H 1/128

L L 1/129

OUT

003001- 12

.1000 E

LAI 800 D

1- 600

400 03

H k- 200

2 0

z

MB501L Vcc

= 5.0V

TA = 25°C

10 20 50 100 200 500 1000 2000

INPUT FREQUENCY (MHz)

993001 - 13

ANALOG DEVICES

FUJITSU

DATASHEET 1/99

MB501L

Integrated CircuitsSpecial Function, Prescalers

AD8307

Integrated CircuitsAnalogue

DATASHEET 1/99

65E

lektor Electronics

1/99

AD8307Low-Cost DC-500 MHz, 92 dB Logarithmic Amplifier

ManufacturerAnalog Devices, One Technology Way, P.O. Box 9106,Norwood, MA 02062-9106, U.S.A.Internet: www.analog.com.

Features- Complete Multistage Logarithmic Amplifier- 92 dB Dynamic Range: –75 dBm to +17 dBm

To –90 dBm Using Matching Network- Single Supply of 2.7 V Min at 7.5 mA Typical- DC-500 MHz Operation, ±1 dB Linearity- Slope of 25 mV/dB, Intercept of –84 dBm- Highly Stable Scaling Over Temperature- Fully Differential DC-Coupled Signal Path- 100 ns Power-Up Time, 150 µA Sleep Current

ApplicationsConversion of Signal Level to Decibel FormTransmitter Antenna Power MeasurementReceiver Signal Strength Indication (RSSI)Low Cost Radar and Sonar Signal ProcessingNetwork and Spectrum Analyzers (to 120 dB)Signal Level Determination Down to 20 HzTrue Decibel AC Mode for Multimeters

Application ExampleRF Decibel Meter, Elektor Electronics January 1999.

Product Description (excerpt)The AD8307 is the first logarithmic amplifier in an 8-lead (SO-8) package. It is a complete 500-MHz mono-lithic demodulating logarithmic amplifier based on theprogressive compression (successive detection) tech-nique, providing a dynamic range of 92 dB to ±3 dBlaw-conformance and 88 dB to a tight ±1-dB errorbound at all frequencies up to 100 MHz. It is extremelystable and easy to use, requiring no significant exter-nal components.

A single supply voltage of 2.7 V to 5.5 V at 7.5 mA isneeded, corresponding to an unprecedented powerconsumption of only 22.5 mW at 3 V. A fast-actingCMOS-compatible control pin can disable the AD8307to a standby current of under 150 µA.

Functional block diagram

Pin configuration

VOUT vs. input level (dBm) at various frequencies.

MB501LTwo Modulus Prescaler

ManufacturerFujitsu. Internet: www.fujitsu.com

Product DescriptionThe Fujitsu MB501L is a two modulus prescaler usedin a frequency synthesizer to make a Phase LockedLoop (PLL). It will divide the input frequency by themodulus of 64/65 or 128/129 The output level is 1.6Vpeak to peak on ECL level.

Features- High Operating Frequency, Low Power Operation

1.0 GHz at 150mW typ. (MB501)1.1 GHz at 50 mW typ. (MB501L)

- Pulse Swallow Function- Wide Operation Temperature TA = –40°C to +85°C- Stable Output Amplitude VOUT = 1.6 Vp-p.- Complete PLL synthesizer circuit with the Fujitsu

MB87001A, PLL synthesizer IC- Plastic 8-pin Standard Dual-In-Line Package or space

saving Flat Package

Application exampleGeneral Coverage Receiver,Elektor Electronics January & February 1999

Pin assignment

Input signal amplitude vs. input frequency

SW MC Divide Ratio

MB501/501L

H H 1/64

H L 1/65

L H 1/128

L L 1/129Note: SW: H = VCC, L = open MC: H = 2.0 V to VCC L = GND to 0.8 V

Block diagram

Clock Data LE

Lock Set.

C1

Vcc 7R

V CC \ICC ./c o 17T

(Max.Vsx 8V)

t6 15 14 13 12 11 10 9

MB87001A

2 3 4 5 6 7 8

77T 07.2 071,2 07Kit

Vcc T

103KSE

8 7 6 5

PR ES CAL ER

2 3 4

1000s, 777

10 OpF

10K

77T

VCO

XI :12.8MHz Zed Vcc :56 z 10% Vs,, 18V Max. C1, C

: depends on crystal oscillator

ox

OUTPUT O

M example of application of MB501/5011/503/504/504L with PLL synthesizer IC MB87001 A 993001 - 17

Sampling scope input point for Input waveform

co

.0 V s 10%

Sampling scope prober point for output waveform

Ce:1000pS Cg:1000pF

Cy: 0.1pF CL: SpF (Including scope and IN capacitance)

21,11 009C01 - I a

MB501L

Integrated CircuitsSpecial Function, Prescalers

AD8307

Integrated CircuitsAnalogue

DATASHEET 1/99DATASHEET 1/99

66E

lekt

or E

lect

roni

cs1/

99

AD8307 Specifications (VS = +5 V, TA = 25ºC, RL ≥1MΩ, unless otherwise noted)

Parameter Conditions Min Typ Max Units

GENERAL CHARACTERISTICS Input Range (±1 dB Error) Expressed in dBm re 50 Ω –72 16 dBm Logarithmic Conformance f ≤100 MHz, Central 80 dB ±0.3 ±1 dB

f = 500 MHz, Central 75 dB ±0.5 dB Logarithmic Slope Unadjusted 1 23 25 27 mV/dB vs. Temperature 23 27 mV/dB Logarithmic Intercept Sine Amplitude; Unadjusted 2 20 µV

Equivalent Sine Power in 50 Ω –87 –84 –77 dBm vs. Temperature –88 –76 dBm Input Noise Spectral Density Inputs Shorted 1.5 nV/√Hz Operating Noise Floor RSOURCE = 50 Ω/2 –78 dBm Output Resistance Pin 4 to Ground 10 12.5 15 kΩ

Internal Load Capacitance 3.5 pF Response Time Small Signal, 10%-90%,0 mV-100 mV, CL = 2pF 400 ns

Large Signal, 10%-90%,0 V-2.4 V, CL = 2 pF 500 ns Upper Usable Frequency 3 500 MHz Lower Usable Frequency Input AC-coupled 10 HzAMPLIFIER CELL CHARACTERISTICS Cell Bandwidth –3 dB 900 MHz Cell Gain 14.3 dBINPUT CHARACTERISTICS DC Common-Mode Voltage Inputs AC-Coupled 3.2 V Common-Mode Range Either Input (Small Signal) –0.3 1.6 VS–1 V DC Input Offset Voltage 4 RSOURCE ≤ 50Ω 50 500 µV

Drift 0.8 mV/ºC Incremental Input Resistance Differential 1.1 kΩ

Input Capacitance Either Pin to Ground 1.4 pF Bias Current Either Input 10 25 mAPOWER INTERFACES Supply Voltage 2.7 5.5 V Supply Current VENB ≥ 2 V 8 10 mA Disabled VENB ≤ 1 V 150 750 µA

NOTES 1 This may be adjusted downward by adding a shunt resistor from the Output to Ground. A 50 kΩ resistor will reduce the nominal slope to 20 mV/dB. 2 This may be adjusted in either direction by a voltage applied to Pin 5, with a scale factor of 8 dB/V. 3 See Application on 900 MHz operation. 4 Normally nulled automatically by internal offset correction loop. May be manually nulled by a voltage applied between Pin 3 and Ground; see APPLICATIONS. Specifications subject to change without notice.

Recommended Operating Conditions

Parameter SymbolValue

UnitMin Typ Max

Supply Voltage VCC 4.5 5.0 5.5 V

Output Current IO 1.2 mA

Ambient Temperature TA –40 +85 ºC

Load Capacitance CL 12 pF

Typical application example

Test circuit

The project is an advanced electrocardiograph witha much greater accuracy than the electrocardio-graphs used in m ost of the world's hospitals today.Rem arkably, the project allows patients to record theirown ECG at home, and send the relevant data tothe doctor's office by modem.

By Jack and Mark Nowinski

it a0/

soft,cis

98.r e99

electrocardiographInternational First Prize

HOW -IS YOUR HEART?

The project, which involves both hard-ware and software, covers all the threethemes of the International PC

Software Competition: measurement,development, and communications.The software is the most instrum ental toolused in the project on account of thehigh accuracy it achieves through theuse of signal transforms and real-timeprocessing of the electrocardiogramsignal supplied by the hardware unit.As for the first theme, measurement,the software program is able to graphand analyze the full signal spectrumcoming from the hardware unit.Through this analysis of the signal, thesoftware program is able to measurevery critical aspects of the electrocar-

diogram (ECG) such as the QRS com-plex, the frequency of the heartbeatand the number of beats per minute(no electrocardiograph in the worldhas such a feature built into onedevice). Due to the powerful instruc-tion set of fourth and fifth generation ofIntel x86 processors, the mathematicaltransforms used for the measurementof the signal enable the most accu-rate electrocardiograph in the worldto be created.This project is full of development; first-ly the hardware and software weredesigned and developed fromscratch. A lot of development taskstook place both on the software andthe hardware. Development will

O

always be ongoing on account of thesoftware, which has the ability to cre-ate databases for people (patients)using the electrocardiographdescribed here.Also, the hardware is has beendesigned to be expandable, and soft-ware function routines are provided todetect if the PC processor has MMXinstructions for future portability andcompatibility. In that way, the projectcan always take full advantage of theprocessor when needed.As for communication there are twoways this project communicates withthe outside world, (1) through a printerport (parallel port), and (2) through amodem. The control program commu-nicates with the electronic hardwarevia the printer port in which data is sentrapidly in a bi-directional fashionbetween the computer and the elec-tronic hardware. Through the printerport the software is able to control theentire electronic hardware. The sec-ond way in which the program com-municates with the outside world is

through a modem. The project is pri-marily intended to be used by peoplein their homes. A heart patient with acomputer at home would apply thestandard medical electrodesto his/herbody and be able to transmit his/herECG to a doctor's office so the doctorwould then be able to study the ECGand determine if medical help is

needed. The ECG signal can be trans-mitted in real-time over the modemand accurate up-to-date informationcan then be processed by a medicaldoctor, the software program itself hasthe capability for medical analysis.

2 - 1/99 Elektor Electronics EXTRA Pa Topics

The project, which involves both hard-ware and software, covers all the threethemes of the International PCSoftware Competition: measurement,development, and communications.The software is the most instrumental toolused in the project on account of thehigh accuracy it achieves through theuse of signal transforms and real-timeprocessing of the electrocardiogramsignal supplied by the hardware unit.As for the first theme, measurement,the software program is able to graphand analyze the full signal spectrumcoming from the hardware unit.Through this analysis of the signal, thesoftware program is able to measurevery critical aspects of the electrocar-

diogram (ECG) such as the QRS com-plex, the frequency of the heartbeatand the number of beats per minute(no electrocardiograph in the worldhas such a feature built into onedevice). Due to the powerful instruc-tion set of fourth and fifth generation ofIntel x86 processors, the mathematicaltransforms used for the measurementof the signal enable the most accu-rate electrocardiograph in the worldto be created.This project is full of development; first-ly the hardware and software weredesigned and developed fromscratch. A lot of development taskstook place both on the software andthe hardware. Development will

always be ongoing on account of thesoftware, which has the ability to cre-ate databases for people (patients)using the electrocardiographdescribed here.Also, the hardware is has beendesigned to be expandable, and soft-ware function routines are provided todetect if the PC processor has MMXinstructions for future portability andcompatibility. In that way, the projectcan always take full advantage of theprocessor when needed.As for communication there are twoways this project communicates withthe outside world, (1) through a printerport (parallel port), and (2) through amodem. The control program commu-nicates with the electronic hardwarevia the printer port in which data is sentrapidly in a bi-directional fashionbetween the computer and the elec-tronic hardware. Through the printerport the software is able to control theentire electronic hardware. The sec-ond way in which the program com-municates with the outside world isthrough a modem. The project is pri-marily intended to be used by peoplein their homes. A heart patient with acomputer at home would apply thestandard medical electrodes to his/herbody and be able to transmit his/herECG to a doctor’s office so the doctorwould then be able to study the ECGand determine if medical help isneeded. The ECG signal can be trans-mitted in real-time over the modemand accurate up-to-date informationcan then be processed by a medicaldoctor, the software program itself hasthe capability for medical analysis.

2 - 1/99 Elektor Electronics EXTRA ——————————————— PC TOPICS

The project is an advanced electrocardiograph witha much greater accuracy than the electrocardio-graphs used in most of the world’s hospitals today.Remarkably, the project allows patients to record theirown ECG at home, and send the relevant data tothe doctor’s office by modem.

By Jack and Mark Nowinski

electrocardiographInternational First Prize

softw

ar

e·

available·on

·cd-rom

PCsoftware'98-'99

The programThe software used to produce or com-pile the software was Borland C++version 5.01 (with in -line assembly).The software program is b ased arounda graphical -user -interface (GUI) of theauthors' own design, see Figures la-d. All software operations take placemainly in the menu on the left-handside of the screen and some measure-ment specific operations take place inthe graphing window. The function ofthe software program is to graph anECG on the screen and if necessarytransmit it to a doctor's office (the doc-tor would have to be using the elec-trocardiograph software program to

receive the ECG signal). The programis able to receive data from two chan-nels (more can be added, but for stan-dard medical and home use one/two

a

b

channel(s) are sufficient). Both chan-nels can be displayed at the sametime, while one may be detecting theso-called QRScomplex in the ECG, theother can detect other arrhythmia's.Another first for an electrocardiographis the use of a TIME BASE for an accu-rate interpretation and analysis of theECG signal. The other options, X-MAG,TRIGGER, MODEM and FILE aredescribed in the interactive multime-dia presentation on the CD-ROM sup-plied (see end of article). To summa-rize, the X-MAG even further dividesthe time base setting by constant inte-ger dividers, the TRIGGER control but-ton sets the triggering level andMODEM control button initializes themodem (along with the telephonenum berto be called)and tellsthe pro-gram to transm it the ECG over the tele-

CH1

CH2

CH162

Tine Rene

X - Mag

Trigger OFF

Trg.Leyel

MEASUREMENT

MODEM

FILE

QUIT

,,,-1,,D - Dso, snd rlsrk ,saInskl

d

phone line. In FILE you can open andsave ECG files and thus help to build adatabase that could be used in thefuture by a medical doctor. The entiregraphics library used by the programhad to be written from scratch, andthe authors also had to configure theinterrupts for the printer port and initial-ize various types of modems and theirprotocols. In summary, the function ofthe software program is to graphincoming ECG signals (from the hard-ware unit through the printer port), andperform a very intense analysis of thesignal and finally save the waveformand/or transmit it to a doctor's office;the software program measures theECG signal, develops a database,and communicates with the outsideworld by transmitting the signal to a

medical facility.

CH1

CH2

CH1D2

Tine Rene

X - Mag

Trigger OFF

Trg.Level

MEASUREMENT

MODEM

FILE

QUIT

File cireateld on:I FrilSep 64 131:28:3O 1998

eTz763.16 neec Frenz1.3 Hz 78.6 BPM

Lod Ayr!: dzylhz.21

CH1

CH2

CH1D2

Tine Rasa

X - Mas

Trigger OFF

Trg.Level

MEASUREMENT

MODEM

FILE

QUIT

ePil0 0Qii

File created on: Fri Oct 2 21:05:8 19'8

Cc/ Copyright 1997-1998 -

Figure la. This screenshot has the ECG waveform already plotted. The control buttons on the left manipulate how the incoming ECGsignal isprocessed. For example, by clicking the right or left mouse button you can increase or decrease the time base on the grid.Figure lb. This screenshot is like the one above but has the measurement feature activated. The time base (lower right) is a factorwhen calculating the exact time (delta t).Figure lc. This screenshot shows the program operating under a different time base (100.0 m sec/d iv).Figure ld. Measurement results for the image above. Notice that the peaks are farther apart on screen but the time base is utilized toproduce an accurate measurement.

PC Topic Ele kto r Electronics EXTRA 3 - 1/99

The programThe software used to produce or com-pile the software was Borland C++version 5.01 (with in-line assembly).The software program is based arounda graphical-user-interface (GUI) of theauthors’ own design, see Figures 1a-d. All software operations take placemainly in the menu on the left-handside of the screen and some measure-ment specific operations take place inthe graphing window. The function ofthe software program is to graph anECG on the screen and if necessarytransmit it to a doctor’s office (the doc-tor would have to be using the elec-trocardiograph software program toreceive the ECG signal). The programis able to receive data from two chan-nels (more can be added, but for stan-dard medical and home use one/two

channel(s) are sufficient). Both chan-nels can be displayed at the sametime, while one may be detecting theso-called QRS complex in the ECG, theother can detect other arrhythmia’s.Another first for an electrocardiographis the use of a TIME BASE for an accu-rate interpretation and analysis of theECG signal. The other options, X-MAG,TRIGGER, MODEM and FILE aredescribed in the interactive multime-dia presentation on the CD-ROM sup-plied (see end of article). To summa-rize, the X-MAG even further dividesthe time base setting by constant inte-ger dividers, the TRIGGER control but-ton sets the triggering level andMODEM control button initializes themodem (along with the telephonenumber to be called) and tells the pro-gram to transmit the ECG over the tele-

phone line. In FILE you can open andsave ECG files and thus help to build adatabase that could be used in thefuture by a medical doctor. The entiregraphics library used by the programhad to be written from scratch, andthe authors also had to configure theinterrupts for the printer port and initial-ize various types of modems and theirprotocols. In summary, the function ofthe software program is to graphincoming ECG signals (from the hard-ware unit through the printer port), andperform a very intense analysis of thesignal and finally save the waveformand/or transmit it to a doctor’s office;the software program measures theECG signal, develops a database,and communicates with the outsideworld by transmitting the signal to amedical facility.

PC TOPICS——————————————— Elektor Electronics EXTRA 3 - 1/99

Figure 1a. This screenshot has the ECG waveform already plotted. The control buttons on the left manipulate how the incoming ECGsignal is processed. For example, by clicking the right or left mouse button you can increase or decrease the time base on the grid.Figure 1b. This screenshot is like the one above but has the measurement feature activated. The time base (lower right) is a factorwhen calculating the exact time (delta t).Figure 1c. This screenshot shows the program operating under a different time base (100.0 msec/div).Figure 1d. Measurement results for the image above. Notice that the peaks are farther apart on screen but the time base is utilized toproduce an accurate measurement.

a

b

c

d

The hardware

As shown by the block diagram inFigure 2, the hardware is composedof amplifiers, active filters, an ana-logue to digital converter, opto-isola-tors, and a DC/DC converter andinverter. A highly accurate low-passand high-pass filter combination(active Butterworth Fourth-order filter)was designed and produced to elimi-nate the extraneous noise producedby power lines and skin movementsunder the electrodes. The filter combi-nation has an extremely high roll-offrate thus making sure that the requiredfrequencies are accepted and otherunwanted frequencies are rejected. ADC/DC converter and inverter were

designed and incorporated becausecertain integrated circuits require neg-ative and higher voltages other thanstandard TTL voltages.An analogue-to-digital converter (12-bit) is used to digitize the analogueECG signal into the digital domain sothe software program along with PCwill be able to recognize it, this alsoimproves the analysis of the signallater in the program. Opto-isolatorsare used to send the digitized signalfrom the hardware to the computer viathe printer port (LPT); opto-isolators areneed in this type of medical equip-ment to isolate the hardware supplyvoltage from that of the computer.Unfortunately, owing to lack of space itis not possible to produce all schemat-

ics and PCB artwork for the hardwaredeveloped by the authors. The circuitdiagram of the ECG input amplifier is,however, given as a sample inFigure 3. The inputs of this circuit areconnected to standard ECG elec-trodes as demonstrated in Jack andMark’s wonderful video clip in whichthey describe the development andbasic operation of their project. Wellworth viewing!

(992010)

All software, source code files, schematicfiles, PCB artwork files and a demonstrationvideo (AVI file) as supplied by the authorsmay be found on a CD-ROM which will beavailable from the Publishers by earlyJanuary 1999.


transducer

input

ECGPreamp

DC - DCConverter

60 Hz(50 Hz)ActiveNotch

+12V

-12V

+12V

-12V

CH0

CS

CLK

DI

DO

SSO

CS

CLK

DI

DO

SSO

CH1

992010 - 11

BATT.

3...16V

+

HP LP

A - D

LP

LPTInterface

PC

LP

T p

ort

JP11

3

5

7

9

11

2

4

6

8

10

12

D1 D3

D4 D2 GNDSENSE

OUTSENSE

INV INP

LH0038

OFFSET

OFFSET

GUARDGAIN1

GAIN2

GAIN3

GAIN4

U1

VOUT

+IN

E+

14

16

E–10

15

11

13

12

9

1

2

3

4

5

6

7

8

C2

470p

C5

100n

C6

100n

C7

100n

C8

100n

C1

100p

C3

100p

R2

110k

R4

110k

R1

200k

R3

200k

R5

330k

C4

1µ

L1

10µH

L2

10µH

LEFT

RIGHT

GND

ACTIVE

NOTCH

4x BAT85

12V

12V

GAIN

JP1 GAIN

7 – 8 200

OPEN 100

4005 – 63 – 4

500

1000

2000

7 – 81 – 29 – 10

11 – 12

J1

J2

J3

992010 - 12

Figure 3. Circuit diagram of the ECG input amplifier. The circuit employs an LH0038 programmable-gain amplifier.

Figure 2. Block diagram of the electrocardiograph hardware.

This professionally designed 16 -channel logicanalyser runs under Windows and is connected thePC's parallel printer port. It requires a minimumam ount of hardware and costs a fraction of a com -parable stand-alone instrum ent. A PC -controlledinstrum ent, it offers ease of control and m any ways ofdisplaying measured signals.

Design by K. Bohm e

50 -MHz logic analyserFirst Prize, Germ any

El50 MHz - Logicanalyser MEM

= 440 ns = 2.6 vsf = 2.27 MHz f = 384.62 kHzCLK. = 22 CLK = 130

i? C

2nd

al

TRun MEI

1=11Exit

Main Specification>Linked to computer via EP port (parallel printer interface)>16 channels (3 V or 5 V input swing)>Sampling rate 1 kHz -50 MHz (or by external clock up to 50 MHz)>Adjustable pre -trigger from 1/8 to .7/8>Triggering by CHOO, CH15, adjustable bit pattern, or external>Adjustable min. trigger pulsewidth from <= 1,4,8,15 samples>Adjustable trigger edge>Frequency output 10 Hz -50 MHz>Output level of frequency output 3 V or 5 V.

The instrument is controlled via a virtu-al front panel which appears on yourPC screen. Apart from the customaryuser interface, the instrument offers alladvantages of a PC running under

Windows, including copying the read-out into temporary storage, and post -processing of measured data by otherWindows programs.The hardware consists essentially of

four Lattice FPGA devices typeispLSI1016 from Lattice. These devicesappear prominently in the circuit dia-grams in Figures 1 and 2. Circuit IC6in Figure 2 is busy handling all datatraffic on the EP port. The device num-ber (3 for the analyser) may be set ontwo jumpers (K3), and the delay for thewait signal, on jumper K4. The IC

receives an address write commandto inform it about the chip select line tobe set. The relevant IC is thenaddressed by the subsequent dataread or write command.The heart of the analyser is formed byIC10. This chip contains the entireprocess control system consisting of apresettable 12 -bit up/down counter

Pa Topic Ele kto r Electronics EXTRA 5 - 1/99PC TOPICS——————————————— Elektor Electronics EXTRA 5 - 1/99

The instrument is controlled via a virtu-al front panel which appears on yourPC screen. Apart from the customaryuser interface, the instrument offers alladvantages of a PC running under

Windows, including copying the read-out into temporary storage, and post-processing of measured data by otherWindows programs.The hardware consists essentially of

four Lattice FPGA devices typeispLSI1016 from Lattice. These devicesappear prominently in the circuit dia-grams in Figures 1 and 2. Circuit IC6in Figure 2 is busy handling all datatraffic on the EP port. The device num-ber (3 for the analyser) may be set ontwo jumpers (K3), and the delay for thewait signal, on jumper K4. The ICreceives an address write commandto inform it about the chip select line tobe set. The relevant IC is thenaddressed by the subsequent dataread or write command. The heart of the analyser is formed byIC10. This chip contains the entireprocess control system consisting of apresettable 12-bit up/down counter

This professionally designed 16-channel logicanalyser runs under Windows and is connected thePC’s parallel printer port. It requires a minimumamount of hardware and costs a fraction of a com-parable stand-alone instrument. A PC-controlledinstrument, it offers ease of control and many ways ofdisplaying measured signals.

Design by K. Böhme

50-MHz logic analyserFirst Prize, Germany

Main Specification Linked to computer via EP port (parallel printer interface) 16 channels (3 V or 5 V input swing) Sampling rate 1 kHz – 50 MHz (or by external clock up to 50 MHz) Adjustable pre-trigger from 1/8 to .7/8 Triggering by CH00, CH15, adjustable bit pattern, or external Adjustable min. trigger pulsewidth from <=1, 4, 8, 15 samples Adjustable trigger edge Frequency output 10 Hz – 50 MHz Output level of frequency output 3 V or 5 V.

software·ava

ilable·on·cd

-rom

PCsoftware

'98-'99

1.1111111

TH200( (extern)

K6 R13M

N33 ZSSAL41

II

vcc-.2

.010

RN2

IC11c

741°C1.7

P"" (C"" R1

33

HMIs (exlem)

VCC-2

H H5

RN3

tvo4

L74AC641VCC

-A '"1 r

0129 74PC14ZS

502

, 03

D1

ty.

Vc

I002 KD10

5184 10'3.-

A01 4,-0203 7,0:rD4 AlmO 5 A051De 0E1

07 A02

O

CEI

vccOCLK-ext TOSC

A 0000_00OADR 31OADR 02OAOADR:34OADR_35OAR_38OADR

013, a OADR OS00¢ 4 -A1 ,----, , ClALIN_10

A11 .- . OAOR_11OADR_12

ff

O

4, 07 CND 07,,Oyu

2k D2

D2Da134

D5130

VCC/01

62

51°4 '11L3V--'

A02

SA00qozqagApo

A10A11 a

GNDA12

tVM 176( 1C-5 MHz

RD

0703

,,T a=A02A03A04

Aa(.=

I420

02

D05126

141OEfril,MTS,

OCIM

ORS.

ICS I ADR-CNT-LS11010

Cl

c.(4550001CH0201-103CH04

0009CH05

01-10701-108 0000CHM 0001---gr CH10 A002CH11 ADO3 .

CHADO4--a 13 .0005r

01414 ADMCH15

GNDADO]

a

ZS SAL41H

V69-73

98 ICileR2

33

SAT43

±CNIP L511011.

VCC

C50. CS1

CSC

A00, A01

A02

: A04A05

VCC

RDwA .r.jmiOM

A10

624A13

A10.1 -LS 1016

RAM R131 TIN 0 eg--,RAM R122 TIN 1 a

11 RAM G ICMPTEXT a

C ACLK

- it RAM VVR

Input -US Dr--,ONO

II

ZSSA1.41747°C14

Fig ures 1 and 2. In essence the hardware of the Logic Analyser consists of four isp LSI c hips from Lattice.

5V/3V

/RD AnteMR /05).IReOet IA516

EPP-LS111216-1

DO ADDD1 AD1D2 AD2DO ADS124 AD4D5 ADSD5 A06D7 AD7

C50ICS1(C52

/C55CS6/C57

with decoding logic. At the start of ameasurement, the pre -trigger value isloaded, and this is counted down tozero. Once at 00, the decoding logicdisables itself, reverts the count d ire c -tion, and then enablesthe trigger con-trol. If a trigger pulse arrives at thatinstant, the counter is reloaded withthe pre -trigger value, and counts up

D

17

OUT IN

'78 G

22009 DI L -A

D7 DIODE

5V

115 V

29V

until an overflow is produced. It thendisables itself, as well as the RAM s andtheir address counters. This concludesthe measurement, and the datastored in the RAMs may be read by thesystem .The same IC also contains the inputmultiplexers for the timebase, the trig-ger source selection logic, and the

C 1 0

L.

992011 - 11

5x100n

C14VCC

992011 - 12

220V

counter for the minimum triggerpulsewidth.

The control signals supplied by thissub circuit are:- RAM RD_1 : read RAM 1- RAM RD2:_ read RAM2- RAM_G: output enable for both

RAMs

6 - 1/99 Elektor Electronics EXTRA PC Topics

with decoding logic. At the start of ameasurement, the pre-trigger value isloaded, and this is counted down tozero. Once at 00, the decoding logicdisables itself, reverts the count direc-tion, and then enables the trigger con-trol. If a trigger pulse arrives at thatinstant, the counter is reloaded withthe pre-trigger value, and counts up

until an overflow is produced. It thendisables itself, as well as the RAMs andtheir address counters. This concludesthe measurement, and the datastored in the RAMs may be read by thesystem.The same IC also contains the inputmultiplexers for the timebase, the trig-ger source selection logic, and the

counter for the minimum triggerpulsewidth.

The control signals supplied by thissubcircuit are:- RAMRD_1: read RAM1- RAMRD_2: read RAM2- RAM_G: output enable for both

RAMs


Figures 1 and 2. In essence the hardware of the Logic Analyser consists of four ispLSI chips from Lattice.

- ACLK: clock signal for addresscounter

- RAM_WR: write signal for both RAMs- BOE: output enable for input

drivers IC4/IC5- Input_UB: goes to Re1 and there

controls the voltageswitchover for the inputand output drivers (IC4,IC5 and IC11).

Circuit IC9 contains the dividers for thevariable output frequencies, two multi-plexers for their selection, and a 12-bitcounter which acts as an addresscounter for the two RAMs.IC1 contains an 8-bit multiplexer forthe reading of the two RAMs, the stor-age registers for the set trigger word,and a comparator which pulls theTFOUT output logic high when the lev-els at the analyser inputs match thoseset up in the trigger word.Circuit IC2 and IC3 are fast RAMs witha capacity of 8 kBytes each. The pro-gram uses about 4 kBytes, an exten-sion to 8 kBytes is already supportedby the PCB layout.Circuits IC4 and IC5 are input driversbetween the inputs and the RAMs. IC11acts as a driver for the signals at theoutput sockets, and at IC10. Relay Re1

allows all driver ICs to work with 3.5 V or5 V logic swing. Both supply voltagesare provided by IC7 and IC8.

Software functions anddocumentation

The 32-bit control program (written inVisual BASIC V5) takes care of all hard-ware control and also provides aneasy way of browsing the 4-kBytememory area. Further functionsinclude setting two marker lines, out-putting measurement values, and set-ting up trigger points and markersusing at the mouse cursor location.You can jump to the marker lines bymeans of a mouse click.Zooming is possible via ‘buttons’, whilean area to be examined in detail maybe enlarged by drawing a rectanglewith the mouse.Program settings as well as entire datafiles may be stored complete withcomment, for retrieval at a later time.For documentation purposes, the con-tents of the active window may becopies into temporary storage, fromwhich it may be loaded by other pro-grams including word processors.All channels may be displayed in acertain colour and they may be given

a name. To make sure the control ele-ments and the actual data display arequickly available without having tojump between partly obscured win-dows, the windows are dividedbetween two register cards (tabs), anddisplayed as fixed elements in a sepa-rate window which also contains thedata display.Apart from descriptions of the circuitdiagram, the circuit operation andprogram installation (in a Word file),the documentation as supplied by theauthor also includes the actual circuitdiagrams, the copper track layoutand component mounting plan of thedouble-sided printed circuit board (ina TIFF file), as well as all source codefiles and sample data files whichshould enable you to check the oper-ation of the program.The complete set of documentationfiles and program files covering theuse and installation of the programmay be found on a CD-ROM whichcontains a large number of winningentries from our 1998 Software DesignContest. This CD-ROM will be availableearly January 1999. More details willbe given in the forthcoming February1999 issue of Elektor Electronics.

(992011-1)


The winning design among the Competition entriesreceived by our Dutch sister magazine is aMicro Pascal compiler form ic ro c o ntro Ile rs from Intel'sMC51 series. The software enables these m icrocon-trollers to b e programmed in a sim p le way using the`Pascal' higher program m ing language. Helped by aseparate ROM emulator, MicroPascal may consider-ably reduce software debugging time.

So f tw a re by J. van de Kamer

MicroPscalt "'"'W,;,,,,alto.e:fi

Pr IZ (14! Net h e r a n

TheOixtode

IsCc

. 3

.000,CiOttiteAld

V.*

410J14

torgevot

tt too

fiLoatclt5

51,-7

004TIAE.

ot-tifffisPSG1311:1A

ao

Viltecwst

0 cep

t e 4 6t

yor ; totl'etall_ %coo'

nee' '0460".°

- 00*

trots? Iv's.41-001 vat el 4.1.W.Dx

....-'`co lgiX oaNvtikt ,. t co

''0*01"'"iv%

vtlto cow- '. t4lie,sot oca-,:ttr

00"

racits-s`t-voLir,

As t e author is ready to.omit,designing a complete caminvolves a vast amoun complexwork. Mr. van de 'amer started offwithin a--511and Pascal environment,but eventually changed to Delphi 1.0,while the final touches to the compilerwere made in Delphi 3.0. After abouttwo years of programming activity(approximately 750 hours) and animpressive 29,731 lines of sourcecode, the compiler was ready in theform as submitted for the SoftwareDesign Contest launched in theJuly/August 1998 issue of ElektorElectronics.

Corn piling is translating

The function of a compiler may becompared to that of a translator.Commands from a higher program -

d s

anguage are converted (trans-lated) into commands from a lower -level language. During the compilingprocess, the translator module will typ-ically encounter three types of `word':- reserved words such as

the commands Begin and End;- words representing a value,

for example, '1' or Joe';- words used as a label.

Comment enclosed in braces is

ignored during the compile process.The syntax of a source code file is

checked using a fixed procedure. Thefirst word that has to be found is

Program or Init. If not, an error report isimmediately returned. Once the rightheader is found, the translator will startto look for the next one, in this case, anidentifier. All relevant information onthe program (name, constants, vari-

a

(z)

s PC,t5L pft3 98, r9

w

bles, etc.) is then gathered andored together with the identifier. The

am ination of the program continuesu Ilan error is encountered, or a fullst (.) is found behind a command.In e compiler, every command is

pro essed as a procedure. If neces-sary, uch a procedure may call itself(this i called 'nesting ').

During the compilation process, thefirst-tim appearance of a new com-mand'causesthe associated machinecode to be generated. Because thecompiler as a flexible structure, thecomm. : to be generated are

red in a special library file calledMCS51.DLL.Before the actual compiler operationstarts, the requisite library has to beloaded into the memory.Initially the compiler does not know theaddresses, interrupts and labels to beused. Consequently, it will first generateso-called relocatable (address-inde-pendent)code. This c ode is stored asafile with the extension `MPU' (forMicroPascal Unit). This intermediatecode must not be changed by the userbecause the compiler will assume thatit has been generated without errors.The file format is universal, allowing theM PU file to be used in combination withother controllers, too.During the second phase, the file is

`linked', which means that labels, pro-cedures and functions set up in thesource code file are coupled to realaddresses. The result is a file that maybe programmed into an EPROM. Forthis purpose, MicroPascal sup ports twofile formats: Intel -hex and CPULink.

8 - 1/99 Elektor Electronics EXTRA PC Topics

As the author is ready to admit,designing a complete compilerinvolves a vast amount of complexwork. Mr. van de Kamer started offwithin a Borland Pascal environment,but eventually changed to Delphi 1.0,while the final touches to the compilerwere made in Delphi 3.0. After abouttwo years of programming activity(approximately 750 hours) and animpressive 29,731 lines of sourcecode, the compiler was ready in theform as submitted for the SoftwareDesign Contest launched in theJuly/August 1998 issue of ElektorElectronics.

Compiling is translating

The function of a compiler may becompared to that of a translator.Commands from a higher program-

ming language are converted (trans-lated) into commands from a lower-level language. During the compilingprocess, the translator module will typ-ically encounter three types of ‘word’: reserved words such as

the commands Begin and End; words representing a value,

for example, ‘1’ or ‘Joe’; words used as a label.

Comment enclosed in braces isignored during the compile process.The syntax of a source code file ischecked using a fixed procedure. Thefirst word that has to be found isProgram or Init. If not, an error report isimmediately returned. Once the rightheader is found, the translator will startto look for the next one, in this case, anidentifier. All relevant information onthe program (name, constants, vari-

ables, etc.) is then gathered andstored together with the identifier. Theexamination of the program continuesuntil an error is encountered, or a fullstop (.) is found behind a command.In the compiler, every command isprocessed as a procedure. If neces-sary, such a procedure may call itself(this is called ‘nesting’).

During the compilation process, thefirst-time appearance of a new com-mand causes the associated machinecode to be generated. Because thecompiler has a flexible structure, thecommands to be generated arestored in a special library file calledMCS51.DLL.Before the actual compiler operationstarts, the requisite library has to beloaded into the memory.Initially the compiler does not know theaddresses, interrupts and labels to beused. Consequently, it will first generateso-called relocatable (address-inde-pendent) code. This code is stored as afile with the extension ‘MPU’ (forMicroPascal Unit). This intermediatecode must not be changed by the userbecause the compiler will assume thatit has been generated without errors.The file format is universal, allowing theMPU file to be used in combination withother controllers, too.During the second phase, the file is‘linked’, which means that labels, pro-cedures and functions set up in thesource code file are coupled to realaddresses. The result is a file that maybe programmed into an EPROM. Forthis purpose, MicroPascal supports twofile formats: Intel-hex and CPULink.


The winning design among the Competition entriesreceived by our Dutch sister magazine is aMicroPascal compiler for microcontrollers from Intel’sMC51 series. The software enables these microcon-trollers to be programmed in a simple way using the‘Pascal’ higher programming language. Helped by aseparate ROM emulator, MicroPascal may consider-ably reduce software debugging time.

Software by J. van de Kamer

MicroPascalFirst Prize, Netherlands

softw

ar

e·

available·o

n·cd-rom

PCsoftware'98-'99

Memory useThe RAM memory is employed for thestorage of variables, procedures, asoftware ALU (arithmetic Logic Unit), astack and intermediate results(scratchpad)Memory is filled from the top to thebottom. For the internal RAM, the top is$FF, for the external RAM, $7FFFF (pro-vided 32 Kbytes of external RAM is

used). Global variables are held in thisexternal memory. One word, for exam-ple, occupies address $7FFFF for the`high' byte and $7FFFE for the 'low'byte. A string made up of 10 bytes isstored in 11 bytes: 10 bytes for thecharacters and number 11 to indicatethe length. The last character is storeat location $7FFFF, the first at $7FFF6,and the length indicator at $7FFF5.

Software ALU

To be able to perform arithmetic oper-ations using variables with a size largerthan one byte, a software ALU is builtinto the program. The size of the ALU isgeared to the largest value supportedby the program. The ALU is d ivided intotwo equal sections, lo_ALU and hi_ALU.The ALU is an important piece in thememory. It is used to move variables toand from the stack, and perform arith-metic operations.

Other points of interest

Once procedures and functions arebeing used, local variables are creat-ed that only exist within the relevantprocedure or function. Obviously, thecompiler has to reserve space forthese intermediate variables. The sim-plest approach is to define memoryspace for any procedure which popsup in the program. In practice, thisresults in a lot of wasted memoryspace. The present compiler first looksfor the procedure or function whichmay be expected to use up thelargest number of variables (in bytes).For this purpose, a special buffer areais made available behind the softwareALU. All local variables generated bythe program are stored in this area.The remainder of the RAM memory isavailable as stack space. MicroPascalworks lop down', starting at the high-est memory location ($FF). This systemhad to be adopted because the con-troller uses memory locations from $00upwards for its internal registers. Theprocessor's own stack is then locatedimmediately above these registers.Both blocks are therefore allowed togrow towards each other as the pro-gram is being executed, and no con -

4 -.:t Micro Pascal - Bounce

File Edit Project Options Emulator Help

h I IA .1 siAltHI 0.01 x I it A. I 1 > I I I

II IAA bounce.rom

0 1 2 3 4 5 6 7 8 9 ABCDEF000000 02 01 1C E6 11 A9 02 F7 08 DC F8 22 A9 02 E7 OA

000010 F6 18 DC F8 22 E6 1A A9 02 F7 22 A9 02 E7 OA F6

000020 22 78 F8 12 00 1B 78 F8 E6 78 F9 F6 E6 64 FF F6

000030 E6 78 F8 F6 78 F8 12 00 15 22 78 F8 7C 02 12 00

000040 OC 78 F5 76 01 08 76 00 78 F7 E6 78 F9 F6 78 F8

000050 E6 78 FA F6 78 F9 78 F5 E6 78 FB F6 78 F6 E6 78

000060 FC F6 78 F9 78 FC E6 FC 78 FA E6 D3 9C 50 12 B4

000070 FF OB 78 FB E6 FC 78 F9 E6 C3 9C 50 04 74 FF 80

000080 02 74 00 64 FF 78 F9 F6 E6 70 03 02 00 99 78 F5

000090 06 B6 00 02 08 06 02 00 48 78 F7 E6 78 F9 F6 78

0000A0 F8 E6 78 FA F6 78 F9 7C 02 12 00 03 78 FA 7C 02

0000B0 12 00 OC 78 FA E6 FC 74 07 D3 9C 50 11 B4 FF OA

000000 78 F9 E6 FC 74 DO C3 9C 50 04 74 FF 80 02 74 00

0000D0 78 F9 F6 E6 70 03 02 00 E3 78 F7 76 64 08 76 00

0000E0 02 01 14 78 F7 E6 78 F9 F6 78 F8 E6 78 FA F6 78

0000F0 F9 7C 02 12 00 03 78 FA 7C 02 12 00 OC 78 F9 E6

000100 24 32 F6 08 E6 34 00 F6 78 F9 E6 78 F7 F6 78 FA

000110 E6 78 F8 F6 78 F7 7C 02 12 00 03 22 7D 00 7A F5

000120 78 FF 76 FF 75 90 7F 78 FD 76 64 08 76 00 74 FF

000130 70 03 02 02 16 78 FD 7C 02 12 00 03 12 00 3A 78

000140 FE 7C 02 12 00 OC 78 FF E6 78 F9 F6 E6 70 03 02

000150 01 B4 E4 A2 90 33 78 F9 F6 12 00 15 78 F9 12 00

000160 1B 78 F9 E6 FC 74 00 6C 70 04 74 FF 80 02 74 00

000170 78 F9 F6 E6 70 03 02 01 89 78 FF 12 00 15 12 00

000180 21 78 FF 12 00 1B 02 01 B1 E5 90 78 F9 F6 12 00

000190 15 78 F9 12 00 1B 7C 01 78 F9 E6 C3 13 DC FC F6

0001A0 12 00 15 78 F9 12 00 1B 78 F9 E6 44 80 F6 E6 F5

L

Figure 1. Screendump of MicroPascal in action. The compiler offersa simple way ofdeveloping code form ic roco ntrollers from the M CS -51 fa m ily. The screens in this a rtic leshow the source code in a window (introductory illustration), and the final machinecode ready forprogramm ing into a PROM or EPROM.

flict will occur as long as sufficientmemory space remains available. If

you run low on available memory, sim-ply reduce the number of variablesand/or procedures being called.When calling a procedure it is alsopossible to convey param eters or vari-ables. If variables are marked with

`VAR' they may be adapted within theprocedure called.To be able to use interrupts, specialprovisions have to be made in thecode. This is necessary because an

interrupt may occur at any moment.The interrupt procedure ensures thatall data of local variables, as well asALU data, are safely stored before aninterrupt procedure is started. The

in to rn a I re g iste rs a re a lso ke pt in a sa feplace. At the end of the interrupt, allrelevant information is retrieved andrestored to its original state.Interrupts may be used to adapt glob-ally defined variables. We should has-ten to add, however, that variableslarger than one byte can not be

Variables and units

Type value minimum maximum

Byte numerical 0 255

Word numerical 0 65535

Shortlnt numerical -128 127

Integer numerical -32768 327767

Boolean numerical False True

Char character 1 character 1 character

String character 0 characters size dependent

Fora string, the size dependson allocated memory space.When no length is spe c ifed, the length is lim ited to 255 characters.

Pa Topic Elektor Electronics EXTRA 9 - 1/99

Memory useThe RAM memory is employed for thestorage of variables, procedures, asoftware ALU (arithmetic Logic Unit), astack and intermediate results(scratchpad)Memory is filled from the top to thebottom. For the internal RAM, the top is$FF, for the external RAM, $7FFFF (pro-vided 32 Kbytes of external RAM isused). Global variables are held in thisexternal memory. One word, for exam-ple, occupies address $7FFFF for the‘high’ byte and $7FFFE for the ‘low’byte. A string made up of 10 bytes isstored in 11 bytes: 10 bytes for thecharacters and number 11 to indicatethe length. The last character is storeat location $7FFFF, the first at $7FFF6,and the length indicator at $7FFF5.

Software ALU

To be able to perform arithmetic oper-ations using variables with a size largerthan one byte, a software ALU is builtinto the program. The size of the ALU isgeared to the largest value supportedby the program. The ALU is divided intotwo equal sections, lo_ALU and hi_ALU.The ALU is an important piece in thememory. It is used to move variables toand from the stack, and perform arith-metic operations.

Other points of interest

Once procedures and functions arebeing used, local variables are creat-ed that only exist within the relevantprocedure or function. Obviously, thecompiler has to reserve space forthese intermediate variables. The sim-plest approach is to define memoryspace for any procedure which popsup in the program. In practice, thisresults in a lot of wasted memoryspace. The present compiler first looksfor the procedure or function whichmay be expected to use up thelargest number of variables (in bytes).For this purpose, a special buffer areais made available behind the softwareALU. All local variables generated bythe program are stored in this area.The remainder of the RAM memory isavailable as stack space. MicroPascalworks ‘top down’, starting at the high-est memory location ($FF). This systemhad to be adopted because the con-troller uses memory locations from $00upwards for its internal registers. Theprocessor’s own stack is then locatedimmediately above these registers.Both blocks are therefore allowed togrow towards each other as the pro-gram is being executed, and no con-

flict will occur as long as sufficientmemory space remains available. Ifyou run low on available memory, sim-ply reduce the number of variablesand/or procedures being called.When calling a procedure it is alsopossible to convey parameters or vari-ables. If variables are marked with‘VAR’ they may be adapted within theprocedure called.To be able to use interrupts, specialprovisions have to be made in thecode. This is necessary because an

interrupt may occur at any moment.The interrupt procedure ensures thatall data of local variables, as well asALU data, are safely stored before aninterrupt procedure is started. Theinternal registers are also kept in a safeplace. At the end of the interrupt, allrelevant information is retrieved andrestored to its original state.Interrupts may be used to adapt glob-ally defined variables. We should has-ten to add, however, that variableslarger than one byte can not be


Figure 1. Screendump of MicroPascal in action. The compiler offers a simple way ofdeveloping code for microcontrollers from the MCS-51 family. The screens in this articleshow the source code in a window (introductory illustration), and the final machinecode ready for programming into a PROM or EPROM.

Variables and units

Type value minimum maximum

Byte numerical 0 255

Word numerical 0 65535

ShortInt numerical -128 127

Integer numerical -32768 327767

Boolean numerical False True

Char character 1 character 1 character

String character 0 characters size dependent

For a string, the size depends on allocated memory space.When no length is specifed, the length is limited to 255 characters.

Setup

Project Target I Environment

-Micro Controller

Type: iisomaissimmummommom_71

1256ISLI

$2000

Internal RAM: External RAM: I RO= Offset: M153E000

=-Debug signals

Clock address:

I$o

Data address:

Iso

Acknowledge address:

'so

K

Figure 2. A special config uratuon window allows the characteristics of the M C S-51 sys-tem ('terg et system') of be set up.

processed. Larger variables are modi-fied with the aid of several instructionshaving a width of one byte. To com-municate with the main program, it is

therefore recommended to use vari-ables with a width of one byte.

Debugging

Eliminating errors in software is almostinvariably a tedious and time-consum-ing activity. MicroPascal hasa numberof extra functions available to simplythe debugging process. Additionalhardware has been designed in the

Statement

Id t

form of a ROM emulator (size

256 Kbytes). The circuit diagram andPCB artwork for this design may befound on the CD-ROM which containsa number of prize-winning entries fromthe 1998 Software DesignCompetition. This CD-ROM will be pub-lished early January 1999. The rele-vant file is called ROM emulator.DOC.

During the linking of a program, eachdebugging location has its own identi-fication (a value between 0 and 255).This value is saved by the compiler,together with the location of the

IdentL,

Co cl on

Cnatlon

MEM El=

MI= =IEl= El=

=I =1 11=11

C ndron

111:111 0 ECM ME EMS NM =I

!dent <DI

Figure 3. This flow diagram shows the ingenious m ethods used by the compiler todeterm ine what is m ea nt by a statement in the source code.

debugging point (file name and linenumber). The debugging procedure isimplemented in the software by call-ing a function from the library that firstsends the identification, and then thecontentsof the local and global mem-ory. MicroPascal 'knows' which vari-ables are visible at the debuggingpoint marked by the identification,and displays the received value plusassociated variables.This approach requires three I/O pinsof the MCS-51 to be given a specificfunction. Two outputs, one carrying theclock signal, the TxD line, plus oneinput employed for handshaking. Asimple handshaking protocol is

employed. The controller pulls bit 0 atthe data output logic low, and placesa low level on the clock line.MicroPascal reads the level and pullsthe acknowledge line logic low. Next,the controller pulls the clock signalhigh gain, whereupon MicroPascaldoes the same with the acknowledgesignal. The first bit is then transmitted.This procedure is com pleted for all bitsthat have to be sent, except the lastone. With this bit, MicroPascal nolonger supplies an acknowledge sig-nal, causing the microcontroller to

enter an infinite loop. During this peri-od, the results may be viewed andanalysed on the PC display. Once allrelevant information is known, you onlyhave to actuate the menu option`Continue' to transmit one acknowl-edge signal and so get the m icrocon-troller out of its infinite loop. The pro-gram is then continues as before.The primitive handshake was chosento make sure reliable communicationis available under all circumstances.However, the price of three controllerI/O pins may be too high in somecases, so that the function is optional.

System requirements

MicroPascal runs under Windows 95and requires about 10 Mbytes of harddisk space. If the optional ROM emu-lator is employed, then you also needa free bidirectional printer port.MicroPascal was developed for MCS-51 microcontrollers and comes with alibrary for these processors only. Usersrequiring support for other microcon-trollers will have to develop their ownlibraries.

(992012-1)

10 - 1/99 Ele ktor Electronics EXTRA Pa Topics

processed. Larger variables are modi-fied with the aid of several instructionshaving a width of one byte. To com-municate with the main program, it istherefore recommended to use vari-ables with a width of one byte.

Debugging

Eliminating errors in software is almostinvariably a tedious and time-consum-ing activity. MicroPascal has a numberof extra functions available to simplythe debugging process. Additionalhardware has been designed in the

form of a ROM emulator (size256 Kbytes). The circuit diagram andPCB artwork for this design may befound on the CD-ROM which containsa number of prize-winning entries fromthe 1998 Software DesignCompetition. This CD-ROM will be pub-lished early January 1999. The rele-vant file is called ROM emulator.DOC.

During the linking of a program, eachdebugging location has its own identi-fication (a value between 0 and 255).This value is saved by the compiler,together with the location of the

debugging point (file name and linenumber). The debugging procedure isimplemented in the software by call-ing a function from the library that firstsends the identification, and then thecontents of the local and global mem-ory. MicroPascal ‘knows’ which vari-ables are visible at the debuggingpoint marked by the identification,and displays the received value plusassociated variables.This approach requires three I/O pinsof the MCS-51 to be given a specificfunction. Two outputs, one carrying theclock signal, the TxD line, plus oneinput employed for handshaking. Asimple handshaking protocol isemployed. The controller pulls bit 0 atthe data output logic low, and placesa low level on the clock line.MicroPascal reads the level and pullsthe acknowledge line logic low. Next,the controller pulls the clock signalhigh gain, whereupon MicroPascaldoes the same with the acknowledgesignal. The first bit is then transmitted.This procedure is completed for all bitsthat have to be sent, except the lastone. With this bit, MicroPascal nolonger supplies an acknowledge sig-nal, causing the microcontroller toenter an infinite loop. During this peri-od, the results may be viewed andanalysed on the PC display. Once allrelevant information is known, you onlyhave to actuate the menu option‘Continue’ to transmit one acknowl-edge signal and so get the microcon-troller out of its infinite loop. The pro-gram is then continues as before.The primitive handshake was chosento make sure reliable communicationis available under all circumstances.However, the price of three controllerI/O pins may be too high in somecases, so that the function is optional.

System requirements

MicroPascal runs under Windows 95and requires about 10 Mbytes of harddisk space. If the optional ROM emu-lator is employed, then you also needa free bidirectional printer port.MicroPascal was developed for MCS-51 microcontrollers and comes with alibrary for these processors only. Usersrequiring support for other microcon-trollers will have to develop their ownlibraries.

(992012-1)


Figure 2. A special configuratuon window allows the characteristics of the MCS-51 sys-tem (‘ terget system’) ot be set up.

Figure 3. This flow diagram shows the ingenious methods used by the compiler todetermine what is meant by a statement in the source code.

New so ftw a re to sim u late lo g is c irc u its se e m s to hitthe streets every month, and most of you will beaware that logic sim ulation on a PC is a good alter-native to wiring up prototypes of d ig ita I c irc u its.

By J. P Strza lka

Logic Simulator 2.0First prize, France

The logic simulation program `Simulat2.0' was awarded the First NationalPrize for Competition entries sent to ourFrench editorial department. The pro-gram is aimed at students, teachersand hobbyists wishing to widen theirknowledge of sequential and combi-natory logic circuit design.Sim ulat 2.0 allows you to drawschematics of logic circuits in a win-dow displayed on your PC monitor.Once you think the schematic is com-plete, you launch a simulation run

using virtual switches, displays or otherindicators and devices to control andvisualize the way the logic circuit oper-ates (or not!). Sim ulat 2.0 is capable ofhandling four schematics at a time.

Program installationThe program is easy to install under theWindows 95 operating system.Select 'Run' from the Start menu, type`X:\setup', and then press the Return

key, where `X:' is the letter of the drivestation that has the 'setup' file.Alternatively, use the 'Browse' option toget assistance from Windows 95 in

looking for the setup program. If youuse the Elektor CD-ROM which con-tains the present project, navigate tothe subdirectory /F/01 where you willfind setup.exe.The installation program builds a

group of three programs. The group iscalled 'Sim ulat', and it will containthree icons. The first icon in the group

allows you to actually launch Sim ulat.To do so, double-click on this icon withthe left-hand mouse button. In this wayyou start the simulator program.

Sc hem atic capture

The program enablesyou to create cir-cuit diagrams. For this function it offersa number of tools that allow users toinclude elementary logic elements likegates, flip-flops, adders, multiplexers,dem ultip lexers, timers or counters in

your schematic. A rather complexexample of such a circuit is shown bythe screendump in Figure 1. This is

actually a counter circuit.

Sim ulation program

The 'pencil' tool allows you to connectlogic gates. Only right angles are pos-sible for the 'wires'. To use the pencil,pick it from the Tools bar, and left -clickon an input or output of a logic oper-ator. Next, move the mouse to theoperator you want to connect up, andthen release the mouse button.The 'Copy' utility of Simulat also sup-ports the use of the Clipboard to cut,copy, delete and paste elements inyour circuit diagram.Sim ulat has deletion utilities: for logicoperators, connections and junctions.To delete an object, select the corre-sponding utility from the Tool bar.

The schematic may be spiced up a bitby text with or without a frame. It is pos-sible to choose fonts, colours andcharacter sizes, as well as text orienta-tion (vertical or horizontal). The usermay choose the frame colour, shade

Pa Topic,s Elektor Electronics EXTRA 11 - 1/99PC TOPICS —————————————— Elektor Electronics EXTRA 11 - 1/99

The logic simulation program ‘Simulat2.0’ was awarded the First NationalPrize for Competition entries sent to ourFrench editorial department. The pro-gram is aimed at students, teachersand hobbyists wishing to widen theirknowledge of sequential and combi-natory logic circuit design.Simulat 2.0 allows you to drawschematics of logic circuits in a win-dow displayed on your PC monitor.Once you think the schematic is com-plete, you launch a simulation runusing virtual switches, displays or otherindicators and devices to control andvisualize the way the logic circuit oper-ates (or not!). Simulat 2.0 is capable ofhandling four schematics at a time.

Program installationThe program is easy to install under theWindows 95 operating system.Select ‘Run’ from the Start menu, type‘X:\setup’, and then press the Returnkey, where ‘X:’ is the letter of the drivestation that has the ‘setup’ file.Alternatively, use the ‘Browse’ option toget assistance from Windows 95 inlooking for the setup program. If youuse the Elektor CD-ROM which con-tains the present project, navigate tothe subdirectory /F/01 where you willfind setup.exe.The installation program builds agroup of three programs. The group iscalled ‘Simulat’, and it will containthree icons. The first icon in the group

allows you to actually launch Simulat.To do so, double-click on this icon withthe left-hand mouse button. In this wayyou start the simulator program.

Schematic capture

The program enables you to create cir-cuit diagrams. For this function it offersa number of tools that allow users toinclude elementary logic elements likegates, flip-flops, adders, multiplexers,demultiplexers, timers or counters inyour schematic. A rather complexexample of such a circuit is shown bythe screendump in Figure 1. This isactually a counter circuit.

Simulation program

The ‘pencil’ tool allows you to connectlogic gates. Only right angles are pos-sible for the ‘wires’. To use the pencil,pick it from the Tools bar, and left-clickon an input or output of a logic oper-ator. Next, move the mouse to theoperator you want to connect up, andthen release the mouse button.The ‘Copy’ utility of Simulat also sup-ports the use of the Clipboard to cut,copy, delete and paste elements inyour circuit diagram.Simulat has deletion utilities: for logicoperators, connections and junctions.To delete an object, select the corre-sponding utility from the Tool bar.

The schematic may be spiced up a bitby text with or without a frame. It is pos-sible to choose fonts, colours andcharacter sizes, as well as text orienta-tion (vertical or horizontal). The usermay choose the frame colour, shade

New software to simulate logic circuits seems to hitthe streets every month, and most of you will beaware that logic simulation on a PC is a good alter-native to wiring up prototypes of digital circuits.

By J.P. Strzalka

Logic Simulator 2.0First prize, France

softw

are

·available·on

·cd-rom

PCsoftware'98-'99

Table of truth

and the frame type. Frame size is auto-matically adjusted to the amount oftext and character height selected bythe user.It is possible to write up some com-ment with each schematic. The lengthof the comment text is limited to 50characters.

Tools

The panoply of utilities found in theworking subdirectory is complement-ed by a powerful 'special logic' cal-culator which is capable of perform-ing special functions like number con-versions and shift -right /shift -left opera -

Calculator

r1

<-1 +Id 11

11

Et 1 Ou 1

2 1 3 1+ I Not] Xor 1

j5 Lsh] Rshl

999999999.0000001024

> Octet> Word

4 ".;t,',.",kK

Basis

Deci

Hexa

Octal

Binary 1

tions. Using binary number notation,the range of operands is adjustablefrom eight bits to double 32 -bit words.In addition, a truth tab le is availab le tohelp you rememberthe function of themain logic operators. This useful littletool pops up on the screen if youselect the corresponding command. Asimple mouse click on one of the logicoperator types in the scroll list instantlyd isp lays the associated truth tab le andthe corresponding logic symbol.

Circuit simulation

Having drawn and saved yourschematic, you are ready to launchthe circuit simulator. The program will

test all connections and display its

findings as the test progresses. Anyerror encountered in this process, is

made known to you by means of dia-logue boxes. If Simulat does notencounter errors, it is possible to oper-ate the virtual switches, indicatorlamps, displays and thumbwheelswitches, and so set up and manipu-late various conditions to which the vir-tual logic circuit responds.The links in the circuit diagram areshown in different colours dependingon their logic level. You can read thelogic level of an operator by left -click-ing on the operator input or output.The cursor will then change to a sym-bol indicating 'zero' or 'one'.Simulat 2.0 also comes with a logicoscilloscope capable of displaying upto five signals taken from points youindicate in the circuit diagram. Theoscilloscope is 'triggered' by either arising or a falling pulse edge detectedon one of the five input channels. If theoscilloscope trace is not stable, youmay insert a trigger delay to keep thisvirtual instrument synchronized.Oscilloscope traces may be copiedonto the Clipboard, or saved in a file.Furthermore, scope traces may besent to the printer to produce hardcopy, at a scaling factor you select.Alternatively, Sim ulat can compute thescale value needed to adapt the sizeof the drawing to that of the paper inthe printer.

Edition of generators

Generator 0 b

Number of work cycle

2.j_jNumber of rest cycle

'.... Start level 1

> Start level 0. .

Generator me

umber of work cycle

Number of rest cycle

Start level 1

> Start level 0 ,

10

10

Oa

Ob

Oc

Logic equations

Sim ulat allows you to determine logicequations based on a schematic.The calculation box identified as`Equations' allows you to print equa-tions that belong with a circuit dia-gram, using a number of options (linespacing, separation, page number,

etc.). Equations may be saved on theClipboard or in a file on your disk. It is

also possible to compute a partialequation. To do so, you click on thecorresponding button shown by thedialogue box. Simulat then temporari-ly closesthe box to enable you to clickon the desired logic gate. Once that is

Equations

,11(001h = Oc

J1K001j = ((f/b)/b) (a -1)+b)

JKOOkk = (0-1 -J1K002Q -111C0031Q -..11000

JIK003h =0c

JIK003j = ((d/b)/b) (a -1 -..11400110-..1

CloseEQUAT

done, the dialogue box pops upagain on the screen, showing thelogic equation against a red back-ground.

Library

Circuit elements may be saved in

order to build a library of basic build-ing blocks.To do so, you choose the option`Copy' and then limit the selection tothe desired elements by means of adashed box. Next, you actuate thecommand 'save model to disk' whichmay be found in the menu called`Library'.Finally, you may select the command`Load model from memory'. This caus-es a certain element to be retrievedfrom the library and placed at thedesired location in the current window.

Utilities & button palettes

The palettes (bars) containing utilitiesand buttons allow you to change theappearance of the program on thePC screen. Modifying the utilitiespalette allows you to increase the

actually visible area in the windowcontaining the schematic you areworking on. In this way, you will beable to see more of your schematic.These palettes may be moved aroundto any location you want on thescreen. Furthermore, the palettes andtoolbars may be 'hidden' to createeven more space on the screen.

12 - 1/99 Elektor Electronics EXTRA Pa To Pi a ,s12 - 1/99 Elektor Electronics EXTRA ——————————————— PC TOPICS

and the frame type. Frame size is auto-matically adjusted to the amount oftext and character height selected bythe user.It is possible to write up some com-ment with each schematic. The lengthof the comment text is limited to 50characters.

Tools

The panoply of utilities found in theworking subdirectory is complement-ed by a powerful ‘special logic’ cal-culator which is capable of perform-ing special functions like number con-versions and shift-right /shift-left opera-

tions. Using binary number notation,the range of operands is adjustablefrom eight bits to double 32-bit words.In addition, a truth table is available tohelp you remember the function of themain logic operators. This useful littletool pops up on the screen if youselect the corresponding command. Asimple mouse click on one of the logicoperator types in the scroll list instantlydisplays the associated truth table andthe corresponding logic symbol.

Circuit simulation

Having drawn and saved yourschematic, you are ready to launchthe circuit simulator. The program will

test all connections and display itsfindings as the test progresses. Anyerror encountered in this process, ismade known to you by means of dia-logue boxes. If Simulat does notencounter errors, it is possible to oper-ate the virtual switches, indicatorlamps, displays and thumbwheelswitches, and so set up and manipu-late various conditions to which the vir-tual logic circuit responds.The links in the circuit diagram areshown in different colours dependingon their logic level. You can read thelogic level of an operator by left-click-ing on the operator input or output.The cursor will then change to a sym-bol indicating ‘zero’ or ‘one’.Simulat 2.0 also comes with a logicoscilloscope capable of displaying upto five signals taken from points youindicate in the circuit diagram. Theoscilloscope is ‘triggered’ by either arising or a falling pulse edge detectedon one of the five input channels. If theoscilloscope trace is not stable, youmay insert a trigger delay to keep thisvirtual instrument synchronized.Oscilloscope traces may be copiedonto the Clipboard, or saved in a file.Furthermore, scope traces may besent to the printer to produce hardcopy, at a scaling factor you select.Alternatively, Simulat can compute thescale value needed to adapt the sizeof the drawing to that of the paper inthe printer.

Logic equations

Simulat allows you to determine logicequations based on a schematic.The calculation box identified as‘Equations’ allows you to print equa-tions that belong with a circuit dia-gram, using a number of options (linespacing, separation, page number,

etc.). Equations may be saved on theClipboard or in a file on your disk. It isalso possible to compute a partialequation. To do so, you click on thecorresponding button shown by thedialogue box. Simulat then temporari-ly closes the box to enable you to clickon the desired logic gate. Once that is

done, the dialogue box pops upagain on the screen, showing thelogic equation against a red back-ground.

Library

Circuit elements may be saved inorder to build a library of basic build-ing blocks.To do so, you choose the option‘Copy’ and then limit the selection tothe desired elements by means of adashed box. Next, you actuate thecommand ‘save model to disk’ whichmay be found in the menu called‘Library’.Finally, you may select the command‘Load model from memory’. This caus-es a certain element to be retrievedfrom the library and placed at thedesired location in the current window.

Utilities & button palettes

The palettes (bars) containing utilitiesand buttons allow you to change theappearance of the program on thePC screen. Modifying the utilitiespalette allows you to increase theactually visible area in the windowcontaining the schematic you areworking on. In this way, you will beable to see more of your schematic.These palettes may be moved aroundto any location you want on thescreen. Furthermore, the palettes andtoolbars may be ‘hidden’ to createeven more space on the screen.

Printing

Simulat 2.0 can provide a Print

Preview. This dialogue box allows youto select the printable area, the scale,page margins, contents of the ID box,and other printer configurations,

110)+1])-(c/b)

K00210)+b

Separation Spacing Jump line

55m55t

N° Page

Frame

Headlin

before the print job is started.The buttons in the dialogue box are`repeat' types, allowing the relevantcommand to be issued again bykeeping the cursor on the button rep-resenting the desired command.Compatibility is assured with any print-er already functioning in graphicsmode, and driven by other applica-tions running under Windows.To print an ID box with the schematic,click on the button `Cartouche'. A dia-logue box pops up which allows you todefine the text fields which are to

appear in the ID box.Paper margins may be set in the dia-logue box 'Page Setup' (Wise en

Pa g e'). In the title bar the program dis-play the paper dimensions and theprintable area. These two values will ofcourse depend on the printer youhave connected up to your PC.

Help system

A powerful, context -sensitive help sys-tem is available.To get help, you click on the`Contextual Help' symbol in the buttonbar. Next, you click your way throughthe Help menu to get to the item youneed help on.

File management

The file management system allowsyou to stay organized as far as yourschematics are concerned.The system offersyou a numberof pos-

NICE® ® 010101-.°]-.9J-.°1-91, :,1

sibilities for copying, moving, loading,opening and deleting your schemat-ics. It also supports displaying file

properties of disk drives, and freespace on the hard disk.

Miscellaneous matters

Because Sim ulat 2.0 runs under theWindows 95 operating system, you areable to copy complete screens intothe Clipboard, and from there movethem to other applications like MS

Write, etc., in order to create docu-ments (useful for educational/didacticpurposes).Simulat 2.0 supports the MultipleDocument (MD) system The programallows you to work on four circuit dia-grams simultaneously. The names ofthe schematics are displayed in the`Window' menu (Tenetre'). You canchange between circuit diagrams inthree ways:- by left -clicking anywhere in the win-

dow of the schematic to be activated;- by clicking on the name of the

schematic as listed in the 'Window'menu;- by selecting the 'Next' option fromthe System menu in a window.

Simulat 2.0 also supports the OLE 2.0system. The linking and embeddingfunctions allow Simulat 2.0 to acceptdifferent data types and objects pro-duced with other programs. The maindifference between linking an objectand embedding it concerns the wayin which data is stored: embeddedobjects form part of a schematic (i.e.,

they are integrated into the file), whilelinked objects remain stored in theoriginal (source) file. The schematicthen only keeps information relating tothe location of the object in the sourcefile.Circuit diagram files may be selectedand moved into the simulation win-dow. Using the file management sys-tem you may pick one or moreschematics and drop them in the sim-ulation window. Once the cursor is onthe program window, release themouse button, and the selected filesare loaded into the simulator. Notethat Sim ulat 2.0 can only handle fourschematics at a time.

System requirem ents

You will need to have Windows 95installed (or a later version). The PCshould have a 386 processor or better,4 Mbytes of RAM, a VGA card and amouse. Simulat 2.0 requires about6 Mbytes of free space on your harddisk.

Note: The version described here supportsa maximum of 100 components. Theauthor can supply a larger version.

(992014-1)

SIMULAT 200 SHAREWARE 11113111Files Edit Text Opo ors Tools Library Te Preference Windows Help

511111111101215111EILILILILILIIILLIIII

-1u 1 1 1°;1°'1 1 I I I I I 1 1Q101

®Q

EditionI

Included

PC Topics - - - - Ele kto r Electronics EXTRA 13 - 1/99PC TOPICS —————————————— Elektor Electronics EXTRA 13 - 1/99

Printing

Simulat 2.0 can provide a PrintPreview. This dialogue box allows youto select the printable area, the scale,page margins, contents of the ID box,and other printer configurations,

before the print job is started.The buttons in the dialogue box are‘repeat’ types, allowing the relevantcommand to be issued again bykeeping the cursor on the button rep-resenting the desired command.Compatibility is assured with any print-er already functioning in graphicsmode, and driven by other applica-tions running under Windows.To print an ID box with the schematic,click on the button ‘Cartouche’. A dia-logue box pops up which allows you todefine the text fields which are toappear in the ID box.Paper margins may be set in the dia-logue box ‘Page Setup’ (‘Mise enPage’). In the title bar the program dis-play the paper dimensions and theprintable area. These two values will ofcourse depend on the printer youhave connected up to your PC.

Help system

A powerful, context-sensitive help sys-tem is available.To get help, you click on the‘Contextual Help’ symbol in the buttonbar. Next, you click your way throughthe Help menu to get to the item youneed help on.

File management

The file management system allowsyou to stay organized as far as yourschematics are concerned.The system offers you a number of pos-

sibilities for copying, moving, loading,opening and deleting your schemat-ics. It also supports displaying fileproperties of disk drives, and freespace on the hard disk.

Miscellaneous matters

Because Simulat 2.0 runs under theWindows 95 operating system, you areable to copy complete screens intothe Clipboard, and from there movethem to other applications like MSWrite, etc., in order to create docu-ments (useful for educational/didacticpurposes).Simulat 2.0 supports the MultipleDocument (MD) system. The programallows you to work on four circuit dia-grams simultaneously. The names ofthe schematics are displayed in the‘Window’ menu (‘Fenetre’). You canchange between circuit diagrams inthree ways:- by left-clicking anywhere in the win-

dow of the schematic to be activated;- by clicking on the name of the

schematic as listed in the ‘Window’menu;- by selecting the ‘Next’ option fromthe System menu in a window.

Simulat 2.0 also supports the OLE 2.0system. The linking and embeddingfunctions allow Simulat 2.0 to acceptdifferent data types and objects pro-duced with other programs. The maindifference between linking an objectand embedding it concerns the wayin which data is stored: embeddedobjects form part of a schematic (i.e.,

they are integrated into the file), whilelinked objects remain stored in theoriginal (source) file. The schematicthen only keeps information relating tothe location of the object in the sourcefile.Circuit diagram files may be selectedand moved into the simulation win-dow. Using the file management sys-tem you may pick one or moreschematics and drop them in the sim-ulation window. Once the cursor is onthe program window, release themouse button, and the selected filesare loaded into the simulator. Notethat Simulat 2.0 can only handle fourschematics at a time.

System requirements

You will need to have Windows 95installed (or a later version). The PCshould have a 386 processor or better,4 Mbytes of RAM, a VGA card and amouse. Simulat 2.0 requires about6 Mbytes of free space on your harddisk.

Note: The version described here supportsa maximum of 100 components. Theauthor can supply a larger version.

(992014-1)

Tern perature Recorder ern ploys a DS1620 transducerand an AT89C1051 RISC m icrocontroller to recordand storing tern perature values. The microcontrollereffectively connects the transducer to the serial port(RS -232) on your PC.

By John Th . Ko kko ri

,tait czo'0

Sp

98.1r r e99

tern perature recorderFirst Prize, UK national entries

C8

V cc 0)

10.63

C9

PIRS -232

0 00 0 O

[C38

14O

15

cc 0 16

62

0

10/63

R2INRI INT2OUTTI OUTGNDVCCVs

VS,

R2OUTRI OUT

T21NTIEN

C2C2+C1C1+

0O

9

2

11

5 C4

MAX232

VCC

1000/16

10/63

10/63

1

Voc

IC1

5

8

9

Rst VccP3.0 P1.7P3.1 P1.6XTALI P1.5XTAL2 P1.4P3.2 P1.3P3.3 P1.2P3.4 P1.1P3.5 P1.0GND P3.7

-§-ovcc18171615

1413

12

Thermometer

3

1C2

11

X1 =T.-- 89C1051

H12M

-05 C6

T33n T33n

Figure 1. C irc u it d ia g ra m of the intelligent interface and its connectio n to the PC 's se ria I p o rt.

Thanks to an Atme189C1051 controllerin the purpose -designed interfaceshown in Figure 1, the system is capa-ble of performing temperature read-ings and storing measured valueswhile the PC is off. Also, the actualther-mom eter may be located at quitesome distance from the host PC. ThePC, when switched on, may requestthe temperature readings from theinterface, and use them to display atemperature graph. The interface is

built on a small printed circuit boardas shown in Figure 2.

Program descriptionTemperature Recorder is a

Win95/98/NT application written in

Visual Basic 5. This program allows youto select the COM Port to which thethermometer is connected, as well asdefine the value of parameter 'RATE'which is the interval between two tem-perature readings. This is defined bythe equation

11ME= ( RATE+ 1 )*30 [seconds]

For example, if RATE is 0, the sampling

DQ VDDCLK TH/RST TLGND TC

7

6

DS1620

0 Vcc

992013 - 11

interval is 30 seconds, and if RATE is 19the interval is 10 minutes.Using 'Read Buffer Data' you canprompt the microcontroller to transmitits buffer contents to the PC. Beforeyou start the transfer you should enter,in the 'Filename' field, the name of thefile which is to contain the temperaturereadings (see Figure 3). Alternatively,or you can choose an existing file byclicking on 'File' and then 'Select File...'. Having selected the file you maypress the 'Start' button, whereupon thediscrete temperature values will be

14 - 1/99 Elektor Electronics EXTRA Pa Topics

Thanks to an Atmel 89C1051 controllerin the purpose-designed interfaceshown in Figure 1, the system is capa-ble of performing temperature read-ings and storing measured valueswhile the PC is off. Also, the actual ther-mometer may be located at quitesome distance from the host PC. ThePC, when switched on, may requestthe temperature readings from theinterface, and use them to display atemperature graph. The interface isbuilt on a small printed circuit boardas shown in Figure 2.

Program descriptionTemperature Recorder is aWin95/98/NT application written inVisual Basic 5. This program allows youto select the COM Port to which thethermometer is connected, as well asdefine the value of parameter ‘RATE’which is the interval between two tem-perature readings. This is defined bythe equation

TIME = ( RATE + 1 )*30 [seconds]

For example, if RATE is 0, the sampling

interval is 30 seconds, and if RATE is 19the interval is 10 minutes.Using ‘Read Buffer Data’ you canprompt the microcontroller to transmitits buffer contents to the PC. Beforeyou start the transfer you should enter,in the ‘Filename’ field, the name of thefile which is to contain the temperaturereadings (see Figure 3). Alternatively,or you can choose an existing file byclicking on ‘File’ and then ‘Select File...’. Having selected the file you maypress the ‘Start’ button, whereupon thediscrete temperature values will be


Temperature Recorder employs a DS1620 transducerand an AT89C1051 RISC microcontroller to recordand storing temperature values. The microcontrollereffectively connects the transducer to the serial port(RS-232) on your PC.

By John Th. Kokkori

temperature recorderFirst Prize, UK national entries

Figure 1. Circuit diagram of the intelligent interface and its connection to the PC’s serial port.

softw

ar

e·

available·on

·cd-rom

PCsoftware'98-'99

written into the file. You will also see thetime, date and the value of the lasttemperature reading. If you press the`Show Graph' button the system pre-sents a graphic display of the mea-surement results.

Installing the program

To install the program you simply runthe 'Setup' file.If you have a different version of VisualBasic installed on your PC you mayhave a problem with the installation. Inthat case, do not stop the installation.Once Setup has finished, copy all files,except `Vbctrls.reg' from the disk'Tem p_Rec / PATCH', to the Windowssystem directory, then run the file

`Vb ctrls.reg' from the 'Temp_Rec /

PATCH' disk. This will update the rele-vant files in the system directory, andthe program should then work.

All files available!

The software as sup plied for the ElektorElectronics Software Design Contestincludes all the source codes for theTemp_Rec Program (in V.Basic 5), theasm, hex files for the microcontrollerand the PCB artwork and schematicfiles (Protel files and image files) for thethermometer.All software components as suppliedby the author may be found on a CD-ROM containing winning entries fromthe 1998 Software Design Contest. ThisCD-ROM will be available earlyJanuary 1999.

992013-1

Fig ure3. Sc re e nd u rn p showing

Tern p e ra tu re Recorder in action.

0

00OO0

00

C9 C310/63 0.1

00O426(3/6O0

IC3r1AX232

810/63

C410/63 O 0

0 0,

CI10/16

XI4M

0 0

0

0c c

OC 5U Lncl °c10(1 0(1

Figure 2. PCB artwork as supplied by the author.

000

00o0

0

0000OO -V-O o oo oo o 0o o IC2 o

DS1620

L1Temperature Recording

Current Temperature

± 0 0rCurrent Date?

1998/10/291s

Current Time

09:15:36

FileName:

- COM Ports

mr COM 1COM 2

C COM 3r COM 4

1 2400/9600

4.0 Read Buffer Dates.

Start _I

Stop I

1?_m_ate

Hide Graph I

ICAProgram Files\Temperature Recording\ThisOctober

emperature History

Pa Topics - Elektor Electronics EXTRA 15 - 1/99

written into the file. You will also see thetime, date and the value of the lasttemperature reading. If you press the‘Show Graph’ button the system pre-sents a graphic display of the mea-surement results.

Installing the program

To install the program you simply runthe ‘Setup’ file.If you have a different version of VisualBasic installed on your PC you mayhave a problem with the installation. Inthat case, do not stop the installation.Once Setup has finished, copy all files,except ‘Vbctrls.reg’ from the disk‘Temp_Rec / PATCH’, to the Windowssystem directory, then run the file‘Vbctrls.reg’ from the ‘Temp_Rec /PATCH’ disk. This will update the rele-vant files in the system directory, andthe program should then work.

All files available!

The software as supplied for the ElektorElectronics Software Design Contestincludes all the source codes for theTemp_Rec Program (in V.Basic 5), theasm, hex files for the microcontrollerand the PCB artwork and schematicfiles (Protel files and image files) for thethermometer.All software components as suppliedby the author may be found on a CD-ROM containing winning entries fromthe 1998 Software Design Contest. ThisCD-ROM will be available earlyJanuary 1999.

992013-1

PC TOPICS —————————————— Elektor Electronics EXTRA 15 - 1/99

Figure 2. PCB artwork as supplied by the author.

Figure 3. Screendump showingTemperature Recorder in action.

electronics on-line · designing electronic circuits almost invariably calls for extensive...

Documents