Scribd Logo
Upload

Welcome to Scribd!

  • Wireless World 1982 09

    Uploaded by

    jacomart
    75 views64 pages

    Original Title

    Wireless-World-1982-09

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    75 views64 pages

    Wireless World 1982 09

    Uploaded by

    jacomart

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 64
    (aie et UE ey a eS Cen oe we ~ ae world etd ard lg C.b. selective call unit TRANSMITTER TEST SET, TTS520 Tests transmitters up to 100 watts rating Der ed For testing base stations: mobile or fixed radios: pocket Ce es CU Eee Ce eee ns ea al Transmitter measurements iaciue: frequency » power » Ss ‘sensitivity, handwidth, capability « call tone modulation Cee ucd Ce Lae nd Creag CO Compatible with Farnell $S6520 synthesized signal generator Se ee eee ‘Split concept (receiver transmitter testing) offers distinct pe ee een cy Ce a OE ad Cee Lae CU radio as received and to verity performance to specification Seo Front covers laser gyro, used in ‘strap-down inerdal navigation ‘ystems, Picture supplied by ‘SFENA, Vilecoublay, France. NEXT MONTH Modular preamplifier. The first of three articles in which John Tinsley Hood describes his features several novel ideas. Electronic compass. Design by Neil Pollock for all-solid-state compass for small boats, using the sensor output or ahard- ‘wired system, which is described, can be used. Low-frequency oscillator. A sine-square oscillator, with tone-burst switching, which covers the 10H2-250KHz frequency range. Caran mye price 0, tack nas st Reta nd rage Coun Units 1&2 Bankside incustal wey Hopton “Set. London Set se ‘on microti: plesse contact House, The Quadrant. Suton, Surrey Advertsing offices: Quss PToaerstte Oscar Suton, Ratshonos: Eon 01.551 2500. Ad- ic 1301-661 3180, segrtmnTelx 652084 BISPRS G Sittin ree yon 14 UX ond Studant rate’ year £838 UK and Pe Poutede Uk Blsmaton:” Guscrant_ House, The aarar”"Suoy Srey SAS ‘oprone 8 Passe ntily a change o address UBA'$50 surtece mall, 898-30 sia US subscriptions, om JC BP, Sun Serioiona Ole, 205 & 2nd Sircet NV ‘corr USA mating agents: Expediters ofthe Printed Word U3, 527 Madison Avonve, Sorte s217, New Yor RY odes end SE Pe atte Prose Ln ie ss WIRELESS WORLD SEPTEMBER 1062 sccrenes WIHFEIESS | comrcsrors covewensee WON vatsenna 31 Land Miss. 32 DIGITAL FREQUENCY STABILIZATION OF A V.F.O. = hy W. Trapman 38 yas 40 25 DIGITAL PHASE CONTROL OF THYRISTORS by B, H. Pardoe 2G DIFFERENTIAL DIRECT CONVERSION by P. E. Gill __ 4G SIMPLE LOW-FREQUENCY OSCILLOSCOPE by R. Taylor 5] SELECTIVE CALLING UNIT | bv. Drury NEWS él 9] TRANSMITTING AERIALS fy J. Wiseman 64 - Chicago Tillett RCA 3140 AND nT L071 + iin et &. hin: a EPROM EMULATOR o: NEW PRODUCTS 90 RANDOM ECHOES The to all your power supply problems. As supplied to International Telecommunications Companies Uninterruptible Power 125va to 50kva, 50 or IB STANDBY POWER, Invaluable for winding down a ‘computer programme on mains failure and wherever ‘ontinuous power is essential IB STABILISATION. :-5% Vital to combat mains voltage fluctuations and ensure the operation of equipment at peak efficiency. Frequency stabilised +-0.1% 47 to65Hz I TRANSIENT ATTENUATION. Provides suppression of mains born interference (spikes). ‘An unbeatable power package at a price to sut all budgets irom Gaatrek ternational UPS ore DOWN to unbeatable plies - UPS CVT Reliability For more information, cut the coupon Galatrek International, FREEPOST, Scotland Street, Llanrwst, nr. Llandudno, Gwynedd LL26 OAL, BRITAIN, Tel No: 0492-64031 1/641298, Telex: 617114 AVB Galahu. Made and Designed in Britain by Galatrek. ALATREK «3 INTERNATIONAL ?_Y Pease send motu etal of your range of ologe Sapiser liters cutouts, generatorsand Ts PleasosendmetlldetaisofUPS Pease sendmearequrementcheckshest t Consultation with GalatrekEngineer o Nome = eee Positior Compatly Address. - es TelNo__ ee “Frade Ci OEM (1 (please ickwhereappropriate) wwe WW — 072 FOR FURTHER DETAILS aaecsal Borst sing with power since they fest ‘thing ne sures rarer ote uate yok sin ato ange of powers ampfrs fare fe oh Mee 2 owen Rane (Cortranas Ms) ‘onisow Naa nent pa rae Breet tom owaow A Ehen Reames un row-smw ——40,a0,160 8 Sonic exéraem ey fers RELIABLE LuneaR ~ tnt ton a tn ieee ga, — Seat psec Hiya (A SE amy & phe reauicemens The HY eet ct power tn cos custsang fs cen» poss. Ake ‘boards can be produced. ~ cveocus our ‘Aetbay tand hse i forte ‘WW ~ 099FOR FURTHER DETAILS. ERASERS {82 DEVELOPMENT LAB oe uviB ‘Asthove bt without sine, Caniage paid U.K) Send cheque oe oc order ‘prompt delivery INUSEIN DESIGN LABS & EDUCATIONAL, ESTABLISHMENTS AROUND THE WORLD Hag hac [NORTHERN ELECTRON! ‘WW ~ 040 FOR FURTHER DETAILS PAA VMK od Sra ereaagquveur rnc este" i} hs y COMPLETE sora ore rcs Win OATA ic ewe ‘01-723 1008/9) Les ex TOMO, .ONON ED oi nacuge erar in Tt Pactra eon 2 WIRELESS WORLO SEPTEMBER 1982 wireless world Editor PHILIP DARRINGTON 07-661.3128 Doputy Editor: GEOFF SHORTER, B.Sc 01-661 8639, ‘Technical Editor: MARTIN ECCLES 01-661 8638 News Editor: DAVID SCOBIE 01-861 8632 Design Editor KERR Drawing Office Man: ROGER OODMAN ‘Technical illustrator: BETTY PALMER, Advertisement Manager BOB NIBBS, A.C. 01-861 3130 DAVID DISLEY 01-861 8647 BARBARA MILLER 07-861 8640, Northern Sales: HARRY AIKEN, 061-872 8861 Midland Sales: BASIL McGOWAN 021-356 4838 Classified Manager: BRIAN DURRANT 01-661 3106 OPHELIA SMITH 01-661 3033, Production: BRIAN BANNISTER (Make-up and copy) 01-661 3600 X3561 LT. and M.LS.S. One of the aims of information ‘Technology Year and the Microelectronics Education Programme isto involve schoolchildrenin theuse of ‘njerocomputers and related electronic devices. There are the M.E.P., the Micros fn Schools Scheme, exhibitions and events throughout the year and beyond. Iti, perhaps, fortunate that Me Callaghan bappened tobe watching television onthe evening the programme “Now the Chips sre Down’ was broadcast and was spurred {nto action tha, or we would probably find the propaganda even more frenetic than that now being put out by the ‘energetic Mr Baker, the prophet of IT Tnformation Technology isa curiously diffuse name for a Year. The official definition, “the acquisition, processing, storage, dissemination and use of vocal, pictorial, textual and numerical Jnformation by a microelectronics-based combination of computing and telecommunications” appears to ‘encompass most ofthe activities of the average person, except eating and one or ‘wo other processes although the use of a ‘computer snot often considered essential to the more basic of these. So ar as its involvement of schoolchildren is concerned, the publicity is decidedly shill, che Minister's aim ‘ring to have a computer in every secondary school by the end ofthe year and even o think about providing them for primary schools, "There can be no argument that young people must be aware of computers and howto use them, but it does seem possible thatthe present blaze of publicity tends to obscure the point that computers area ‘means, not an end. There i also the duestion of how the micros are to be used in schools. ‘According tothe ith eition ofthe Concise Oxford Dictionary (now, ‘ximitedly, modified), a computers “a ‘alulator~ an elecvonic calculating ‘machine” — an unfortunate description, taken to literally by atleast some of those responsible fr introducing youngsters to computing, withthe result thatthe school ‘cro is offen given to the senior maths teacher to guard with his life, presumably fon the grounds that computers are lator drifts a litle, but presents no prob- Jem. The real problem is with the last dig of the counter, which in successive counts (fit were provided with a display) would show two values ~ it would ‘iter’. The stabilizing counter does the same thing, Fig. 4 Analogue weighting circuit doing the s9me work as many gates. Results stored in cross-coupled gates. Input resistors should have narrower tolerance when ‘more and more signals are to be weighted. Supply lines are well regulated. ‘Hoo Fig. Simple delay line using propagation delays in butfrs. 2 even when the incoming frequency is perfectly stable since there is no phase relationship between the reference oscilla- tor and the v.£.0. This is most disturbing ‘and leads to noise on the correcting voltage tw the oscillator. To get a clear picture I drew a few pulses ‘on graph paper, as in Fig. 3, which repre sents @ 5 Hz signal as it appears before ancl after the gate of a frequency counter. The flip-flops in the counter toggle on the posi- tive-going slope. When a window of exactly one second moves over the $ Hz pulse train, we see immediately what hap- pens: half a pulse has just the same posi- tivergoing slope, which is what is counted, halons 9 Boi count STITT when the two waveforms glide past each other. Only when the situation is exactiy stable (no gliding) do we get consistent answers from the counting process, Figure 3 (©) shows that stretching the pulses out (by lowering the frequency to 4.5 Hz) does ‘not produce a different result; even 4.1 Hz does not make the fifth pulse ‘fall out of the window’. ‘The only way to make this one-pulse ambiguity less important is to count more pulses, for instance by counting during ten seconds or by making ten, separate, one- second counts and averaging the answers. ‘The last method was used very often when counters had no switch position for —< = | r Integrator TokSovr Reset nsec. Systal canal Fig. 6. Block dlagram of equency-stablizedv.t.o. Each of odd number of counters gets a ‘slightly different ‘cut from continuous pulse simultaneously in from the oscillator. Counters work Frequency variation of output is reduced when corrections come fester, dri boing Peau hibedh coun Even otros times str corecon rates worthwhile improvement. (.. over pulse from tage oye ‘owes | | ters | Fig. 8. Adding digital information without clipping is much better. Digital store is trozen whan switchs open. ™ ‘multiple frequency’ and the resolution ‘was just one decimal place too low. Anew method All this may be interesting but does not look very promising for our v.f.0., until a ‘new idea is brought in. Ite not necessary to do the individual one-second mea- surements one by one: the measuring times may overlap. The aim is to average ‘out the one-pulse variation and obtain the hhigh resolution that may be expected from a long counting time, but before placing endless rows of decades on the breadboard it might well be asked why one should use four flip-flops to get a single 0 to 1 out of them? Well, it depends largely on the stability of the oscillator when the control- ling loop is disabled, and on the nature and magnitude of any impulsive noise that ‘might enter the system. With a simple four-bit counter there is plenty of margin on both sides of the dividing line between 7 and 8, but with a decade counter one side is limited; wo pulses extra make a zero from that 8 to give a ‘correction’ in the ‘wrong dirction, ‘Because the decade on the breadboard ‘worked well once the starting time was over, I decided that two-bit counters were all T needed. There were times I wished I hhad used the SN 7493 four-bit counter, but some of the problems turned out to be hardware-oriented and not so much a system fault. Voting logie ‘The majority-vote is taken by the analogue circuit of Fig. 4, which works exsily with 21 inputs ~ an odd number to get a ‘majority in all situations. The input resis- tors are of the carbon-film type, selected for close tolerances fora larger batch with the cireuit of Fig. 4 itself (without R,-R,). AA volumeter at point P indicates output voltage during the selection; the setpoint and sensitivity potentiometers are partly ‘incorporated for this occasion. Even this circutis« simplificaon from the bread- board circuit, where the digital outputs of ‘the counters were stored by “buffer flip- flops. These buffers are not required if a reasonable amount of time is available for the voltage at P to take on the proper value. If take over and reset must take place in 1 or 2 us, it Becomes necessary 9 tse the buffer store. Counting in the parallel, but delayed, counting operation, it is immate To verieap “T Fig. 8. Push button is pressed 10 seconds after power is switched on, reducing loop {ain while charging capacitor to average Value. H.. decoupling of control line not shown, WIRELESS WORLD SEPTEMBER 1982 Fig. 10. Block diagram of improved version. The ‘more counters give a too high’ signal, the greater the contro! voltage to the oscilator. Beware of ht ‘on control line, Transistors with 20 V supply amplify the digital signals to avoid varying levels int Y Ped Pr \ ute 9a bt counters i utter/ | oetay ‘ine vv aT ‘whether the clock is delayed or the v.f.0. signal. For oscilloscope testing, it is very pleasant to be able to delay the highest frequency (the v.f0.): it would take a sto- rage oscilloscope or one with dual timebase to display 10 Hz signals with small delays. The end of the counting, take over and reset and start can now be the same forall counters, which keeps maters simple. The last question to resolve is, with a5 MHz v.£.0., over what period must the delays be extended and how many counting systems tare needed? In 2 practical system, with cheap t-t1. building blocks, there are only ‘wo possibilities for building a delay lin. ‘The first one to think of is a series of ‘monostable’ delays, SN 74121 being the monostable block.” A second way to achieve delay is the use of the intrinsic propagation delay of logic elements, the standard inverter SN7404 giving, in pairs, delay of about 20 nanoseconds, according to the data sheer, although measurement shows the delay of n pairs often to be shorter than n times 20ns. To get a maxi- ‘mum delay of at least one period of the 5 ‘MHz, the delay line consists of a series string of inverters with outputs every to pairs, which takes six blocks, ene more ‘being used to make sure the input to the delay line has the appropriate switching Complete system A block diagram of a complete system gives Fig. 6. The veto. frequency of 5,123,456.789 Hz is counted during 0.1 ‘second in a modulo-four counter which is another way of saying that it contains, ‘when the counting is over, the remainder of the division-by-four of the number of input pulses. if the remainder is 0 or 1 20 level is presented to the analogue weighing network, a count of 2 of 3 producing a 1. ‘The nine counters get a 0.1 second gate signal, the closing of the gates immediately being followed by storing a 0 or 1 in the Aiptlop, depending on the dominance of the counts 0 and I (‘too low’ signal) or 2 \WiRELESS WORLD SEPTEMBER 1052 P and 3 (too high’ signal). The integration time constant is related to the repetition rae of the measurements, which ten pet ‘The stabilization points (the grid om the tuning dial) have such frequency valves that the average remainder in the counters wobbles between I and 2. The next grid point i one eyele of the counters further, Or four more counts. Four counts in one tenth of «second means 40 in a second, #0 the grid poins have 40 Tz spacing: the ++0/+1 irregularities are not so well aver- aged out a8 I had hoped. A lide calou- [tin and comparison with other processes where a random disturbance is reduced in ‘prominence by a form of averaging leads to the unpleasant conclusion that these vara tions ate reduced by ‘VN, where N i the ‘numberof eounters. This is disappointing and means that the nine counting systems five only three-fold reduction in random “This should be regarded as a by-product of the process, the main advantage being the faster response of the loop. In the system of Fig. 6, drift of the oscilator leads to all counters giving the same signal, 2g. “to high’, and this lads to a consis tent | in the buffer, which is something different from the random | appearing ‘cccasionally. Before examining this more closely, it is necessary to remember that the vifo. stabilizer is not perfect ~ not only i the numberof counters N very low, ‘but the delay-line sections do not have ‘qual delays. Then there isthe question of ‘the length of the delay line: one feels by {ntuition that che discrete length ofthe fine is important, end the same intuition tells tus that when N is realy high, this diserete- ess is not quite so important. ‘Another point to remember is the fact that the “feel” of the turning knob is 0 ‘much better, because of the improvement in response speed, compared with a one second gate period. The ripple on the fre- quency is probably as shown in Fig. 7 and even a 3x improvement looks good. “ake asec Gate tine LPL Further development At this stage of the experiments a set of flipflops was in place, and for a reason I cannot remember now these storage flip- flops were reinserted between the counters and the weighting network, Instead of the ¥.f.o, a crystal oscillator was connected to the input and a digitsl-to-analogue converter with a set of equal resistors feed- ing an indicator in the form of a $0 WA ameter (Fig. 8). The frequency ofthis oscil- lator can be varied slightly by varying the supply voltage so, by breaking the line controlling take-over of information a cer- tain times by means of the switch and changing the oscillator frequency by small ‘mounts, the effects ofa drifting oscillator can be seen on the meter. There is n0 closed loop and we see just the informa tion-gathering part ofthe process. ‘The demonstration was set up without thinking very much of it, but it became apparent that this ‘analogue’ method of adding the information from the counters is very good; the occasional +0/+1 varia- tion is much smaller than the swing of the pointer when # small frequency shift cecurs, several counters responding in the same direction It really come down tothe fact thatthe limiting applied by the 709 in Fig. 4to the sum signal should be avoided. What is needed is linear amplification, which can be done with the left-hand part of Fig. 4, the storage in the right-hand part now being redundant. The control line to the v.f.o, may not go negative, but litle 4.c, amplification is required. Some amplification can be_had_ before the d-to-a conversion takes. place by providing a higher voltage difference be- tween logic levels, The SN 7406 or discrete transistors may be used, connected be- tween the buffer flip-flops and the sum- ‘ming resistors. This was not tried, but seems an easy way to get some gain when a swell-regulated supply of 20 V or more is already available. This method gives well- dlfined voltage levels for ‘high’ and ‘low’, ‘but the system is very tolerant of faults in 8 aes sd Fig. 11. Delaying clock (with gates) is easy but does not lead to anything useful H aha H wo} ee Hoke ove it a ee LT Fig. 12. Command pulses are derived with monostable delays. Exact value of gate time. hotimportant and any odd crystal may be used. _ beacon cua aoe ensl na0is Tests asians Fig. 18. Further possibilty is number of erystal oscillators. E /\ ¢ ‘ ost) Joes the logic levels and even a distinct fault of: one of thet... blocks feeding the resistors ‘must be more than 0.3 volt to become a nuisance. (I know this very well a a user of out-of specification ics.) Starting-up ‘There is a small problem to be solved in ‘connexion with the starting conditions: the integrating capacitor should be charged t0 the average value on the line and the loop disconnected when starting up. The large drift of the oscillator in the fret minute or so after the power is switched on is often not due to the oscillator itself, the tran- sistor taking only a few milliwatts, but 10 setting down of the regulated” power supply, where more power is dissipated, the threshold voltages being temperature dependent. ‘During the experiments, a push button ‘was used, as shown in Fig. 9, its contacts ‘connecting the capecitor tothe sider of the 1 KO preset potentiometer across. the ‘power supply. The low impedance of the potentiometer reduces the loop gain when ‘the button is pressed. When the voltage on the slider is set ro the average value of the ‘voltage on the capacitor the loop gain can ‘be adjusted with the impedance across the ‘burton without undue frequency shift. Further refinements like suppressing the ‘output of the v0, during warm-up are possible, Delayed clock ‘The combination of Figs. 6 and 8 (shown in Fig 10) work well. There is still the feeling left, however, that the subject of the parallel working counters should be looked upon from another point of view. This is stimulated by a remark made by ‘one of my friends that the number of pack= ages is not very high, afterall. Besides, a vf. with several frequency bands and a Tong dial is always a fairly large part of any telecommunications apparatus, and who ‘would object tothe number of .cs used ifa better performance is obtained? So 1 started to trace the decisions made (and reported here) and found a badly argued ‘one in the preference for a delay line con- nected to the v.f.0. signal. ‘Knowing now it can be done that way, I wondered if T could delay the clock with ‘equally good results. Schemes using a long delay, like the one with the timing diagram of Fig. 11, all seem to use more time with- fout any improvement in performance, compared to that of Fig. 10. The idea behind such schemes is that whem the first gate closes, some information about the lunknown frequency is available, this in- formation getting improved every time a agate closes, Here is still an opportunity for thought and experiment as I can see n0 practical way to achieve something useful along these lines. Clock oscillator ‘The clock used in the experiments consists of e crystal oseilaror, buffer amplifier and pulse shaper, followed by a chain of coun- ters. When experimenting with longer or shorter gate times, complicated switching WIRELESS WORLD SEPTEMBER 1982 to get the command pulses at the right times was avoided by taking the exact re- petition frequency used from the divider in question (¢-g. 0.1 Hz, 0.2 Hz, 0.4 Hz, ctc.), and triggering a 74121 monostable block with it. The time between monosta- ble pulses is used as the gate time. A few ‘more monostables produce the command pulses ‘take over" and ‘reset’ as in Fig. 12. ‘The gate time is now not exactly 0.1 second, etc., but this is unimportant. The stability ofthe gate time is still good, as the iis from the 74121 is only a small part of it and a drift to 1.1 ysis unlikely to happen ina short time. Further possi Next to waiting for the proper integrated circuit to appear on the market, containing a large number of fast gates and counters, the most straightforward approach would be to use a number of nominally equal crystals, as in the scheme of Fig. 13. This is especially attractive in countries where ‘ex-second world war supplies are available ‘and cheap. Of course, very efficient decou- pling and buffering is necessary to prevent the oscillators from locking on to each other. The v.c.0. frequency may be chosen freely, taking into account the division ratio N ofthe divider chain, ‘This should preferably be a power of 2, providing symmetrical output waveforms from all intermediate tap points. The ‘number of flip flops in the divider should be such that the output frequency of the divider is, say, 10 10 20 Hz, As outlined before, there is no need for separate flip flops controlling the gates, this being done just as easily with monostable circuits. ‘The use of 27 MHz crystals was not successful, these being probably less stable ‘than the surplus types, which were around 3 MHz. The lower oscillator frequency ‘makes the isolating problem less severe, too. It is not necessary to make the fre- quencies of the crystal oscillators very nearly equal when the weighing network is large, as itis here. Starting from a nominal v.coo, frequency (e.g., the one that gives exactly 0.1s gate time) it is easy to find out what is the maximum allowable spread in ‘crystal frequencies, the limits being given by the size of the counter. Obviously, one requirement is that all counters overflow the same number of times. as can be ex pected, a better alignment of the crystal oscillators makes it possible to reduce the ‘numberof resistors inthe network ‘Reducing the size of the counter from ‘modulo 16 or modulo 10 to modulo 4 or ‘even modulo 2 is worth thinking over, if the v.c.0. is extremely good. Starting up sand locking is going to be much more ‘troublesome; this makes other components necessary (automatic loop gain reduction when turning the v.c.o. wheel). With ‘modulo 2 counting no working system is ‘obtained, however, because in theory no frequency difference is allowed that leads toa zero in some counters and a one in the ‘others. In this case there is no discrimina- tion between a higher and a lower ¥.c. frequency. For some frequencies, or grid points this is not true, a form of stabiliza- tion is taking place near certain points of the dial ~ more a drift reduction than true stabilization. eA Computing ‘Apple interfacing, by J. G. Titus, D. G. Lar sen and C. A. Titus. 206 pp., paperback Prentice-Hall International, £7.85. (SBN 0- 672-2186, in theory and practice by the use of experiments. A general-purpose. bread- board assists in circult construction and there is enough information to enable the ‘experimenter to build a printed-board type. Only nodding acquaintanceship with digital electronics ig assumed. ‘The Logic Design Of Computers, by M. P. Chinitz. 413 pp., paperback. Prentice-Hall International, €11.15, ISBN 0-000-218-008. Neither hardware nor software are the pro- vines of this book, but rather the logical [processes which are performed. by former when instructed by the latter. A short historial section and a piece “on fe chapters on theory. Logie. olements are treated. in ‘outline and the final chapter doals with ‘microprogramming, ‘The 6809 Companion, by M. James. 88 pp. paperback. Bernard Babani, £1.85. SBN 0- 35934-0775, {A discussion of the characteristics of the 8809 microprocessor from the pro grammer’s point of view. This is a small Book, but nonetheless a very useful contr: bution. A slight drawback i that the text was produced by a Centronics printer, Which is good for its purpose but not, ‘16-bit Microprocessor User's Manual. 231 pp, paperback. Prentice-Hall International, £271.20. ISBN 0-13-566695-3. This is the third ection of what is effect vely Motorola's handbook forthe MC5B000 WIRELESS WORLD SEPTEMBER 1862 f microprocessors. Full information pects of instruction, signal and bus Operation end memory management re provided, with all characteristics. ‘The Explorer's Guide to the ZXB1, by M. Lord. 120 pp., paperback. Timedata, £4.95. In the words of the author, this “attempts to carry the reader on from where the ZX 81 manual leaves off”. There is no intro: duction to ‘ming ids which help to use the limited Fam. available’ more economically. A number of games are included as ilustra- tions of techniques. There is a cha machine languag help the inexperienced, and the final sec- tion deals with practical matters, such a loading from cassettes and connecting monitor. Timedata Ltd, 57 Swallowdale, Essex S815 582. Analogue design ‘Analog instrumentation Fundamentals, by VIF. Leonord, jr. 18 po. paperback. Pron all International, £1295: ISBN 0-672. 718366 “Anyone looking for an explan principles of 8. and de. polntertype ‘Svumens, bridges, transducers, RC fiters land attonustors will probebiy’ find this Book useful. Those needing practical help vllnot find ithere since, a8 to tie clearly Sates these are fundementals, which ‘tractvely treated ina readable svi ion of the ications and Design with Analog Cir ceuite, by J. M. Jacob. 498 pp, hardback Prentice-Hall International, £22.45. ISBN 0- 8359.0248-5, Essentially practical design information on the use of operational amplifiers in most of free students, the treatment is such that Tees advanced readers will find it accessi- Ql Operational Amplifiers and Linear inte- ‘grated Circuits, by R. F. Coughlin and F. F. Driscoll. 376 pp., hardback. Prentice-Hall International, £16.45, (SBN 0-13-637765-8, Writen at @ slightly less advanced level than the previous book, this Is even more practical and a litie wider in scope. It is intended for @ one year course, all the ci cuits described having been tested, Stu- dents not specializing in electronics will fing the book helpful and the practial ap- proach is ideal for amateurs, Radio frequencies Introduction to Radio Frequency Design, by WH Hayward, 363 pp hardback, Pre {eral Intemational, £2096. SBN 0.3. oat ‘way into lactronies by way of digital circuitry and ‘components, rf. design being well in the background: Thisis an attempt to do for rf what thousands of texte at this level do for digital design. It is practical, with many examples, mathematics up to about A level boing used. A working knowledge of ‘simple circuit theory is needed to make full tse of the book. All the components of @ modern receiver are described In detail, including a chapter on transmission lines and the Smith chart HF. Antennas for All Locations, by L. A Moxon, GBXN. 260 pp,, hardback, RSG1. £5.00. (£6.67 by post, from RSGB, 35 Doughty Street, London WCTN 2AE. ISBN (0900612-57-6 This is @ remarkably helpful book, which deals with a subject ofton regarded as be- ing akin to black magic, Ina throrougly practical manner. That is not to say thet theory is neglected ~ itis not — but itis kept in its place. The author has been well served by his publishors, who i out a well-produced book at an extremely Tow price ~ well under half the price of ‘many technical paperbacks of half the size, ” CLASS ‘S’ A novel approach to amplifier distortion If the load presented to the voltage amplifier by the power output stage appears very high, crossover distortion can be reduced and the voltage amplifier is mucl ‘A basic cause of distortion in power ampli fiers is quite simply that they have to drive loads, something which is clearly seen in Fig. 1 (b), when the higher resistance load ‘of 2.40) can have +25 volts of swing gener- ated across it whilst the low resistance load ‘of 0.80 can have only +10 volts across it before i limits. Another way of looking at the situation is to try and generate £25 volts across the 0.80 resistor, when severe distortion in the form of limiting will ‘occur. In other words, it is much easier to drive a high-resistance load than a low-re~ ‘The object of Class ‘Sis to perform the task of making a low-resistance or im- ppedance load appear to the voltage ampli- fier as a high, ideally infinite impedance. Basic configuration Let V, (Fig. 2) represent the voltage out- put of the voltage amplifier and Iy- the distortion yuo aaa Fig. 2 the current from the current ‘amplifier equals that into the load Fi, tho Voltage amplifier sees an infinite Impedance. Voltage Vy controls the current amplifier. by A.M. Sandman, M_-Phil.(London),M.I-E.R.E. ‘output of the subsidiary current amplifier ‘which also provides load power. Ry is @ small resistor which develope a voltage Vix proportional to the load current Ip.Rt is the load resistance If Vo is used to make the subsidiary amplifier develop a current I, as close in value to I, as possible then Iy—I,)=0 and Ziq wil tend to infinity “This is the simple idea behind a whole family of circuits, one of which will now be described Basic circuit ‘This consists, as seen in Fig. 3, ofa voltage amplifier, comprising » standard opera- tional amplifier with feedback (Ay,Ry,Rs) easier to design. and a voltage-controlled, high-impedance- output current amplifier (A> with its asso- ‘iated bridge of Ray Rays Rg, Re). ‘As the non-inverting (+) and inverting (-) inputs to A, are virtually atthe same potential, irrespective of what this poten- tial is (@ basic property of operational am- pilifers), it follows thatthe bridge must be in balance and that, ideally, since Ip = I. ‘Ay sees an infinite load. Demonstration circuit ‘The circuit of Fig. 4 works on the prin- ciple of the basic circuit, with the addition of a complementary push-pull pair at the output (Fry, Tr,) to provide more current than the second 741 and its voltage fol- lower Tr isa capable of. During the crossover region time when, say, Try stops conducting and Tz has not yet started to conduct, ‘Trs provides the ‘output voltage and the impedance it sees the basic volage-amplifer stage, 2 high-resistance load provides a larger swing Fig. 3. Basie circuit of Class A amplifier. Az iscurrent amplifier. ‘The author ‘Mr Sandman was born in 1933 of Jow- Ish parents and almost become a tlior! for gaining City and Guilds Full Technological Cortificte in 1967, he ob- tained his M.Phil. degree from London University in 1978. Many years ago, he published 2 ‘scheme for automating the railways 25 art of a scheme for road trafic euto- ‘mation ~ he tells us that his chief plea- | ure lies inthe invention of now | ‘Schemes such as a bandwidth compres: ‘sion system, currently in the pipeline, and “Error Take Off,” @ method of ro- ducing amplifier distortion published in Wireless World in October 1974, ‘A deeply held, though hardly nove, belief ie that we shall nover resolve our problems as a country until, we pay Our fengineors about twice as much 25 we donow. WIRELESS WORLO SEPTEMBER 1982 drops from, ideally, infinity to 1220. However, this is only fora small excursion voltage and current and so Tes can easly handle it. What is more important is that ‘when the transition is completed, the im- pedance seen switches from low to high impedance, which will produce a voltage spike a the output. However, this spike is of very low am- plitude due to the low ousput impedance of the first 741 and its voltage follower and, ‘more to the point, at least one other class “circuit, which will not be described at this time, exists in which this minor prob- Jem does not occur to-a measurable degree. Resistor Ry may be taken tothe junction of the 220, 1000 and 6.240 resistors, so reducing the output impedance to zero. ‘The crossover distortion of Az, Tr, Tr} (WI) has a very small effect on the output (W2) and this effect (W3) is less than 6mV (pike voltage) at the virtual eaxth of Ay corresponding to 2mV at the output (allowing for attentuation by the two 10k0 resistors). That is I2mV in 24V p-p, or 0.05%. Circuit analysis Referring to Fig. 3, Yw'=Val rap) Ro where BERG? @ 1 whence Viq'=Vay 1 T ([t-23] ) a wd nef Br ReeRS” ‘RELESS WORLD SEPTEMBER 1982 Fig. 4. Addition of power stage to augment Az output current forms working cireut. Let Ro=Rg' and Ru=Ry’ RutRo, @ from (3), and applying Binomial Theorem. Ane RyRo'=Ru’Re & Rut. from (2), Rut Ri Rue’? Rw’Ro Roe whence Ria 1 therefore Ane Ro=Re(I+d) Ry=Rw'(l+p) whence, from (1) and (2), and thus, by binomial approximation, Rut Ri, Roe A—p 2 O1 (1% resistors), (Ry RLV0.02=50 (Rye+ Ry), so that for Ry + Rr=8+1=90, the load appears as 4500 to Ay. The formulae (4) (6) and (6) are useful as initial design ‘guides, but the full formula (2) should be used for the final calculations. The circuit is a stable one and has the ‘reat advantage of dealing with the prob- Jem of crossover distortion whilst not needing any setting-up, even though a class B amplifier is employed in the sub- Fig. 5. Wavoforms in circuit of Fig. 4. Top shows crossover distortion at output of ‘river stage, middle picture showing that output of eurrent amplifiers clear Bottom trace shows small spike at input of voltage ampli caused by both ourput transistors Coming back into conduction after being cut off during crossover. sidiary amplifier. It is also insensitive to the effects of temperature, specimen and ageing variations on cross-over perform: ance, unlike the standard class B amplifier. TOUCH-SENSITIVE V.D.U. read with interest News of the Month, Wireless World, p.69, July 1982, to be informed that ‘Tony Booth and his enginering staff of Bish Telecom, London, have ‘invented the systen? called City Business System, described a5 ‘du, terminal with 2 touch-sensitive screen, the senitivity being obtained using a matrix of infra-ed beams which can sense which part of the sreen i being touched. ‘Worthy of note a this may be, [fel I should ‘bing to your attention that an identical sytem to that described was designed and. bult Gaveated?) by myself as part of a computer ‘controlled system on a lage research particle {eceeraor called Nimrod back in 1971, and was {in uaeuatl dhe machine shut down in 1978. ‘A report containing sufficient information to ‘make such # device was avaiable to anyone, price 20p., from HMSO or through any good bookseller ented: "A Transparent Touch-Screen Device for In- teractive Computer Graphic Displays RHEL IR248 1972. ‘Sorry sll been done before. EM. Mott ‘Rutherford and Appleton Dideot Laboratories Oxfordshire LAWS OF MOTION Mr Frost (uly ‘Leners) should consider the difference berween science and poetic imagi- ration. The views of Lilretius ‘were unsup. ‘ported by experiment, not open to test by others, (Chore being in any case no instruments about at the time to test them with) and, as matters of ‘opinion, have in fat about the Same scientific statusas the views of lying saucer nuts; perhaps less, If Mr Frost wil read. Aristotle he wil rapidly realise that sloppy theorizing is sloppy even when 2,000 years old, and that good ‘lence good even when five hundred years ‘older stil T suggest tha the briefeit survey of human history shows up the truth that an idea may be sid ¢o have arived when the engineers stop ‘sing rule of thumb ‘and apply. mathematics ‘Cannon had been around for hundreds of years before Newton, demonstrating all thrce of his laws with classical elegance. ‘The ballistics ex pert had i all ia font of them but missed ‘Nevwon didn't, Newton was a genius. Even 50, the needed an intellectual climate in which to function; Gallo, perhaps the greater scientist, jst missed it « Newton's great work i called the Mathemat- ‘al Principles of Natural Philosophy, and that is the indicator of his achievement; he synthesized the Known physical world in mathematica terms. Te may be that the central problem of ‘modern pays is either that 1) we are thinking bout everything all wrong, or 2) the right ‘maths has not yet come along, or 3) the ight ‘maths is avalable but lying about unnoticed, For Mr RG. Young, however, the rt has already sec in, He doubts Special Relativity be- ‘ause he thinks of electrons a8 peat in. drain- Fig unig wood at, Pes ee’ hard, M.D. Bacon ‘New Milton Hans DESIGN EXHIBITION have been asked to help organize an exhibition of ‘Design Since 1945, and as one of the ex: hibits, T require a 1948 Ekeo P63 “Princes” personal portable valve radio in green plastic fee. I wonder if any of your readers could perhaps help met locate one ofthese radios LEAKY FEEDER RADIO SYSTEMS ave read with interest the recent articles on underground radio systems by my former col- league, Dt David Mari. ‘Whilst I am in general agreement with most of his paper, I fel thar it would be desirable to ‘larfy the positon of the National Coal Boar. "There are about 130 leaky feeder radio systems installed underground, the majority of ‘which aze of the disy-chin type. This tech- nique, which requires separated transmitter and receiver, and which makes spurs difficult 10 {achieve io now considered obsolescent. Over the pest few yeas, experimental installations have been made of 2 numberof alternatives, includ ing the BDR system (one installation) the back if system (one installation) and the forward: drive system (vo installations). The two-way ‘epeater is, however, now being installed rou {Uney, and thee are cusrently 13 installations of {his type operating successfully underground, the largest being at Easington Colliery with & total length of about 10 km. The repestere used are more expensive than comparable unidiree tional repeaters, but the difference does ot significantly effec total system cost; and their performance is entirely adequate even in terms ‘Sfeumulatve nose and intermodulation distor tion for any NCB requirement expected over the next few years In addition, the technique is conceptually simple and needs only ewe types of Tine module; anormal repeater, and a spliting repeater at each branch pint. Tn the medium term (ay $ 1010 years) un- derground mobile communications syrtems may ‘eed to provide additonal services such a8 high speed dat or vision signals. Such requirements would involve considerable changes to our eut- eat systems, which flly meet present needs. It {sour intention to embark on further de- velopment as fature needs become clearer R. Js Parsons, lead of Communication Applications Group, ‘National Coal Board 3D TELEVISION ‘The publication of your article on 3D ty, fol lowing the ‘Real World” programme on TVS on ‘May 4th prompts me to write and reiterate that ‘this county could have had 3D ev 30 years ago, without the ned for ealou spectaces. "You published a leer of mine on the subject. some three years ago when mentioned that my father (he late Granville Bradshaw who, though designing in many fclds, was better Known in that of the internal-combustion engine, ‘zamples of which ae to be seen inthe Science ‘nd other museums had developed inthe 1930s ‘system of 3D photography, which was widely ‘sed in advertising displays. ‘With this system in mind I designed for hima system for television. Both the BBC and the TTA were extremely interested but had to admit In the end that they had no bret proceed with 3D, let alone the money with which to experi ‘The result — another Britis invention stifled at birth for being, so it would now appear, at least 30 years ahead ofits time. Geotrey Bradshaw ‘Leatherhead Surrey INTENTIONAL LOGIC SYMBOLS Referring to the latent Later July '82), why sp bickering about old BS symbuls, now de- functor otherwise unknown this side of the (Channel? There exists set of ANSUTEEEAEC approved logic symbols, that provides ‘roughout diveriction, ease of drawing (30, iris, po Bows) and inenionaligy one Wes Plewe refer oe. the 1981 Supplement to The TTL Data Book by Teas Istuments God eidon, pp 318343. ‘Ofcourse, « manuacure’s handbook, all of te fanaa provided should be fully doc- ‘ented, so the schemas loka bi weird. But Procice shows that alte pruning bee and {hore yields draings cat are very intentional Indeed and cay wre Ing, J Byckinans| Steruiden, Begum S-100 BUSES ‘Do you know if anyone hes made a comparison chart between the various S100 computers in {erms of the bus definition? Some of the early "S100" equipment differs with that designed after the specification was finalised. Can your readers lp? S.B. Hayes, Toltec Computer, Cambridge WIRELESS WORLD SEPTEMBER 1982 IT’LL DO — OR WILL IT? {am involved in various aspects of electronics including design, application, rouble-shooting, repair and construction. Like many people who have a general knowledge ofthe subject, T have {an appreciation ofthe principles ofeach aspect {nda respect for experts in each field. Some of iy spare time is given to private repair work, 1, calculators, eleevonic musical instruments, ‘EA, et, a0 that I consider myself not inexpe- enced in the area of consumer elecronis Recently I completed two jobs that provided teal food fr thought "The first was a Goodman's ‘One-Ten’ runer amplifier, which sports a very elaborate power fmplifier with several transistor sages in dc Bethe ‘The peblen ws simply traced to the Spier powcr evitchy which bed spy born She mcg es ey povwecuipt sage renga) about ump. of cure, bu ths an sited by ‘Salina own not up theo. These exemplify the Gree Bel do” ace that has de Japan Wha iso ‘Whe yar of watering are ery not em. Ihe for sch prods oi ite mere care ould vest aches a dependable the ave tdi ofthe "Or that fst ey and ory of te remarable development ht hat ve sto the fs dha repo nto products of nme dsion Mike Feeney Neth Gsforch ‘Neweanieupon-Tyne Presumably Mr Feeney is aware that depending ‘upon environment, posible misuse of equipe ‘ment and any number of other variable factors, {tis powible for oxidation to occur in siilar simuatons which bave noting to do with “ama- teutish design”. This is the only such ease that ‘weave been made aware of on this product and ‘many thousands are sli operation, providing ‘excelent performances to thei very satisfied T would farther point out thatthe last One ‘Ten tunecfampliier was manufactured. some ght years ago, TVE.W. Bennett Goodsnans Loudspeakers Lid he lamp with he it tht Mi the Bich omeiner ecco ndiny gee erally acknowledged ax baving been conser Sbly iftenced by lack of atenton to qualiy ‘matters. By the same token I believe it aso ‘widely ecktowledged that Ferguson mee one of the at a vead the sigs andy Iter with thee ‘Queen's -Avard for ‘Technolog! Achieve ment were one ofthe more sues in > tering the wend. AC least we are sll hte to respond titer WIRELESS WORLD SEPTEMBER 1962 Having sid this, I do no feel that Mr Feeney ‘lusrates his case well The Goodmans unit. hich he seers was engineered by one of the Countey’s leading hi-fi designers of the period, the late Frank Rodwell, no less than 12 years ago. Perhaps more significantly the last ones were actually made about 8 years ago. If cor- Foded fase holders were a general faut on these ‘models, then it would worry me more, bur they fare not, and Tam sue that there must be many ‘easons apart from intrinsic bad. quality why Frolated cases of contamination might emerge ‘over soong a peiod. "The ‘Revolver’ falis into an entirely different category. This is an o-e.m. product sourced for reverse, and as it happens power and switch areas alvays get particular scrutiny for safety ‘easons apart from anything else. Tis quite imposible to jodge the rating of a sovteh merely by looking at itt Approval speci fications run to many ages, and tests ake hundreds of hours to conduct aver tens of thou sands of cycles of operation. A switch banding ‘a amp. oF so on low volage in a small audio product would very likely be lighter in appeat- nce thin & television mains component which ould be ated a over 100A surges butt should note condemned fort This failure may wel be a symprom of Far ener manufacturers having chinks in thet D.R. Topping Director of Engineering ‘Thorn EMI Ferguson Lid, THEORIES AND MIRACLES ‘My initial reaction to the article by De Murray in the June ese of Wires World was “Why all the fos All scenic theories canbe shown 19 erate anomalies if pushed fr enough, indeed fe would be suspicious of one that didnot “Anomalies are not the catastrophe that Dr Mur- fay seems 0 think, they serve only to indicate the limits of present theories and are therefore the starting point for the next development. Dr ‘Murray also atacks ‘pragmatism’ in tiene; a5 shall show below itis the only measure, of ‘which we canbe certain, that ean be applied toa ‘lente theory. Dr Murray scems ro believe that there is an sbsolute truth or ultimate scientific theory 10> ‘wards which we should be moving, that we can ‘measure our theories against the rel world and so improve and up-grade them, that one day we Will achieve this abolute truth, In this he Ts sadly mistaken, for two reasons. Firsly in {esting our curren theories we must observe the ‘world, and in doing so we must also accept as tue other theories that explain the operation of cour observing apparatus (including the pbysical ‘bases of our own sense organs), and those theo- ‘es may in turn depend on the truth of other theories, including in some cases, the theory we ae testing. Secondly (and much more funda- rmentaly) not only must we observe via our senses (and other apparatus) but we must also interpret what we observe, and these interpreta Sons are, at base, limited by our human psyeho- logy, we are, as it were, bounded by our hhumaness. We can never observe the word ireedy, but only a human version oft, ‘There i nothing We ean do about this imi- tation, no advance in our observing apparatus an affect it. Any interpretation we cae vo put ‘on our observations, no matter how obtained, ‘aust be human centred, bot there wil always be the possibilty that the same observations ould be interpreted by a non-human thinking nity in another but equally valid way. Even it Wwe extend our mental capacities by building ‘ficial or machine inteligences that can in- terpret observations ofthe world in other ways, they could ot explain what they perceived directly tow without translating tito human ‘Dr Murray may have, if he wishes, his belief in the simple truths of nature, but as 1 have shown above, they are (and must always be) limited and based truths John W. T. Smith Hatield Hens RF RADIATION HAZARDS {read with interest the wo ates referring to 1.f radiation hazards; the Gist, WW, October 19aL, page 42 and the more recent article WW, ‘Apa 1982, page 8 "There are, in Australia, many hundreds of smal craft fied with microwave radar. The majority would be professional fishing boas. However, privately owned pleaure craft are ‘being equipped in increasing numbers. Many of the installations that T have seen have been installed withthe rotating antenna barely above the head of the helmsman, This applies in part cular to those vessels which have’ fly bridge shove the main sape-stuctue. T would be iterested to know if any of your readers have observed, or have knowledge of, ‘ny adverse effets resulting from exposure to ‘microwave radiation in the situation outlined above, John Allan, VKSUL, ‘Ovingham South Australia NIKOLA TESLA [Nikola Tesla is a fascinating subject for study snd I join with Desmond Thackeray (May "Let- ters) in the toart “Tesla Lives!™ But surely the centennial for Tesla was in 1956 and wasnt the ‘Tesla Society formed forthe celebration? ‘AS for Tela's claim to having invented wire- Jess telegraphy, his patent 645,576 was upheld by the US Supreme Court agaist the claims of the Marconi Wireless ‘Telegraph Company of America on June 2ist 1943. Ironically, Tesla had ied in January of tha year. ‘A factual account of what Tesla accom- plished, particulary towards the wireless trans. fission of power would make most intererting reading; may we hope for an article along these lines? Martin W, Berner 9Y4TAM, ‘Teinidad, Wa. “ COST EFFECTIVE IGNITION Far from being criticised (Letes, June), Mr Cooper is 10 be on producing a design for electronic ignition using component thar ean be obtained bythe home constructor. ‘Wireless Word has never claimed tobe a mag: azine merely catering for the amateur constr: tor and inthe course of demonstration the lest, technology his, quite rightly, published circuits ‘using components not generally avaiable, and ‘requiring alignment equipment found in’ only ‘the peviegedlaboretores ‘However, certain designs are sill leary in- ‘tended tobe but even though articles with the constructional detail of, for example, the Dinsdale high-quality audio amplifier of JanunryiFebriary 1965 are no longer published, Designers should therefore consider catfully ‘whether any_component they are fortunate ‘nough to have avaiable may not be avaliable to ‘themajrity of readers. ‘Before quoting past numbers, Mr Pickavance should conser the atid of many firms, and ‘ethape also his, to request from private indi- iduals for the’ lest-common components in small quantitis. €.G. Gardiner Chelmsford Exeex have read Me Cooper's article on electronic fgiton in March Wales World and wonder iT ‘might be permitted summarize my own ex- periences ‘and. views on alternative ‘ignition systems? Ifthe cause of bad starting isnot in the igni- tion aystem but, say, in the fel supply, as it ‘often, no alternative form of ignition sytem Will help. Unless you want to beable 1 start {your ca with a neatly Nat barter, Ido no think ‘changing the ignition system will help. Ii not instarcing the ear thatthe fangs ofthe Ketter- ing system have most effect. "As Mr Cooper points ot, if your engine only bas 2 6,000-6500 rpm. maximum (as most have) thee is nothing the Kewering system can'tachieve if kept in god order. have used many differen systems for many years but have never realy noticed any signif- ‘andy better fuel consumption. Ihave not made ‘my detailed esta but over say a tank fll, no ‘change is noticeable, I would lve to think that Deter consumption resulted ~ itis a tangible and worthwhile benefit in these days of high fel cots ~ but Thave grave doubts wheter it exits In their freedom from maintenance, lters- tive systems can, I think, produce benefits. The Fiat atk is relieve the contact breaker of the high current demanded by the Ketering system. Contrary 10 Mr Cooper's opinion, T believe this alone is worthwhile. Ifthe system chosen also obviates adjusumeat of the dwell langle, #0 much the beter. This leaves any resi- ‘dual mechanical weaknesses of the ordinary fontact breaker. These do not manifest thea Selves under say 7,000 rpm. ~ probably more save for wear. The contacts may not wea but the fibre block on the eam can and does and his wear is not compensated for by contact meat ‘The contact gup may not be important i alter snave systems but there must be one! Do altr- native systems affect the plugs? Those which produce more energy in the spark (ie. C.D. ystems) do but aly in the seae thatthe plugs a ‘will goon working long efter they would misfire with the Kettering system. Ihave never noticed ‘any reduced wear and it would be interesting if ‘Mr Cooper ‘could elaborate on the short paragraph in ie article on this aspect (what ian tlectronic spark?) Ihave noticed that wide gaps {up t0 60-70 thou) stil work and so do sooted plugs = due, Thave always assumed, (0 the sreater energy available with C-D. Incidentally, fas anyone noticed smoother’ running at ‘minimum throtde openings just to maintain speed, with wide plug gaps? ‘On the subject of eurrent consumption, doce in really mater (within limits) bow much cur ‘ent the ignition system uses unless you are in ‘the habit of leaving it on without the engine running? CEH. Benson ‘Market Drayton ‘Shropehire bad to abandon any idea of constructing Wa SKinson's Opto-lectronic contact breaker (WW, April 1981) simply because T was unable 10 ‘obvain the 1c, specified through normal real ‘channels, In the ight of Red Cooper’ remarks bout using readily avaluble components con tained in your leter in the June fue, T p- ‘proached the contruction of his ignition system ‘with more hope. Als, T have come up against the same problem of supply in respect of the ‘Mallard FX 3720 ferrite power transformer. (Of two main stockists tried, Philips and Mar- shall, the former would only supply for their ‘own equipments, andthe later, whom T rang, said they didnot tock the item! So far as ean See none ofthe usual suppliers who advertise in WW and elsewhere stocks the FX 3720, and I a beginning to suspect that i is again one of ‘hose desirable components avaliable only to industrial users and electronic engineers ‘through privileged channels of supply. ‘As to the circuit itelfy this ems 10 have sullered somewhat in the reproduction process for publication and I would appreciate your help inmy difficulties of intepretation. You will have noticed that Dy, unlike Dy. has not been Specified. This may bea misprine for 23A fuse, oF possiby a decoupling capacitor. The capaci- tor values in the milivbrator have been ‘mitted, Perhaps Mr Cooper would confirm thar 12W resistors were imended throughout the circuit, JE. Stevenson, Purley, Surrey. The autor replies agree with Mr Stevenson's comments about ‘components for construction ates: such com- ponents should be available through accesible apply channels and not just to profesionals, ‘but ie s fact that some ofthe most teresting items remain relatively obscure because they are ‘only avalable via the privileged channels, which Tekin is pity. It needs a change of heart on the part of cerain Brish component suppliers fand manufacturers o alter this situation. ‘However, this is not the case with the X3720, whichis a Mullard ferrite available rough’ their stockists. My local “Mullard stockist is Hawat Electronics Ltd, Fiswood Read, Garretts Green, Birmingham B33 OQT, and they will be pleased to supply this ferrite, ‘which they stock in depth. The FX3720 is 4p Det piece (you necd two pisces to make a tans {ormer) and the bobbin DT2723 to suits 35p. ‘This firm has a minimum charge of £2, which should cover a transformer, bobbin, v.2t. and posuage- “There is a world of diffrence berween the ‘non-availability of silicon diodes as discussed in ‘the WW ‘Levers column and the imited avalla- bility of a more specialized component through ‘manufacturers selected agents, such as Hawnt lecuonics. Unfortunately its all oo easy for an author to slant his constructional {Towards the professional supply channel realising iI this case, 1 checked that the FX3720 was available without spotting that there might be a problem for people ike Mr Stevenson who may not know the supply chan. el ‘Regarding the other comments, corrections sppeared inthe August isue. As for resistors, I didnot give wattage ratings for resistors because 1 think most readers of WA wil se at a glance what is needed; it is quite easy to use the for- mula VIR and then make an allowance for fmblent temperature and duty cycle, My view {hat a constructional project should not be cookery-book type recipe to be fllomed blindly, but should involve at least a lie thought by whe constructor about what he i doing. Inthe same vein, T hope you will ead about ef. suppression before winding a trans former, as suggested in the article ‘The 3 A rectifier diode, D;, was not speified because itis non-real; its only function is to provide reverse-poarty protection and itcan be by-passed with shordng lnk ence the unit is installed and working corrcty, POOR DEAL FOR AMATEUR RADIO 1 feel { must reply to G3DRN’s letter criticising ry own leter ‘A poor deal for amateur radio’ ‘Like him, I shall endeavour to answer and ene: ‘ment on his ‘point in tra. Firsy, his ‘All Fools’ comment ~ Is this the best he can manage? Likewise, his claims to impartiality can be dismissed with ite (or bet. ‘erst no comment. "AS ¢-member of the RSGH (fighting from within) feel that Ihave evry right to question ite actions and expect a reply. As regards my

    You might also like