up-to-date electronics for lab and leisure 1 jsi'Isi A ft ^ The easy way to a PCB ...the Seno33 system! a uni« u ' h ;c | *' n slV totally ef , Ling ° f Ie ' and P ol H L rdS , .;na 3 . -ted -ich * he ndg* al sirnP lv ° The oj r . SJJ 0(1 to resisted tr 3cK L ne n e^ fh> n c"> a, >' copP e p r i >h ks [tes- 0 UI , C L in ^'"Lnens'hg Etching inKdJf Tipd i n " n <50 ' sS*5« fini i .* * * 9 INO E»i A revo^s o> etC ; A ^ rob^ ceal ed , «" e . P T Uni« u ! =t h.''i tn« r , Un ,q c the ri , pCB : >5* 5 .vste^Lnce.sLssoC' 3 : SVS nve^ n 1^° cal P r ° bled 1 ® et ch 3n o< » c ined t0 : $<< t b° a ' SAP -rflP |ete All prices post & VAT inclusive Data sheets free of charge Seno33 — The Laboratory in a box From your usual component supplier or direct from •• DECON LABORATORIES LTD. Ellen Steet, Portslade, Brighton BN4 1EQ Telephone: (0273)414371 Telex = IDACON BRIGHTON 87443 idvertisement elektor june 1976 — 601 raiT Many Elektor circuits are accompanied by designs for printed circuits. For those who do not feel inclined to etch their own printed circuit boards, a number of these designs are also available as ready-etched and predrilled boards. These boards can be ordered from our Canterbury office. Payment, including £ 0.15 p & p, must be in advance. Delivery time is approximately three weeks. Bank account number: A/C No. 1 1014587, sorting code 40-16-1 1 Midland Bank Ltd, Canterbury. circuit number issue price % VAT AUDIO austereo 3-watt amplifier HB1 1 5 1.15 (12.5) austereo power supply HB1 2 5 0.55 (12.5) austereo control amplifier HB1 3 5 1.55 (12.5) austereo disc preamp HB1 4 5 0.65 (12.5) edwin amplifier 97-536 6 1.25 (12.5) miniature amplifier 1486 6 0.55 (12.5) equa amplifier 1499 1 1.55 (12.5) equin 9401* 13 2.15 (12.5) tap preamp 4003 4 1.95 (12.5) tap power tap preamp front panels: 9072* 7 2.- (12.5) power 1626A 7 1.65 (12.5) input 1626B 4 1.65 (12.5) volume 1626C 4 1.65 (12.5) tone 1626D 4 1.65 (12.51 width 1626E 4 1.65 (12.5) 730/740 (1C control amplifier) 9191* 8 1.15 (12.5) disc preamp 76131 4040A 3 1.05 (12.5) preco, preamplifier 9398* 13 2.- (12.5) preco, control amplifier 9399* 13 1.15 (12.5) dnl 1234 11 0.85 (12.5) electronic loudspeaker 1527 2 0.55 (12.5) compressor 6019A 3 1.50 (12.5) RF coilless receiver for MW and LW 3166 5 0.85 (12.5) super-plam, main p.c.b. 601 2-1 b 11 2.05 (12.5) super-plam, detector a 601 2-2a 11 1.10 (12.5) super-plam, detector b 601 2-3a 11 1.25 (12.5) ssb receiver 6031* 11 2.25 (12.5) mini MW receiver 9369* 9 0.70 (12.5) FM/TV pll stereo decoder MCI 31 OP 1477 9 0.75 (12.5) CA3090AQ stereo decoder 9126* 5 0.80 (12.5) tv sound 6025 2 1.75 (12.5) tv sound, front-end 9357* 11 1.30 (12.5) ota pll feedback pll fm receiver 6029 7 1.15 (12.5) (3 boards) 9356* 9 4.05 (12.5) aerial amplifier 1668 1 1.15 (12.5) integrated indoor fm aerial 9423* 13 0.80 (12.5) CARS car power supply digital rev counter (control 1563 4 1.50 ( 8 ) p.c.b. only!) 1590 1 1.35 ( 8 ) car anti-theft alarm 1592 4 1.70 ( 8 ) GAMES beetle 1492 4 2.35 ( 8 ) tv tennis, main pcb 9029-1 A* 7 4.40 ( 8 ) tv tennis, modulator/oscillator 9029-2* 7 1.05 ( 8 ) tv tennis, 5-volt supply 9218A* 7 0.90 ( 8 ) tv tennis extensions 9363* 13 5.15 ( 8 ) die 9169* 8 0.80 ( 8 ) RHYTHM AND SOUND minidrum gyrator 1465A 2 0.95 (12.5) minidrum mixer/preamp 1465B 2 0.60 (12.5) minidrum noise 1465C 2 1.25 (12.5) minidrum tap 1 621 A 2 0.90 (12.5) minidrum ruffle circuit 1 621 B 3 1.15 (12.5) automatic bassdrum 1 621 C 3 0.90 (12.5) microdrum 1661 2 1.05 (12.5) rhythm generator (M252AA) 9110* 5/12 0.90 (12.5) rhythm generator (M253AA) ic drummer, instruments 9344-3* 12 1.20 (12.5) mother board ic drummer, instruments 9344-2* 12 1.70 (12.5) daughter board 9344-1 * 12 0.30 (12.5) screening for master oscillator 4011 A 10 0.30 112.5) master oscillator AY10212 401 IB-2 10 1.30 (12.5) ancillary supply for m.o. 401 IB-13 10 0.95 ( 8 ) big ben 95 5028 2 1.50 (12.5) 7400 siren 9119* 5 0.80 (12.5) circuit TIME-KEEPING mos clock 5314 clock circuit mos clock 5314 display board mos clock timebase versatile digital clock clamant clock, alarm clamant clock, time signal clamant clock, striking system car clock (2 boards) car clock front panel (transparent red plastic) kitchen timer digital watch (date only) digital watch (day/date) DISPLAYS twin minitron display twin led display twin decade counter maxi display UAA170, 270° meter, basic board UAA170, 270° meter, front panel TEST EQUIPMENT universal frequency reference distortion meter a/d converter recip-riaa dil-led probe frequency counter, control logic frequency counter, minitron display board frequency counter, led display board frequency counter, counter/ latch/display driver frequency counter preamp frequency counter, —5 V supply tup/tun tester tup/tun tester front panel p.c.b. and wiring tester fm test generator capacitance meter versatile logic probe UAA170, LED voltmeter, basic board UAA170, 16 LED display board MISCELLANEOUS tap sensor mostap light dimmer ttl +5 V supply integrated voltage regulator automatic call-sign generator stylus balance Polaroid timer number issue price % VAT 1607A 1 1.65 ( 8 ) 1607B 1 1.20 ( 8 ) 1620 4 0.85 ( 8 ) 4414B 6 1.40 ( 8 ) 4015-13 7 1.45 ( 8 ) 4015-16 7 1.- ( 8 ) 4015-27 8 1.35 ( 8 ) 7036 6 2.20 ( 8 ) 7036-3 6 1.15 ( 8 ) 9147* 5 0.85 ( 8 ) 9397-1* 10 0.80 ( 8 ) 9397-2* 10 0.80 ( 8 ) 4029-1 2 1.95 ( 8 ) 4029-2 2 1.95 ( 8 ) 4029-3 2 1.95 ( 8 ) 4409 2 2. - ( 8 ) 9392-1 * 12 1.40 ( 8 ) 9392-2 12 1.90 ( 8 ) HD4 5 1.40 ( 8 ) 1437 1 2.20 ( 8 ) 1443 3 1.05 ( 8 ) 4039 2 0.90 ( 8 ) 5027A+B 2 2.45 ( 8 ) 9033* 7 1.50 ( 8 ) 9312* 7 1.30 ( 8 ) 9313* 7 1.30 ( 8 ) 9314* 7 1.05 ( 8 ) 9031-1* 8 1.30 ( 8 ) 9031-2* 8 0.85 ( 8 ) 9076* 4 2.05 ( 8 ) 9076/2A 4 2.30 ( 8 ) 9106* 5 0.60 ( 8 ) 9155* 10 0.85 ( 8 ) 9183* 5 0.85 ( 8 ) 9329* 13 1.05 ( 8 ) 9392-3* 12 1 .- ( 8 ) 9392-4* 12 0.85 ( 8 ) 1457 1 0.85 ( 8 ) 1540 2 1.30 ( 8 ) 1487 6 0.55 ( 8 ) 4046 7 1.05 ( 8 ) 7043b 11 0.80 (12.5) 9017* 10 2.85 ( 8 ) 9343 12 0.40 (12.5) 9379* 12 1.20 (12.5) NEW digibell 9325* 14 1.90 ( 8 ) led light show 9403* 14 2.55 ( 8 ) fet front, preamp 9413* 14 0.80 ( 8 ) fet front, probe 9427* 14 0.70 ( 8 ) * with solder mask All prices include VAT at the rate shown in brackets. 602 — elektor june 1976 publisher's notici elektor decoder What is a TUN? What is 1 0 n? What is the EPS service? What is the TQ service? What is a missing link? Semiconductor types Very often, a large number of equivalent semiconductors exist with different type numbers. For this reason, 'abbreviated' type numbers are used in Elektor wherever possible: — '741' stands for juA741, LM741 , MC741 , MIC741, RM741, SN72741, etc. — 'TUP' or 'TUN' (Transistor, Universal, PNP or NPN respectively) stands for any low frequency silicon transistor that meets the specifications listed in Table 1 . Some examples are listed below. - 'DUS' or 'DUG' (Diode, Universal, Silicon or Germanium respectively) stands for any diode that meets the specifications listed in Table 2. - 'BC107B', 'BC237B', 'BC547B' all refer to the same 'family' of almost identical better-quality silicon transistors. In general, any other member of the same family can be used instead. (See below.) For further information, see 'TUP, TUN, DUG, DUS', Elektor 1 2, p. 458. Table 1 . Minimum specifications for TUP (PNP) and TUN (NPN). VCEO.max 20V *C,max 100 mA hfe.min 100 p tot,max 100 mW ^T,min 100 MHz Some 'TUN's are: BC107, BC108 and BC109 families; 2N3856A, 2N3859, 2N3860, 2N3904, 2N3947, 2N4124. Some 'TUP's are: BC177 and BC178 families; BC179 family with the possible exception of BC159 and BC1 79; 2N241 2, 2N3251 , 2N3906, 2N4126, 2N4291 . Table 2. Minimum specifications for DUS (silicon) and DUG (germanium). DUS DUG VR.max * F,max ' R,max Ptot,max ^D,max 25 V 100mA 1mA 250mW 5pF 20V 35mA IOOju A 250mW lOpF Some 'DUS's are: BA127, BA217, BA218, BA221, BA222, BA317, BA318, BAX13, BAY61 , 1N914, 1 N4148. Some 'DUG's are: OA85, OA91 , OA95, AA116. BC107 (-8, -9) families: BC107 (-8, -9), BC147 (-8,-9), BC207 (-8, -9), BC237 (-8, -9), BC317 (-8, -9), BC347 (-8, -9), BC547 (-8, -9), BC171 (-2, -3), BC182 (-3, -4), BC382 (-3, -4), BC437 (-8, -9), BC414 BC177 (-8,-9) families: BC177 (-8, -9), BC157 (-8, -9), BC204 (-5, -6), BC307 (-8, -9), BC320 (-1,-2), BC350 (-1,-2), BC557 (-8, -9), BC251 (-2, -3), BC212 (-3, -4), BC512 (-3, -4), BC261 (-2, -3), BC416. Resistor and capacitor values When giving component values, decimal points and large numbers of zeros are avoided wherever possible. The decimal point is usually replaced by one of the following international abbreviations: P (pico-) = 10“ n (nano-) = 10“ (micro-) = 10“ m (mill i-) = 10“ k (kilo-) = 10 3 M (mega-) = 10 6 G (giga-) = 10 9 A few examples: Resistance value 2k7: this is 2.7 k ft, or 2700 £7. Resistance value 470: this is 470 n. Capacitance value 4p7: this is 4.7 pF, or 0.000 000 000 004 7 F . . . Capacitance value 10 n: this is the international way of writing 10,000 pF or .01 pF, since 1 n is 10' 9 farads or 1000 pF. Mains voltages No mains (power line) voltages are listed in Elektor circuits. It is assumed that our readers know what voltage is standard in their part of the world! Readers in countries that use 60 Hz should note that Elektor circuits are designed for 50 Hz operation. This will not normally be a problem; however, in cases where the mains frequency is used for synchronisation some modification may be required. Technical services to readers — EPS service. Many Elektor articles include a lay-out for a printed circuit board. Some — but not all — of these boards are available ready-etched and predrilled. The 'EPS print service list' in the current issue always gives a complete list of available boards. — Technical queries. Members of the technical staff are available to answer technical queries (relating to articles published in Elektor) by telephone on Mondays from 14.00 to 16.30. Letters with technical queries should be addressed to: Dept. TQ. Please enclose a stamped, self addressed envelope; readers outside U.K. please enclose an IRC instead of stamps. — Missing link. Any important modifications to, additions to, improvements on or corrections in Elektor circuits are generally listed under the heading 'Missing Link' at the earliest opportunity. Volume 2 Number 6 Editor Deputy editor Technical editors Art editor Drawing office Subscriptions W. van der Horst P. Holmes J. Barendrecht G.H.K. Dam E. Krempelsauer Fr. Scheel K. S.M. Walraven C. Sinke L. Martin Mrs. A. van Meyel UK editorial offices, administration and advertising: 6 Stour Street, Canterbury CT 1 2XZ. Tel. Canterbury (0227) - 54430. Telex: 965504. Bank: Midland Bank Ltd Canterbury A/C no. 11014587, Sorting code 40-16-11, giro: no. 315 4254. Assistant Manager and Advertising : R.G. Knapp Editorial : T. Emmens Elektor is published monthly on the third Friday of each month, price 40 pence. Please note that number 15/16 (July/August) is a double issue, 'Summer Circuits', price 80 pence. Single copies (including back issues) are available by post from our Canterbury office to UK addresses and to all countries by surface mail at £ 0.55. Single copies by air mail to all countries are £ 0.90. Subscriptions for 1976 (January to December inclusive): to UK addresses and to all countries by surface mail: £ 6.25, to all countries by air mail £ 11,—. Subscriptions for 1976 (July/August to December inclusive): to UK addresses and to all countries by surface mail: £ 3.05. All prices include p & p. Subscribers are requested to notify a change of address four weeks in advance and to return envelope bearing previous address. Letters should be addressed to the department concerned: TQ = Technical Queries; ADV = Advertisements; SUB = Subscriptions; ADM = Administration; ED = Editorial (articles submitted for publication etc.); EPS = Elektor printed circuit board service. For technical queries, please enclose a stamped, addressed envelope. The circuits published are for domestic use only. The submission of designs or articles to Elektor implies permission to the publishers to alter and translate the text and design, and to use the contents in other Elektor publications and activities. The publishers cannot guarantee to return any material submitted to them. All drawings, photographs, printed circuit boards and articles published in Elektor are copyright and may not be reproduced or imitated in whole or part without prior written permission of the publishers. Patent protection may exist in respect of circuits, devices, components etc. described in this magazine. The publishers do not accept responsibility for failing to identify such patent or other protection. Distribution: Spotlight Magazine Distributors Ltd., Spotlight House 1, Bentwell road, Holloway, London N7 7AX. Copyright © 1976 Elektor publishers Ltd — Canterbury. Printed in the Netherlands. :ontent$ elektor june 1976 — 603 selektor 608 channel quadrupler — Dipl. Ing. H. Weidner 610 A single-beam oscilloscope is often insufficient nowadays for testing electronic circuits. This article describes a multi-channel switch with which four signals may be displayed simultaneously. measuring pencil — J. Hajek 612 vhf fm reception 613 Owners of FM tuners may often wish to know what sort of signal level they can expect to receive in the locality in which they live. This article investigates the rules governing VHF propagation and shows how received signal strengths may be estimated with a few simple calculations. triac control 617 a.m. mains intercom 618 Mains intercoms of a more or less reasonable quality are still a bit expensive on the market. Consequently, there appears to be a fair demand for a cheap a.m. intercom which will be useful in certain applications where mains interference is not excessive. fet front 620 HF preamp and FET probe for frequency counter. missing link: link 75 624 ejektor 625 An invitation to investigate, improve on and implement imperfect but interesting ideas. vertical fets 628 The Japanese have now developed a semiconductor called 'vertical field effect transistor', intended for use in high power output stages. This article describes the V-FET's construction and operation, along with its appli- cation in commercial circuits. led light show 634 Small is beautiful, as they say, and the LED light show can brighten up the front panel of an amplifier, tuner or tape recorder. digibell — R. Janssen 638 A design for an electronic doorbell that plays the well-known 'Westminster Chime'. The only tuning necessary is a single adjustment to set the entire melody in the required key. market 641 604 — elektor june 1976 advertisemei ELECTRONIC COMPONENT DISTRIBUTORS P.O. Box 25 Canterbury Kent Telephone: Canterbury (0227) 52139 Elektor printed circuit boards and binders. Prices as in this issue Component kits and individual components available by return of post from stocks at Canterbury. All prices include VAT and P.+P. 741-8/1 4DI L £0.22 E.1109 . £6.08 7400 . .£0.15 7038A . .£3.68 7401 . .£0.15 TBA231 . . . . £1.02 7405 . .£0.19 CA1310AE. . £2.27 7447 . .£1.20 CA3080 . . . . £0.79 7473 . .£0.30 CD4011AE . .£0.23 7495 . .£0.69 CD4017AE . .£1.76 74121 . £0.32 CD4049AE . .£0.67 MM5314N . . £4.77 TBA120 . . . . £1.24 SFE 6 mA (6 MHz Filter) £0.90 l/C EXTRACTOR £0.55 NE555V £0.65 l/C INSERTER £1.40 2N3055 £0.54 Crystals for any requirement PH 100 5-digit 30 MHz freq. counter £99.50 Above types are partial examples. Write or phone for details. 7 p.m. to 9 p.m. phone Dover (0304) 812332 Stock List free on request. FM Aerials The Fuba UKa Stereo range of aerials provide complete coverage of high quality and high performance FM aerials. The Fuba Uka Stereo 8 is famous for really good long distance reception. v UHF TV Aerials The famous Fuba XC 3 range of UHF TV aerials, and the unique AKV 450 “Active Capsule” amplifier which converts the passive aerial to an active one, will often provide out of area TV* \ \ channels. v \ XC391 The unique Fuba indoor active FM aerial gives a level of performance not previously obtainable with indoor aerials. 'Full set of leaflets sent on receipt of large SAE with 9p stamp. Please mark envelope Cat. E. Prices will be included. SPECIAL OFFERS for a limited period only. Stolle Automatic aerial rotator with control unit at £ 39.80 including VAT and 15% discount. Carr free. Exa UKW 2 FM aerial 2 elements; anodised type finish, £4.85 including VAT, carr and 20% discount. 5 Way rotator control cable, 20 metere length £4.00 including VAT and carriage. Teldis 2 stage ultra broadband mast head amplifier and mains power supply unit, £ 13.70 including VAT and carriage and approx. 25% discount. All offers apply for when ordering.’ 4 weeks from date of publication. Please refer to this advertisement All products available from us CWO, Access or Barclaycard. Audio Workshops Ltd 29 HIGH STREET ROBERTSBRIDGE. SUSSEX. TEL ROBERTSBRIDGE 88058C (STD 058C Phoenix Electronics iSolent) Limited 46 OSBORNE ROAD, SOUTHSEA, PORTSMOUTH, HANTS. CALLERS AT OUR SHOP WELCOME Components and tools supplied guaranteed from franchise agreements. Member of AFDEC — the franchised component distribu- tors association. Our catalogue sent for 20p including postage. All prices include VAT. Carriage is 20p extra on each order — mini- mum value £ 1 TTL -93 TYPES AVAILABLE 7400 £0.17 7401,2,3,4,5,10 £0.21 7413,70,74 £0.33 7441,2,91 £0.95 7481,2,151,1 57 £1.35 7490 £0.75 7492,3 £0.89 74121 £0.59 74190,1 £2.20 74192,3 £ 1.99 LINEAR CIRCUITS TRANSISTORS BC1 07,8,9 £0.11 301 A, 307, 709, 741 £0.39 BC1 77,8,9 £0.15 723 regulator £0.49 BC1 82L,3L,4L £0.09 537,558 dual £0.59 BC204,5,6 £0.14 747 dual 555 timer £0.69 £0.85 BC207,8,9 £0.09 309K regulator £ 1.50 BC21 2L,3L,4L BDY56 £0.10 £1.72 DTL - 1 1 TYPES AVAILABLE BD1 35,6 2N221 9,20,21,22 £0.37 £0.22 Gates from £0.29 2N2904,5,6,7 £0.24 Flip-flops from £0.39 2N3442,BDY53 £0.88 DIODES AA143.4.0A200 1N914.6 ,4148,9 1N4001-7 1 N5402-6 ZENERS 5% ’/2 Watt 3.3-1 5V 1 Watt 4.7-24V 10 Watt 8. 2-24 V BRIDGES 1-AMP P.C. WOO 5 50V W08 800V £0.08 £0.06 £0.07-£0.13 £ 0.1 7-£ 0.21 £0.11 £0.22 £0.83 £0.27 £0.40 RESISTORS ’A WATT Metal film 2% 30R-300K CAPACITORS Aluminium electrolyc, Polyester, Ceramic disc. Tantalum CONNECTORS, CLIPS. KNOBS, LAMPHOLDERS, SWITCHES, FUSEHOLDERS, FUSELINKS All available in a wide range of sizes and designs SOLDERING EQUIPMENT Weller 'Marksman' £3.02 Adcola 'Invader' £3.11 Weller 'Expert' Gun £7.84 'marksman' Stand £2.18 'invader' Stand £1.86 R500 Desolder Iron £11.60 SOLDER AND DESOLDERING BRAID AVAILABLE. WIDE RANGE OF SOLDERING ACCESSORIES INTRODUCTORY OFFER TO ELEKTOR READERS SPECIAL VALUE BARGAIN PACK INCLUDING: Small signal transistors — Triacs — Plastic power transistors — Diodes — HV rectifiers — Zeners. At least 50 pieces of dis- crete semiconductors. Integrated circuits — TTLand linear. Panel hardware selection including: Knobs — Jacks — Con- nectors — Fuses — Aluminium electrolytic, ceramic and film capacitors. Total value at least £ 10 — sent with our new catalogue at an inclusive price of £ 3, post free. advertisement elektor june 1976 — 605 Join the Digital Revolution Teach yourself the latest techniques of digital electronics Computers and calculators are only the beginning of the digital revolution in electronics. Telephones, wristwatches. TV, automobile instrumentation — these will be just some of the application areas in the next few years. Are you prepared to cope with these developments? This four volume course — each volume measuring Ilf" x 8f " and containing 48 pages — guides you step-by-step with hundreds of diagrams and questions through number systems. Boolean algebra, truth tables, de Morgan's theorem, flipflops. registers, counters and adders. All from first principles. The only initial ability assumed is simple arithmetic. # At the end of the course you will have broadened your horizons, career prospects and your fundamental under- standing of the changing world around you Design of Digital Systems Bookl • • • „ • • «• • • • * • Also available - a more advanced course in 6 volumes: 1. Computer Arithmetic 2. Boolean Logic 3. Arithmetic Circuits 4. Memories & Counters 5. Calculator Design Computer Architecture £6.20 plus 80 p Offer Order this together with Digital Computer Logic & Electronics for the bargain price of £ 9.70, plus 80 p p & p p & p. Design of Digital Systems contains over twice as much information in each volume as the simpler course Digital Computer Logic and Electronics. All the information in the simpler course is covered as part of the first volumes of Design of Digital Systems which as you can see from its contents also covers many more advanced topics These courses were written so that you could teach yourself the theory and application of digital logic. Learning by self-instruction has the advantages of being quicker and more thorough than classroom learning. You work at your own speed and must respond by answering questions on each new piece of information before proceeding to the next. Guarantee — no risk to you If you are not entirely satisfied with Digital Computer Logic and Electronics or Design of Digital Systems, you may return them to us and your money will be refunded in full, no questions asked. Designer Manager Enthusiast Scientist Engineer Student Digital Computer Logic and Electronics A Self -instructional Course C PG.*ne MA (Cantab) A W Un«m BA (Cantab) Book 1 Basic computer logic Book Q 1 Logical circuit 1 elements Book J 1 Designing circuits to carry out ‘ logical functions Book A Flipflops L and r registers Digital Comptr Logic and Electronics Book tlrai £4.20 plus 80 p packing and surface post anywhere in the world Quantity discounts available on request. Payment may be made in foreign currencies. VAT zero rated. r Ei4 1 To: Cambridge Learning Enterprises, FREEPOST, St. Ives, Huntingdon, Cambs PE 17 4BR * Please send me set(s) of Digital Computer Logic & Electronics at £ 5.00 each, p & p included. | * or set(s) of Design of Digital Systems at £ 7.00 each, p & p included. • or combined set(s) at £ 10.50 each, p & p included. Name I Address * delete as applicable No need to use a stamp - just print FREEPOST on the envelope J 606 — elektor june 1976 advertisement I wouldn t give a dime for an elektor circuit You don’t have to on average they cost about half a dime (2 p) each! HIGH QUALITY PROFESSIONAL AUDIO MODULES Over the years there has been a growing demand for high quality audio power amplifier modules. At last and never before has such quality been offered at such a reasonable price. All our modules at JPS are fitted with full output protection, and have a super High Stability Intergrated Circuit front end containing within itself no less than twenty transistors. All JPS products carry a two year guarantee. POWER OUTPUT: -i r A 170WR.M.S. 13U FREQUENCY RESPONSE: D.C.- 21kHz - 0.2dB POWER BANDWIDTH: D.C. -20kHz -0.2dB POWER OUTPUT: 65 W R.M.S. FREQUENCY RESPONSE: D.C. 21kHz 0.2 dB POWER BANDWIDTH: D.C. 20kHz 0.2 dB SLEWING RATE: 8.4V per microsecond T.H.D: .03% INPUT SENSITIVITY:* i OdB (0.775V) 150 W INPUT IMPEDANCE:* 47k HUM and NOISE: >100dB below 150 Watts DAMPING FACTOR: >200 to 1kHz POWER REQUIREMENTS: ±55V D.C. PRICE: £32.02 (inc. VAT) SLEWING RATE: 8.6 V per microsecond T.H.D: .03% INPUT SENSITIVITY:* OdB (0.775V) 50W 8 ohms INPUT IMPEDANCE:* 47k HUM and NOISE: >100dB below 60W DAMPING FACTOR: >200 to 1kHz POWER REQUIREMENTS: ±35 V D.C. PRICE: £19.20 (inc. VAT) *At a small extra charge , these parameters may be changed to order. Impedance: Max 250 K - Sensitivity: lOOm/V RMS \ As both modules are totally D.C. coupled, they may also be ordered as D.C. amplifiers, the Power Response then being -OdB + 0.5dB - D.C. to 20 kHz POWER SUPPLIES below are available upon request:- PS50 (Drives single 50 Watts) £11.0/ PS150'1 (Drives single 1 50 Watts) £18.15 PS502 (Drives dual 50 Watts) £14.57 PS.150'2 (Drives dual 150W'atts) £23.00 N.B. It is not recommended that the D.C. Modules are used for audio For For Q/50 watt modules @ £32.02 fine. VA Tj Q 50 watt modules (® £1 9.20 (inc. VA T I Quantity Discounts upon request . . . NAME ADDRESS installations as extreme low frequencies damage loudspeaker systems. Please send me further information Q I herewith enclose Cheque/Postal Order for £ made payable to JPS Associates. associates All JPS Products carry a full Two Year Guarantee. BELMONT HOUSE STEELE ROAD PARK ROYAL Postal Code LONDON NW10 7AR TELEPHONE 01-961 1274 Wl advertisement elektor june 1976 — 607 B. BAMBER ELECTRONICS DEPT E, 5 STATION ROAD, LITTLEPORT, CAMBS., CB61QE Tel: ELY (0353) 860185 (2 lines) Tuesday to Saturday. ALL ITEM + 8% VAT (UNLESS OTHERWISE MARKED) . BARGAIN OFFER MIXED COM- PONENT PACKS, containing resistors, capacitors, switches, pots, etc., All New, (Random sample bag revealed approx. 700 items) £ 2.00 per pack, while stocks last. TRANSISTOR HEATSINKS, to take 2 x TO 18 transistors, screw in clamps, block size Ix'/ix Kin, with.holes for mounting 3 for 50p . MINIATURE 2 PIN PLUGS & SOCKETS (Fit into K " hole, pins onclosed, with covers for chassis mounting, or can be used for in-lino connectors). Bargain pack of 3 plugs + 3 sockets + covers 50p • PROGRAMMERS (Magnetic Devices) Contain 9 microswitches (suitable for mains operation) with 9 rotating cams, all individually adjustable, ideal for switching disco lights, displays, etc., or industrial machine programming. (Need slow motion motor to drive cams, not supplied) 9 switch version £1.50 •.HEAVY DUTY HEATSINK BLOCKS, undrilled, base area 2K" x 2", with 5 fins, total height 2K" 50p each •9V RELAYS, Continental type, 2 pole change over 35p • RUBBER MAGNETS K" square, with mounting hole 20 for 30p • SPERRY 7-SEGMENT P.G.D. DISPLAYS, digit height 0.3 in red, with decimal points, 150V to 200V (nominal 180V) operation. These are high-volt industrial type and therefore brighter than normal displays. All brand new, AT THE BARGAIN PRICE OF 50p PER DIGIT. Typo 332 (two digits in one mount) £1.00 each Typo 333 (three digits on one mount) £1.50 (Sorry, no single digit available). Data Supplied. BSX20 Transistors 3 for 50p BC108 (metal can) 4 for 50p PBC108 (plastic BC108) 5 for 50p OC200 Transistors 6 for 50p BSY95A Transistors 6 for 50p BFY51 Transistors 4 for 60p BCY72 Transistors 4 for 50p PNP audio type TO 5 Transistors 12for25p Telephone Type earpiece insert .... 50p IK" Polythene chassis mounting fuse- holders 6 for 30p LES Lamps. 24V 1.2W 10for40p Mullard Tubular ceramic trimmers. 1-1 8pf 6 for 50p I.C.'s, some coded, 14 Dl L type, untested mixed 20 for 25p Miniature slider switches, 2 pole 2 way 5 for 50p DIECAST BOXES The range of aluminium alloy diecast boxes has the unique feature of internal slots for dividing parts, e.g. screens, printed circuits boards etc. For use as screened sub assemblies, as rigid chassis, as junction boxes, for test sets and for complete equipment. In the following sizes: 4.3" x 2.3" x 1.2" approx 85p 4.8" x 2.8" x 1.5" approx 75p 4.8" x 3.8" x 1 " approx 85p 4.8" x 3.8" x 2" approx £1.00 6.8" x 4.8" x 2" approx £1.45 4.8" x 3.8" x 3" approx £1.55 6.8" x 4.8" x 4" approx £2.25 8.6" x 5.8" x 2" approx £1.85 10.6" x 6.8" x 2" approx £2.25 * NO INTERNAL SLOTS PLUGS & SOCKETS T.V. PLUGS (Metal Type) . . 5 for 50p T.V. SOCKETS (Metal Type) . 4 for 50p T.V. LINE CONNECTORS (Back-to-back sockets) 4 for 50p DIN 3-pin LINE SOCKETS 15p each DIN 3-pin PLUGS 15peach DIN 6-pin Right-angled PLUGS 20p each PLEASE ADD 25% VAT TO ALL DIN & TV SOCKETS. BARGAIN PACKS MIXED DIELECTRIC CAPACITORS (Approx 100) £1.50 ELECTROLYTICS (LOW VOLTAGE TYPES) £1.50 (Approx 100) ELECTROLYTICS (HIGH VOLTAGE TYPES) £1.50 (Approx 50) RESISTORS (MIXED WATTAGES) MIXED VALUES £1.00 PLEASE ADD 12,5% VAT TO ALL RESISTORS & CAPACITOR PACKS HIGH QUALITY SPEAKERS, 8 1 /*" x 6” elliptical, only 2" deep, inverse magnot, 4 ohms, rated up to 10 W, £ 1.50 each, or 2 for £ 2.75 (qty. discount available + 12,5 VAT. WELLER SOLDERING IRONS WELLERSOLDERING IRONS EXPERT Built-in-spotlight illuminates work Pistol grip with fingertip trigger. High efficiency copper soldering tip. EXPERT SOLDER GUN £6.80 + VAT (54p) EXPERT SOLDER GUN KIT (SPARE BITS. CASE. ETC.) £9.80 + VAT (78p) SPARE BITS PAIR, 26p + VAT (2p) MARKSMAN SOLDERING IRON Unbreakable heat-resistant handle. Stain- less steel barrel. Special steel heating elements with Mica insulation bodded in ceramic. SP15D 15W £2.52 + VAT (20p) SP25D 25W £2.52 + VAT (20p) SP25D 25W+ BITS etc, KIT £3.12 + VAT (25p) BENCH STAND with spring for MARKSMAN IRONS £1.80 + VAT (14p) Spare sponges, as for TC PI . SPARE BITS MT8 for 1 5W 42p + VAT (3p) MT4 for 25W 35p + VAT (3p) TC PI TEMPERATURE CONTROLLED IRON Most versatile soldering tool yet designed. 48W pencil out-performs uncontrolled irons several times its weight and consumption. Temperature controlled iron & PSU. £18.43 + VAT (£1.47) SPARE TIPS Type CC SINGLE FLAT, Type K DOUBLE FLAT FINE TIP. 90p each + VAT (8p) ALL SPARES AVAILABLE. MULTICORE SOLDER Size 5 SAVBIT 18swg in alloy dispenser 32p + VAT (3p) Size C1SAV18 SAVBIT 18 swg 56p + V AT (4p) Size 12 SAVBIT 18swg on plastic reel £1.80 * V AT (1 5p) MINIATURE PLIERS HIGH QUALITY "CRESCENT" MADE IN USA £4.35 + VAT (35p) SIDE CUTTERS HIGH QUALITY "CRESCENT" MADE IN USA £5.45 + VAT (44p) SOLDER SUCKERS (HIGH QUALITY, PLUNGER TYPE) STANDARD £4.50 + VAT (36p) WITH SKIRT £4.95 + VAT (40p) Spare nozzles (PTFE) 60p each + VAT (5p) RUBBER BULB TYPE A handy inexpensive tool for the quick removal of solder. Small, lightweight and easy to use. Teflon tiplets easily changed or replaced. No. 881 Complete Tool. £1.20 + VAT (lOp) No. 8810 Nozzle only 50p + V AT (4p) Terms of Business: CASH WITH ORDER. MINIMUM ORDER SAE with ALL ENQUIRIES Please. PLEASE ADD VAT AS £1.00. ALL PRICES INCLUDE POST & PACKING (UK ONLY). SHOWN. ALL GOODS IN STOCK DESPATCHED BY RETURN. QUARTZ CRYSTAL UNITS from ★ ★ ★ ★ TO DEF 5271-A 1 .0 — 60.0M H z FAST DELIVERY HIGH STABILITY TEL. HYTHE 848961 STD CODE 0703 Write for Leaflet AT-1 Mcknight CRYSTAL Co. Ltd. Hardley Industrial Estate, Hythe, Southampton S04 6ZY (248) sav i mu saw it in elektor and don't forget...... Printed circuit boards are available for many projects published in Elektor. Refer to the 'eps print service' list in this issue and quote ref. no. when ordering. The Semicon International e ~ Transistor Index Easy alpha-numeric reference to the ratings and characteristics of some 24,000 transistors of international origin. European, U.S.A., Japanese. Essential guide for all Engineers, Technicians, and Buyers. Over 450 pages of basic information. By far the best manual of its kind available anywhere to-day. * EXTENSIVE SUBSTITUTION GUIDE * CV & BS DEVICES & EQUIVALENTS * TERMINATION DRAWINGS * ALTERNATIVE SOURCES OF SUPPLY Remit with order C^Q Elsewhere £10.90 price. UK only. ■ WW Surface mailing. Please send copies of the Semicon Transistor Index, 6th Ed. to: Name Address □ / enclose cheque/postal orders for £ □ My Access/ Bard ay /In ter bank Card No. is Refund if not satisfied and book is returned within 14 days. Semicon Indexes Ltd. 2 Denmark St. Wokingham. Berks. RG1 1 2BB TEL: WOKINGHAM (0734) 786161 608 — elektor june 1976 selektoi rasnnrseueKTDrsei.1 Build a Digital Watch — the easy way For those who want to make a digital watch but are deterred by the degree of miniaturisation involved, a range of preassembled watch modules is available from Litronix. Prices have fallen so rapidly in recent months that it is now cheaper to buy one of these modules than to purchase the individual components for a watch, and it is only necessary to fit the module into a case and install batteries to obtain a fully- working watch. The timekeeping function is integrated into a single ion-implanted CMOS chip and the display drivers are contained in two silicon bipolar IC’s. These, together with the LED display, 32.768 kHz crystal, capacitors and resistors, are mounted on a ceramic substrate using hybrid assembly techniques. This in turn is mounted in a high impact plastic frame to provide mechanical protection for the module. Gold-plated switch contacts are provided around the periphery of the module to activate the display and timesetting functions. The module is powered by two RAY- 0-VAC RW44 silver oxide batteries (or equivalent) and battery life is said to be one year with up to twenty interrog- ations per day. The cheaper version of the watch module, the LWM-653 1 , has an hours and minutes display with a flashing colon for seconds. One-off price is around £ 1 5. The more expensive version, the LWM6560, retails for about £ 1 8 and displays hours and minutes, seconds, day-of-week and date. It also incor- porates a light sensor to adjust the display brightness to suit ambient light and thus extend battery life. Litronix House , 593 Hitchin Road, Stopsley , Luton, Bedfordshire . Vegetable Growing by Computer Researches in Naaldwijk, Holland, wish to know exactly how fast tomatoes ripen and how well cucumbers, for example, grow in greenhouse conditions. Assistance of a computer has been called on at the research and experimen- tal institute for the cultivation of fruit and vegetables. At the present time this is the largest research project of its kind in the world. A Siemens 330 process computer monitors and controls the environmental conditions in 24 growing compartments within the large green- house. In addition to the central processing unit with a main memory of 64 K words ( 1 6 bits each), the computer system comprises a series of data input/output devices. These include, process inter- facing devices for connection to the measuring and control systems of the individual climatic chambers. The task of this process periphery is to collect 800 analog and 80 digital signals, while also handling 450 digital output signals. Twenty-six analog signals are recorded in each of the 24 climatic chambers (each 56 m in area). These include temperature values for the air and soil, relative humidity, C0 2 concentration of the air, temperature of the water in the heating system, signals indicating window positions and the various values for the irrigation and drainage systems and the heating system. Sixteen digital outputs in each climatic chamber control the motors of the valve drives and window adjusters. In addition, the weather station parameters can also be included in the calculations and control operations. The signals issued from each sensor are scanned once every minute. The process computer calculates the manipulated variables and the setpoints of the secondary analog backup controllers by direct digital control so that a smooth changeover to analog control is possible if required. Every day an estimated one million measured values, signals and commands are exchanged between the process computer and its peripherals. All relevant data can be logged in tabular form, using the typewriter. In addition, the case history of 256 values over the the previous 96 hours can be graphically represented on the colour curve display station. The aim of the investigations in Naaldwijk is not only to research the optimum conditions for growing fruit and vegetables but also to determine the climatic conditions under which the plants are least susceptible to diseases. It should then be possible, to a certain degree, to forego the use of chemical agents and pesticides. The investigations are at present con- cerned with tomatoes and cucumbers. The research program is later to be extended to other types of vegetable, e.g. red peppers, aubergines, beans and lettuce. Fruit and flowers are also to be investigated at a later date. At present, tomatoes grown in Dutch greenhouses have annual production value of approximately £ 9.6 million. Siemens AG Zentralstelle fiir Information Postfach 3240, D-8520 Erlangen 2 Federal Republic of Germany 'Noiseless' Discs and tapes A new process which completely eliminates surface and background noise from disc recordings and captures for the first time the full dynamic range of the music is announced by dbx, Incor- porated, manufacturer of noise re- duction systems for the professional studio and the audiophile. The process permits commercial discs as played in the home to equal the per- formance quality of studio master tapes. This is accomplished by elec- tronically compressing the recorded elektor elektor june 1976 — 609 ■JUST afij MH tHU p HKfWSi Wilmk *W$m pmb ignal by a factor of 2: 1 at the time the naster disc is cut, and expanding the ignal by a complementary factor of :2 at the point of playback. If a dbx ncoded master tape is used, full lynamic range and freedom from noise vill be realized upon playback. Master apes produced with other types of loise reduction systems, or with no loise at all, may be used and the played )ack disc will sound equal to the naster tape. Ordinary discs are limited o a dynamic range of some 60 dB, vhereas the dbx encoded disc has a ange well in excess of 100 dB. rhe dbx process compresses the lynamic range of the music to a dynamic range envelope’ which fits :onveniently within the inherent imitations of the record medium, then expands the music to its full original dynamics at the point of playback, rhis compression and subsequent expansion reduces record surface noise md other unwanted background noise to inaudibility, so that when the musical program stops, no sound of any kind is heard from the playback system. Hie complete absence of background noise also makes the quiet portions of the music easier to hear, and the defi- nition of individual voices and instru- ments in ensemble music is dramatically improved. A significant advantage of the dbx disc encoding process is that it does not obsolete any existing manufacturing technique or equipment presently in use in the recording industry, nor does it increase actual product cost in any way. On the contrary, dbx encoding offers numerous opportunities for reducing the cost of recorded music without compro- mising quality. For example, with dbx encoding, record grooves may be placed closer together, increasing the amount of music on a record by up to 30%. Electronic expansion circuitry, similar in cost and complexity to Dolby B and quad matrices, is required at the point of playback to properly decode the dbx compressed signal. The decoder cir- cuitry is now available to audio equip- ment manufacturers for inclusion in consumer audio components and sys- tems on a license basis. Also, many audio component dealers are now selling the 120 consumer series of compressor/ expander noise reduction systems which have disc decoding circuitry built in. The 1 20 series uses the same 2: 1 linear decibel compression/expansion principle used in dbx professional studio equip- ment, but the sensing circuits have been optimized to best complement the bandwidth requirements of consumer grade reel-to-reel, cassette and cartridge tape recorders. The 120 series is not compatible with dbx professional format tape noise reduction systems used in recording studios. The new systems allow consumer grade recorders with signal-to-noise ratios as low as 45 to 50 dB to produce full dynamic range tapes which are audibly free from tape hiss and background noise. In excess of 30 dB noise re- duction can be realized with the 1 20 system, along with 10 dB headroom improvement which reduces the likeli- hood of tape saturation. The 120 noise reduction format is also used by record manufacturers for pro- cessing of dbx encoded records, and a 120 family decoding device is required for playing the dbx encoded discs. Two models are initially available in the 120 family of noise reduction units. Model 122 is a two-channel record or playback unit. That is, it will either record or playback, and is switched from one function to the other. Model 1 24 is a four-channel record or playback unit suitable for the full range of quadraphonic activities, and having the added feature that when used in a two-channel system it can record and playback at the same time, permitting the recordist to monitor the decoded or normalized signal during recording. Commercial record labels currently releasing stereo discs in dbx encoded format include Klavier and Creative World, and negotiations are in progress with a number of other lables to release material in dbx encoded format. Both Models 122 and 124 will decode presently available dbx encoded stereo discs, dbx disc encoding is equally applicable to the production of quadraphonic releases as well, regard- less of whether they are produced in discrete (4-4-4) or matrixed (4-2-4) format. dbx Incorporated , 296 Newton Street , Waltham, Massachusetts 02154. Ion beams etch chip structures New etching process for superfine structures on semiconductor chips. Semiconductor chip structures have become unbelievably small, and this trend will continue. The wavelength of the light beam, used to create com- ponent contours on chips using the photomasking processes, is nowadays often not short enough. This is why electron beams with their markedly shorter wavelengths must be used. Such beams are capable of producing structure spacings of 1 and less. However, it is then necessary to use a different method to create the struc- tures, which are defined by a photo- mask. This is because the chemical Comparison of chemical and ion beam etch- ing. The chemically etched structures in the left-hand photo are characterized by under- cutting below the photo-mask and the curved side walls. The ion-etched structures in the right-hand photo are characterized by smooth, straight side walls. Siemens Photo etching processes that are being used now have an undercutting effect — whereby the side walls of the structures are washed away from underneath. This just cannot be tolerated in the case of superfine structures in the submi- crometer range. A process is at the present time being worked on at the Siemens laboratories, this process uses an ion beam, working mechanically like a sand blasting jet. It cleanly cuts out even the very finest chip structures. The fast argon ions required for this job are produced in a plasma chamber, then are directed onto the photo- resisted silicon chips. The structures are transferred practically without change in dimensions independently of the resist adhesion. The particular advantage of this process is that the side walls of the chip structures are smooth in contrast to chemical etching, and the angle of the side walls has a uniform value of around 65 (see photo). Siemens AG Zentralstelle fiir Information Postfach 3240 , D-8520 Erlangen 2 West Germany 610 — elektor june 1976 channel quadruple Dipl. Ing. H. Weidner A single-beam oscilloscope is often insufficient nowadays for testing electronic circuits. This article describes a multi-channel switch with which four signals may be displayed simultaneously. Two-channel oscilloscopes are, by now, commonplace, and various types of two- channel switches are available for extending single-beam oscilloscopes. By somebody-or-others law, however, we continue to build circuits which we are incapable of testing, and even two-beam displays are often insufficient. Bear in mind that the available two-channel switches are expensive, and we are left with a demand for a simple and reliable multi-channel switch within the pur- chasing power of the amateur. The requirements of such a switch are: 1. four channels; 2. unity gain and a facility for attenu- ation; 3. Y-position separately adjustable for each channel; 4. both chopped and alternating switching modes; 5. facility for selecting each channel separately. This article describes a switch to meet these requirements, with the design being kept as simple as possible by starting off with a suitable integrated circuit. A survey of ICs available on the marke led to the selection of the HA240! made by Harris. This consists of fou operational amplifiers (opamps), onl\ one of which is activated at any ont time depending on the information a the two control inputs (pins 1 5 and 16) The outputs of the four opamps art internally connected to the commoi output amplifier, so that the output o the activated opamp is available at tht common output point (pin 10). Other wise, the IC behaves as expected for ai opamp. The maximum output voltagt hannel quadrupler elektor june 1 976 —611 igure 1. Circuit diagram of the four-channel witch. igure 2A. Pin connection diagram for the IA2405. igure 2B. Pin connection diagram for the M 318. igure 3. Front view of the (German!) proto- /pe. ariation is ± 10 V, the gain and attenu- tion being adjustable in the usual way vith feedback resistors. rhe circuit ; igure 1 gives the circuit diagram of the witch. As can be seen, the four opamps n IC1 provide the central part of the ircuit, with their gain and attenuation >eing controlled by the input resistors 11 . . . R4 and the feedback resistors . . . R8. The gain of the first opamp s controlled by the ratio of R1 : R5 similarly for the other opamps); the naximum gain in this case being unity 0 dB), and the attenuation being con- rolled by the variable resistors 15 . . . R8. Calibration of the attenu- ition has not been included (since phase omparisons are usually of more nterest) but it would be fairly simple o include calibration if required, either )y fitting a calibration scale for the 'ariable resistors, or by replacing them vith multi-position switches between ixed resistors. The maximum input voltage should not exceed ± 10 V, which s sufficient for most applications, ilthough the input voltage range could tlways be extended by using voltage- iivider probes with an attenuation of 10 : 1 . Hie Y-position of the signals displayed pn the screen is controlled by the potentiometers R17 . . . R20 which control the voltage to the non-inverting nputs of the opamps. The values of the esistors in series with these potentio- neters have been chosen to give a full- screen display on the oscilloscope with he input sensitivity set to 1 V/cm. Other values may be substituted if other atios are required. Since the IIA2405 performs an inversion petween input and output, IC6 has been ncluded to invert the signal again so hat the correct polarity is obtained at he output of the switch. Also included lere is a feed-forward circuit R30, CIO, which improves the slew rate. VIoving on to channel selection, this is ichieved by the switch S6, through which digital information is supplied 2a HAS4Q5 . 15 14 13 12 11 10 16 s ^2 3 4 5 6 7 8 7044- 2A Top view 7044 • 2 B T05 package 612 — elektor june 1976 channel quadrupler measuring penc to the two controlling inputs. S6 is a five-position switch of which positions 1 . . . 4 are used to select opamps (i.e. channels) 1 ... 4 respectively, by providing the following binary signals at the input pins: pin 1 5 pin 1 6 selected channel 0 0 1 0 1 2 1 0 3 1 1 4 Position 5 allows the outputs from the dual flip-flop IC5 through to pins 1 5 and 16, so that all four channels are displayed on the screen. The requirement for alternative switching modes is more difficult to achieve although in theory both chopped and alternating mode can be provided, and indeed have been provided in this particular circuit. Practical diffi- culties may however intervene as explained shortly. The mode is selected by means of the Figure 4. Display of a square wave using the switch. Top: input signal. Bottom: output signal. X-scale: 2 ps/division. Y-scale: top: 2 V/division; bottom: 0.2 V/division. switch S5, which connects one of the two control sections to IC5. In chopped mode the switching voltage is generated by IC2 and IC3 of which the latter (timer 1C type 555) is connected as a multivibrator producing a square wave of 1 00 kHz. This signal drives the monostable IC2, which converts it into spikes whose negative edges trigger the flip-flop 1C5; the frequency of the switching signal is 50 kHz. The output from IC3 also drives a further mono- stable IC4, whose output is converted by the circuit of T3, T4 into negative pulses which suppress the beam during switching. The ‘beam suppression’ signal should be connected to the Z-input of the oscilloscope. If the oscilloscope is not provided with a Z-input the simplest solution is to ignore beam suppression, in which case IC4, T3, T4 can be omitted. The absence of beam suppression is only noticeable when the signal frequency and the chopper frequency are similar. At this point the only solution is to use alter- nating mode in which one complete signal or channel is ‘written’ on the screen, then the next signal is ‘written’ etc. This mode can also be useful in its own right (e.g. when amplitude compari- sons are of more interest than phase ones) not just when chopped mode fails. The switching pulses required for alter- nating mode come from the oscilloscope itself at the end of each deflection period. To make use of these pulses the sawtooth voltage for the horizontal deflection (if necessary via a square wave voltage shaper) or the gate voltage must be available. This is not true of many single-beam oscilloscopes, so it is recommended that such an output be provided if working in alternating mode is to be achieved. The gate signal or the pulse derived from the sawtooth voltage is amplified by the circuit of Tl, T2 to provide a signal to IC5. Technical data Figure 4 shows a 100 kHz square wave displayed on an oscilloscope using the switch with the attenuation set to 10 : 1. Note that as this setting, the slope of the edge of the output signal is still good, but when the ‘gain’ control is increased to maximum (1:1) the band- width of the unit is restricted to such an extent that only sine waves up to 100 kHz can be satisfactorily displayed. The input impedance of the switch is less than 100 k, while the input impedance of most oscilloscopes is 1 M. When measuring with this switch, there- fore, the extra loading on the measuring point must be taken into account. As far as power supplies are concerned, three stabilised voltages are needed, vis: ± 15 V for the opamps and beam suppression, and +5 V for the ICs. The current consumption of the circuit is about 25 mA from each 15 V supply, and 60 mA from the IC supply. To keep the circuit as compact as possible, it is recommended that integrated voltage regulators be used. In the prototype, L 1 29 and L 1 3 1 (SGS) were used. H J. Hajek measuring pencil Here is a device to help with that of recurring requirement — an extra pai of hands. This measuring pencil can b< used as a probe for inspecting voltages in addition to its conventional appli cation of writing down the results. The measuring pencil consists of j propelling pencil made of a syntheth resin. A flex is soldered to the push button such that the propelling mechan ism is unimpaired, and the other end o the flex is provided with a plug to fi the measuring instrument (e.g. a volt meter). To use the pencil, connect tht circuit to be tested to the earth termina of the meter, and the pencil to the inpu socket. Using the pencil as a probe, the circuit voltages may now be measurec at various points and the results writter down with the same instrument. The measuring pencil is particularly suitable for circuits with low voltages (< 42 V AC or < 60 V DC) as the insu- lation of the pencil is not then a problem. The contact resistance between the measuring point and the meter is less than 1 £2 in this design, and when used with modern high impedance equipment, the measuring error is negligible. u HF FM reception elektor june 1976 — 613 ITHFFM reception )wners of FM tuners may often /ish to know what sort of signal jvel they can expect to receive in he locality in which they live. 'he simplest and most accurate nethod is to obtain a direct eading using a field strength neter, but of course very few leople possess one of these nstruments. The charts iroduced by the BBC of the ervice areas of their various ransmitters provide a useful wide, but it is often possible vith a sensitive receiver and good ierial to obtain a usable signal lutside the service area. This rticle investigates the rules loverning VHF propagation and hows how received signal trengths may be estimated with i few simple calculations. : igure 1. The optical line of sight is approxi- nately Dh = \/2R • h, + \/2R • h 2 . After ntroducing the radius of the earth this >ecomes Dh = 3570 • (y/hi + n/F^). h,, h 2 ind Dh are in m. The optical line of sight days a part in the empirical formula for field trength calculations. : igure 2. Bending effects increase the recep- ion range of radio waves to beyond the >ptical line of sight. The range is then Jh = 4120 • (Vh, +S/M. : igure 3. As a result of reflections at the roposphere, considerably greater distances ire bridged. The intensity varies, however, as i result of cosmic influences. 10000 1000 A > a. 0,001 1525 3 1000 km 500 miles 614 — elektor june 1976 VHF FM receptio 4 n O O o field strength microvolts/meter o o in o in o o o in oo o o oo o in CM o o CM o in o o o in o n *7 I in CN in 0 ) O C ro CO • MM TJ (D 0 o o o o o o o c o o o o o o o o o o o in o • o • o in • o in CM o m CM »— . . o o o m m rM o o in o in cm o o o 05 o oo o CD o in o o 00 o CM microvolts/meter ro I in CN in gain in dB HF FM reception elektor june 1976 — 615 o o Lf) o LO o o o LO CO o o CO o LO CN o o CM o LO O O o LO in CN in E 0 ) o c 03 4 -» CO • T 3 Q) o o o CD O 00 O o to o LO o o CO o CM o o Cl ■c I in CN m 616 — elektor june 1976 VHF FM receptic Figure 4. The curves of figure 4A and 4B (land) and 4E, 4F (sea) give the field strength versus the distance for various heights of the transmitting aerial. In this it is assumed that the receiving aerial is at ground level (0 m). The curves relate to a radiation power of 1 kW. The curves of figure 4C, 4D (land) and 4G, 4H (sea) give the gain which is achieved for certain heights of the receiving aerial. The curves give reliable results for the frequency range of 70 MHz to 150 MHz. Figure 5. The value of the exponent 'n' in the empirical formula, as a function of fre- quency. In the early days of wireless it was be- lieved that radio waves of short wave- length (VHF) could only be transmitted over line of sight distances. It was thought that radio waves were rapidly attenuated once the receiver was below the optical horizon of the transmitter (figure 1). This is now known not to be the case. Due to differing electrical properties of the layers of the atmos- phere reflections and refractions of the radio waves occur, so that the waves follow a curved path. Radio waves can thus be received at reasonable signal strength even when the receiving aerial is below the optical horizon (figure 2). Figure 3 shows how the field strength of radio waves re- ceived by tropospheric reflection (shown dotted) are considerably greater than the field strengths of the ground waves at the same distance from the transmitter. The factors that determine if a usable signal can be received from a particular transmitter may be tabulated as follows: 1. transmitter output power (it is as- sumed that transmitters radiate omni- directionally); 2. the distance between the transmitter and the receiver aerials; 3. the height of the transmitting and re- ceiving aerials (obviously if the aerials are higher the transmitter and receiver can be further apart before the receiver falls below the ‘radio horizon’); 4. the gain of the receiving aerial (rela- tive to a simple dipole); 5. the minimum signal strength required by the receiver to produce a reasonably noise-free signal. Using a folded dipole aerial at 100 MHz the signal produced by the dipole is ap- proximately numerically the same as the field strength i.e. a dipole in a field strength of 1 fiW/m will produce a signal of 1 juV. The signal produced by other types of aerial can be obtained by multi- plying by the aerial gain (or adding the gain in dB). The curves of figure 4 may be used to establish the field strength that can be expected at a given distance from a transmitter. These curves are based on a transmitter power of 1 kW, but the field strengths with other transmitters can easily be found. This is best illus- trated by means of some examples. Example 1. The distance between the transmitting and receiving aerials is 200 km* on land, the transmitter power is 100 kW, the transmitter aerial height is 500 m*, and a folded dipole at a height of 15 m is used for the receiving aerial. What is the voltage expected at the tuner input? From figure 4B it can be seen that the expected field strength 200 km from a 1 kW transmitter with a 500 m high aerial is around 0.1 /iV/m. However, the transmitter power is actually 100 kW so the field strength is increased by 10 log ^-^(power ratio!) or 20 dB. Furthermore, the receiving aerial height is 15 m, and from figure 4C this gives a further increase of 20 dB. * 1 kilometer (km) = 0.62 miles; 1 meter (metre, m) = 39.37 inches. The expected field strength is thu 40 dB up on 0.1 fiV/m or 100 time (voltage ratio!) i.e. 10 /iV/m. Since th aerial is a simple dipole the signal inpu to the tuner (assuming negligible losse in the aerial downlead) is 10 /zV, whicl is adequate for most tuners. Example 2. The distance between th transmitter and receiver is 175 km am the path is across the sea. The trans mitter power is 100 W and the aeria height is 500 m. The receiving aerial i a 4 element Yagi array with a gain o 10 dB mounted at a height of 10 m. From figure 4F it is apparent that th received field from a 1 kW transmitte would be 1 /zV/m. However the trans mitter is only 100 W, so the fieh strength is reduced by —10 log - 10 dB. From figure 4G the aerial height o 10 m provides no additional increaS' (0 dB), so the output voltage from folded dipole would be 10 dB down oi 1 fjiV. However, the aerial provides ; gain of +10 dB, which cancels out thi 10 dB loss due to the reduced trans mitter power. The voltage at the inpu to the tuner is thus 1 /zV. Example 3. This may be useful fo would-be spies . . . The distance be tween transmitter and receiver i 150 km, the height of the transmittinj aerial is 50 meters, and the power i 100 watts. What type of receiving aerial must b< used to deliver at least 0.5 to the re ceiver? Figure 4B shows that the basic fielc strength from a 50 meter transmittinj VHF FM reception elektor june 1976 — 617 aerial at 150 km is 0.01 juV/ meter. Since a voltage of at least 0.5 fiW is required at the receiver input, the basic gain requirement is equal to 2° log ■ 34 IB. However, that’s for a 1 kW transmitter but the power in this case is only 1 00 W. Therefore the field strength is reduced by a further 10 log — 10 dB. Therefore, the receiving aerial will have to give a total gain of 44 dB (10 dB for the low transmitting power and 34 dB for distance). One possible way to achieve the 44 dB gain figure is to attach a folded dipole to a balloon floating at 250 meters. A more prac- tical (?) solution is a 6-element yagi array with a gain of 14 dB mounted at 50 meters. Example 4. This is of a more practical nature, and can be modified for indi- vidual circumstances. A 4-element aerial with a gain of 10 dB is mounted at a height of 20 meters, what receiving range can be expected for reasonably noise free FM stereo broadcasts? First, some assumptions must be made: most FM transmitters have power out- puts of 100 kW or more with aerial heights of over 200 meters; most re- ceivers need approximately 100 juV for noise-free stereo reception. Adding the basic receiving aerial gain (10 dB) to the gain due to receiving aerial height as derived from figure 4C (22 dB), a total aerial gain of 32 dB is found. Since the receiver requires 100 /iV, the 32 dB aerial gain means that a field strength of 2.5 £/V/m is re- quired at the aerial. If the transmitter power was 1 kW a look at figure 4B would show the distance. However, the transmitter power is 100 kW. This means that the field strength at the receiving aerial is in- creased by 10 log 100 1 = 20 dB for any given distance. 20 dB in field strength is a factor 10, so if a 100 kW transmitter gives the required 2.5 yu V/ m at a certain distance, a 1 kW transmitter would give a field strength of y~= 0.25 juV/m at the same distance. According to figure 4B, if the trans- mitting aerial height is 200 meters a 1 kW transmitter will give a field strength of 0.25 yuV/m (so the 100 kW transmitter will give 2.5 juV/m) at a distance of 1 50 km. Or, looking at it the other way round, using the 4-element aerial at 20 meters height with an ordinary receiver it should be possible to get good (stereo) reception from any transmitter within a range of 1 50 km over land. Of course, the results obtained from the graphs of figure 4 are true only where the transmission path is over the sea or over relatively flat terrain. If the re- ceiving aerial is on top of a hill then the received field strength will be greater since the effective height of tile aerial is greater. Conversely, if the receiving aerial is in a valley the field strength will be less since the effective height of the aerial is smaller and the aerial will be screened by the walls of the valley. Instead of using graphs, the received field strength may be calculated empiri- cally from the equation given below, which takes account of the terrain by including a term for the distance to the optical horizon (obviously the distance to the optical horizon is greater for an aerial on flat terrain than for one mounted at the bottom of a valley). The equation is as follows: E _ 88 • VP • h, • h 2 • Dh' 2 b XDn Where E is the r.m.s. value of the field strength in volts per metre; P is the effective radiated power in watts; hi is the height of the receiving aerial in metres; h 2 is the height of the transmitting aerial in metres; Dh is the distance of the optical horizon in metres (figure 1 ); X is the wavelength in metres; D is the distance between transmitter and receiver in metres; n is a frequency dependent exponent (see figure 5). Both the curves of figure 4 and the empirical equation apply only for nor- mal atmospheric conditions. During unusual weather conditions (such as inversion layers) VHF reception at distances of several thousand kilometres have been observed. This, of course, is not the norm, and is really of interest only to those interested in DX activities. M triac control In simple triac phase angle control the trigger circuit contains only R4, PI, Cl and the diac. Cl cannot fully discharge in that circuit, so on later half-waves the triac fires sooner. This gives a ‘snap-on’ effect called hysteresis. Adding resistor R2 and R3 and diodes D1 . . . D4 leads to equal starting con- ditions for every trigger-cycle. In this way, hysteresis effects are avoided. Triac control units are notorious for causing interference on radio and TV. The easiest way to eliminate this is to add the LC low-pass filter consisting of LI, C2, C3 and Rl. Coil LI should be placed in the neutral line. M BTW 10-400 618 — elektor june 1976 A.M. mains intercom Mains intercoms of a more or less reasonable quality are still a bit expensive on the market. Consequently, there appears to be a fair demand for a super simple and cheap a.m. intercom which despite a modest performance will be useful in certain applications where mains interference is not excessive. The type of mains intercom we are dis- cussing here has already become popular as a babyphone. The house mains wiring is used not only to power the two posts of the intercom, but also as the signal connection. Each post is therefore plugged in and the exasperating task of laying out and rolling up lines is elim- inated. Intercoms using the mains as their signal connection are, however, susceptible to mains-born interference. The unit described here uses amplitude modulation (a.m.), as do most of the commercially available units. This gives a reasonable compromise between sim- plicity and performance. It is worth noting here that a more complex design using frequency modulation (f.m.) to obtain high quality results will be published in a future issue of Elektor. Arrangement Every intercom post consists of a transmitter and a receiver. Figure 1 shows the block diagram of one such post. In the position ‘speak’ (or ‘trans- mit’) a simple oscillator produces a carrier which is amplified by an output stage. The resulting signal is amplitude modulated by the amplified microphone signal. The high frequency signal is then fed into the mains via a special trans- former. In the position ‘listen’ (or ‘receive’) the high frequency signal transmitted from another post is picked up from the same transformer (at point A) and fed to a high frequency preamplifier. From there it goes to a modest low frequency amplifier where it is brought to a level sufficient to drive a small loudspeaker. Switching between ‘speak’ and ‘listen’ is achieved by switching the supply volt- age between transmitter and receiver. Transmitter The transmitter for the intercom is shown in figure 2. The oscillator is a simple multivibrator built around T3 and T4. The output from the oscillator is fed to a class C output amplifier T6, via a buffer stage T5. The collector volt- age of this output stage is controlled by the amplifier T2/T7 which is adjusted to its maximum. This amplifier is in turn driven by the microphone amplifier Tl, whose gain may be varied by the poten- tiometer PI in order to vary the modu- lation depth. The final result is that an amplitude modulated high frequency signal is fed into the mains via trans- former Trl. Diodes D1 . . . D4 protect the output stage against voltage peaks at switch on. Capacitors C9 and CIO isolate the circuit from the mains. Receiver The receiver, which is very simple in design, is shown in figure 3. The trans- mitted signal is picked up at point A in the transmitter circuit and fed to the re- ceiver. The received signal is amplified considerably by the circuit around T8 and T9, and then detection takes place in the simple demodulator D7, D8, Cl 5. The automatic gain control (a.g.c.) cir- cuit formed by R18, D5, D6 is designed to operate only on very high input levels so protecting the listener from high level mains-born interference. The low frequency amplifier is simple but adequate. The output power is about 250 mW, which is sufficient for good intelligibility. The volume may be 4.M. mains intercom elektor june 1976 — 619 930 1-2 12V 2b0mA B = B40C400 Figure 1. Block diagram of one post of the mains intercom. Each post is a combination of an a.m. transmitter and a conventional 'direct' receiver. Figure 2. Circuit diagram of the a.m. trans- mitter. Either high or low impedance micro- phones may be used. Figure 3. Circuit diagram of the receiver and the l.f. amplifier. Figure 4. A suitable power supply for the intercom. SI switches between transmitter and receiver. controlled by P3. It is advisable to set P2 as low as practi- cable (i.e. so that the modulation is at the audible threshold) as a considerable number of components will then remain below the detection level hence limiting the interference on reception as much as possible. Conclusion The simplicity of this design gives more than a hint of the quality if its perform- ance. Since the transmitting power is fairly low (about 1 W) and interference suppression cannot reasonably be com- pared with that obtained with a narrow- band f.m. system, good performance can only be expected in a conventional one-family house. For a number of applications, this will be sufficient. Thanks to the low transmitting power though, the current consumption is very low, and a simple supply will do, figure 4 shows the circuit diagram of a suitable supply using an IC LI 30. The switch SI changes the supply between transmitting and receiving circuits. The circuit is not very critical and can therefore be built without too much difficulty. The only obstacle may be the transformer which must be wound. For the prototype, a potcore AL250 with a diameter of 18 mm was used. Various manufacturers (e.g. Siemens, ITT and Philips) supply these potcores in a range of versions and sizes. u 620 — elektor june 1976 FET front Specification Usable Frequency Range: 20 kHz-45 MHz Sensitivity: 4 mV r.m.s. at 20 MHz Input Impedance: 1 M in parallel with 5 pF Rise Time: approx. 5 ns Trigger Level: adjustable HF preamp and FET probe for frequency counter The frequency counter design published in the November 1975 issue of Elektor was accompanied by a design for an input preamplifier (Elektor 8, December 1975 p. 1235). Whilst this design gave good performance from 0-20 MHz it was decided that for r.f. work a preamp with a higher sensitivity would be useful, since it is here that signal levels are smallest. To avoid problems due to input cable capacitance the preamp is equipped with a FET input probe. To be of any practical use in the ma- jority of applications a frequency coun- ter must have a high input sensitivity and high input impedance. The fre- quency counter described in Elektor 7 has, in its basic form, the input connec- ted direct to the TTL logic circuitry, whose input impedance is low and asymmetric. In addition the input volt- age swing required to trigger the logic circuitry is of the order of 2 V. This is clearly not of much use except for per- forming measurements on other logic circuits. The preamp described in Elektor 8 had an input impedance of 1 M in parallel with a few picofarads and an input sensitivity of around 40 mV, rising to 100 mV at 20 MHz. It was felt that for r.f. use a higher input sensitivity was desirable, and it was decided that by sacrificing the low-frequency response (for reasons explained later) a high gain could be obtained with a simplified circuit. With the original design the preamp was mounted in the frequency counter case. However with this design there was a problem: reactive loading of the signal source by the capacitance of the input cable, which can be over 100 pF/m for coaxial cables. For this reason it was decided to split the new preamp design into two sections, a FET probe with a high input impedance and low output impedance capable of driving a coaxial cable, and a preamp, mounted in the case with the counter, to provide most of the gain. Design Targets The performance requirements for the preamp are similar to those given for the earlier design, except that higher input sensitivity is aimed at, while the low-frequency response is unimportant, © = 1,4 v Parts List to Figure 1 Resistors R 1 = 4k7 R2 = 1 k R3 = 2k7 R4 = 1 k8 R5 = 100 n R6,R9 = 220 n R7 = 68 n R8 =47 n R10 = 1 k5 R 1 1 ,R 1 3 = 470 n R 1 2 = 10 k PI = 1 k Capacitors Cl = 100 n C2 = 470 n C3,C5,C6,C1 1 = 10 n C4 = 47 m/6 V C7,C10 = 15 m/ 3 V C8 = 220 p C9 = 100 p Semiconductors T 1 ,T3 = BF494.BF1 94.BF1 95 T2 = BC557.BC1 57 FET front elektor june 1976 — 621 since the circuit is specifically intended for r.f. work. The requirements are tabulated below. 1. Bandwith. It was decided that the usable frequency range of the preamp should extend from the top end of the audio band to above the upper fre- quency limit of the counter ( 1 8 MHz). 2. Input sensitivity. It was decided that 10 mV was a useful value and could be obtained without resorting to complex circuitry. 3. Input impedance. This should be as high as possible i.e. input resistance should be high and input capacitance low. These design requirements are met, and in some cases exceeded, by the new design. The usable frequency range of the preamp + probe extends from below 20 kHz* to above 45 MHz. At 20 MHz the input sensitivity is around 4 mV, whilst at 45 MHz it is still only 17 mV, which is better than the l.f. sensitivity of the original design. These figures refer to the r.m.s. input voltage necess- ary to cause an output voltage swing that will reliably trigger the TTL Schmitt input of the frequency counter. The impedance of the probe input is 1 M in parallel with 5 pF. Preamp Circuit The preamp circuit is shown in figure 1 and consists of a simple, three-stage, direct coupled amplifier. To minimise the effect of transistor capacitances and stray circuit capacitance the resistor values around the circuit are kept low. As a consequence of this the low- frequency response is sacrificed, since to extend the l.f. response down into the * Frequencies below 20 kHz can be measured , provided the rise time is suf- ficiently short - less than 10 ps. audio band excessively large value electrolytics would have been required for C2 and C7 (especially C7). Quite apart from the size consideration the parasitic inductances of such large electrolytic capacitors can cause un- desirable resonances at higher fre- quencies. The preamp is provided with two in- puts, an a.c. input, which is normally fed from the probe output, and a d.c. input intended principally for low- frequency measurements on logic cir- cuits. At low frequencies the d.c. input sensitivity is compatible with TTL logic levels. The a.c. trigger level control PI sets the d.c. operating point of T3 and hence determines the input level at which it will turn on. With this control it is possible, when measuring complex waveforms, to trigger the frequency counter from either the fundamental or one of the harmonics as required. Figure 3 shows the effect of varying the trigger level control. The upper trace of each oscillograph is an 18 MHz input signal with a high 3rd harmonic content. The lower trace in each case is the out- put of the preamp, with different set- tings of the trigger level control. The graph of input sensitivity versus fre- quency for the preamp is given in fig- ure 2. This shows the input voltage (r.m.s. mV) required for a 4 V peak-to- peak output swing. As can be seen from Figure 1. Circuit diagram of the preamplifier. PI adjusts the triggering level. Figure 2. Graph showing the preamplifier sensitivity as a fuction of the input signal fre- quency. Figures 3a, b, c. Oscilloscope shows the effect of trigger level adjustment by control PI (see text). 622 — elektor june 1976 FET front LI ®= 1 V D1 =D2= 1 N4148 (§)= 3 V * see text ©=4,2V Parts List Resistors R1 = 1 k R2 = 1 M R3,R7 = 180 n R4 = 2k2 R5 = 3k3 R6 = 68 H R8 = 18 ft Capacitors Cl = 3n3 C2 = 10 m/ 6 V C3,C5 = 1 n C4 = 220 m/ 4 V C6 = 100 n C7 = 150 p Semiconductors D1,D2 = 1N4148 T1 = E300,BF245C T2 = BF494,BF194,BF195 Inductors LI = choke 100 mH the graph the required input voltage rises sharply above about 30 MHz, until at 45 MHz it is about 22 mV. Even so this is better than the original preamp design and is further improved by the addition of the probe, which also pro- vides some gain. Probe Circuit To reduce the cost and simplify the circuit it was decided to use a single FET as the input stage instead of the dual FET used in the original circuit. T 1 operates as a source follower to provide a high input impedance, with T2 providing a gain of about 2 and an output impedance of 68 £2 to drive the coaxial cable. Diodes D1 and D2 clamp the input voltage to ± 0.6 V maximum to protect the FET. The equalization network in the emitter of T2 helps to maintain a relatively flat frequency response, though of course a ‘ruler-flat’ response is not important in this application, provided the input sensitivity is adequate over the required frequency range. The FET used in this circuit is the tried and trusted Siliconix E300. Other FET’s, such as the 2N5397, 2N5398, BF245C and BF256C may also be suitable. If an alternative FET is used it should be selected for a zero gate volt- age drain current of at least 10 mA. R3 should then be selected to give a drain current of 3 to 5 mA with the device in the circuit. Frequency Response The gain of the probe circuit versus fre- quency is shown in figure 5 with the probe fed from a 50 £2 source. At the low frequency end the gain is about 2, and the response exhibits a slight rise up to about 60 MHz, after which it falls off. If the probe is fed from a high source impedance then the shunt capaci- tance of the probe impedance quickly attenuates the signal at higher fre- quencies. This is shown in the dashed curve for a source impedance of 10 k. When the probe is combined with the preamp the overall frequency response is as shown in figure 8. At 1 MHz the FET front elektor june 1976 — 623 Figure 4. Circuit diagram of AC FET probe. Figure 5. Fet probe gain as a function of fre- quency. The probe is terminated into 50 £7. Spectrum analyzer display shows frequencies from 100 kHz to 50 MHz (left to right). The input signal to the probe was 0 dB. The gain of the probe by itself isn't of much import- ance, because its main job is as an impedance match. Figure 6. Preamplifier p.c. board and com- ponent layout (EPS 9413). Figure 7. FET probe p.c. board and com- ponent layout (EPS 9427). Figure 8. Probe-plus-preamplifier sensitivity as a function of the input signal frequency. input sensitivity is around 2 mV, while at 45 MHz it is still only 17 mV, which is sufficient for many applications. Figure 9 shows an oscillograph of the preamp output to the probe at 40 MHz. The upper trace is the 40 MHz input signal (scale 20 mV/cm) whilst the lower trace is the preamp output (scale 1 V/cm). The timebase speed is 20 ns/ cm. Construction A p.c. board and component layout for the preamp is given in figure 6, and for the probe in figure 7. Normal r.f. prac- tice should be followed when mounting the components on the boards i.e. the component leads should be kept as short as possible, especially the transis- tor leads. The preamp may be mounted in the fre- quency counter case, and is connected to the FET probe by a length of 50- 75 £2 coaxial cable. The supply lead to the probe can be run alongside the cable. The housing for the probe is a matter of individual taste. The board is sufficiently small to fit in a small box folded from sheet aluminium. A test prod made from brass rod may be connected to the probe input through an insulating bush in the end of the box so that the probe may be used as a hand-held unit. The ground connection to the probe input may be made with a crocodile clip on a flying lead. Alternatively the probe input connections may be made using 4 mm sockets, as shown in figure 11. Short Hying leads ter- 624 — elektor june 1976 FET front missing link Figure 9. Probe-plus-preamplifier response at an input signal frequency of 40 MHz (see text). Figure 10. Excessive DC input signal levels should be cut down to approximately 0.5 V, since the maximum frequency limit is con- siderably lower for large signals than for small signal levels. Figure 11. The completed FET probe. minuting in test prods or crocodile clips may be then be plugged into these and connected to the circuit under test so that the user’s hands are left free. The unit could also be used with a prod made from 4 mm brass rod which would plug into the signal input socket. Power Supply Both the preamp and the FET probe can derive their supply from the +5 V rail in the frequency counter. The —5 V supply which was necessary with the original preamp design is not required. The total current consumption is around 25 mA. Operation In use the probe/preamp combination requires no adjustments except for the trigger level control and should function as soon as power is applied. In the event of a malfunction test point voltages are provided in figures 1 and 4 as an aid when faultfinding. THE SHIE Bli.ll Cumulative index of 'Missing Links'. The Link will appear each year in the June issue of Elektor. It contains an index to all Missing Links concerning articles published in the previous year. The intent of the link is to assist the home constructor by listing corrections and improvements to Elektor circuits in one easy to find place. A simple check of the Link will show whether any problems were associated with a project. Tunable Aerial Amplifier (El); February 75, page 229. Steam Whistle (El ), April 75 (E3), page 458. TV Sound (E2); June 75 (E4), page 660. CA 3090 AQ Stereo Decoder (E5); February 76 (E10), page 230. BC5 1 6/BC5 1 7 Transistor problems; March 76 (El 1 ), page 354. Diagram for the CA 3080, page 755 of E5, is incorrect; see September 75, page 952. TV Tennis (E7); January 76 (E9), page 148; May 76 (E13, page 508. Also for good ideas see March 76 (E 1 1 ), page 3 1 8. Lie Detector (E7); April 76 (El 2), page 454. TCA 730/740 (E8); January 76 (E9), page 148. Pre-amp for counter (E8); April 76 (El 2), page 454. Missing Links concerning articles in volume 2: Feedback PLL for FM (E9); February 76 (E10), page 230. Capacity Relay (E9); February 76 (E10), page 230. Digital Master Oscillator (E10); April 76 (El 2), page 454. Morse Typewriter (El 0); May 76 (E13),page 508. ejektor elektor june 1976 — 625 These pages offer our design staff — and, we hope, our readers! — a long wished for opportunity to eject more-or-less wild ideas. It often happens that promising ideas cannot be converted into practical cir- cuits. The problem may be lack of specialised knowledge, lack of equip- ment or even simply lack of time. Several examples of this kind of thing are still floating around the Elektor laboratories, such as a spot-sinewave generator and an OTA-gyrator. For the time being, development of these pro- jects was stopped after unexpected technical problems arose. We simply cannot afford to spend any more development time on these projects at the moment. Sometimes projects are put on ice at an even earlier stage, especially when it is obvious from the start that develop- ment will cost a disproportionate amount of lab time. In this case the design may never get past the block diagram stage, or it may be de- veloped bit by bit in the course of (several) years. An example of this type of thing is the one-line intercom. That basic idea dates back to 1971, but it is only now nearing completion. It is rather frustrating to see a recent Philips Press release describing a very similar arrangement developed at their research laboratories This means that our design staff are regularly coming up with interesting ideas that are only published several years later, if at all. Our readers also regularly submit circuits that contain an interesting idea, but are not (quite) suit- able for publication because of technical imperfections. Usually our editorial staff can add the final touches, but this, too, costs development time and man- power — which is not always available. Somehow, we want to eject these ideas. Somebody may be able to use them or carry on where the designer stuck. From now on, interesting ideas which can not (yet) be implemented in practi- cal circuits may be published in ‘Ejektor’. It is not the intention to use these pages for publishing ‘dud’ circuits; on the contrary, the intention is to publish interesting ideas. This may, of course, include circuits that look as if they should work but don’t. Also, some of the ideas may well prove completely impracticable on fundamental theoreti- cal grounds. If so, we hope that the reader who discovers this will let us know . . . Our editorial and design staff are quite enthousiastic about the new oppor- tunities offered. It should also be quite a challenge to our readers. We hope to be able at a later date to publish practical circuits based on the ideas presented here, after our readers or our design staff have found time to investigate them further. OTA gy rotor Comparison of the basic gyrator formulae with the basic OTA formulae shows that the OTA should be an ideal active device in gyrator circuits. Using an OTA gyrator it should be possible to construct a filter that can be swept through the whole audio band , while maintaining either constant bandwidth or constant Q. Such a filter could be used for spectrum analysis , electronic music (synthesiser!), equaliser, LF PLL, etc. The basic principles of the gyrator were discussed in a previous article (‘How to gyrate — and why’, Elektor 2, p. 255). It was shown that when a gyrator is used to simulate a parallel LC tuned circuit (figure 1), the following formulae apply: Q = VSgR, where: f Q = resonance frequency; g = gi = g 2 = gyration constant; Q = quality factor; C = Ci = C2 ; R= Ri = R 2 . Note that both resonance frequency and quality factor are linear functions of the gyration constant (g). The gyration constant is equal to the absolute value of the slope (or trans- 9051 — 3C 102E-1 626 — elektor june 1976 ejektor conductance) of the two amplifiers used in the gyrator, so for each amplifier: •out = ± g-Vin (the + sign for the non-inverting ampli- fier, the — sign for the inverting ampli- fier). Compare this with the basic OTA formula: •out = ± Sm- V in- The similarity is obvious! A bias current sets the value of g m . For a CA 3080, say, the transconductance equals: gm = 19.2 x IabC- This means that if OTAs are used in a gyrator circuit, the gyration constant is a linear function of the bias current OABC)- From this it follows that the resonance frequency must also be a linear function of the bias current. The quality factor will also be proportional to the bias current, provided the input impedance of the OTAs is large compared to R. If, however, the OTA input impedance determines the value of R, the quality factor will be almost constant over the whole band. This means that a simple DC adjust- ment will suffice to sweep such a filter over a 1000 : 1 frequency range (0.1 pA < I ABC ^ 1 00 At A), while main- taining either constant bandwidth or constant Q! A possible circuit is shown in figure 2. The only thing wrong with it is that it doesn't work properly . . . Provided the input signal level was kept very low, the filter worked as expected. At slightly higher signal levels, however, a sort of ‘lock-on’ effect occurred: the output level suddenly jumped to a much higher value, and maintained this higher value over quite a broad frequency range. Outside that range it would suddenly drop back again to the original low level. Literature: 4 How to gyrate - and why’: Elektor 2, p. 255; ‘ OTA V Elektor 6, p. 927. one octave lower Musicians nowadays tend to use more and more bass guitars, bass clarinets, and the like - as far as possible, that is. For this reason, they are faced with the problem of having to buy and carry around more and more (expensive) instruments. Electronic circuits that would transpose the sound of particular instruments down over one or two octaves would be a welcome relief. A = Compressor B = filter C = expander E103 Several factors determine the ‘sound’ of a particular instrument: wave shape, attack, decay, non-harmonic sounds (e.g. wind noises), changes of harmonic content as a function of amplitude, etc. For this reason it is not normally possible to retain the same ‘sound’ when the original signal is passed through a simple divide-by-two stage to transpose it over one octave. There arc, however, several possible approaches to the problem; the best approach for one instrument may be of no use for any other instrument. A few ideas will be given here — further development is left to electronic musicians or musical electronics engin- eers Guitar. It has been found that a simple divide-by-two circuit followed by suit- able filters can give quite reasonable results for a guitar. However, during the decay time the system tends to ‘stutter’ as the input signal drops below the trigger level of the divider, so this basic system is musically useless. It will there- fore be necessary to add a compressor stage in front of the divider to keep the input to this at a relatively constant level; an expander before or after the filters can restore the original amplitude relationships. A block diagram of this arrangement is shown in figure 1 . Trombone. Several brass instruments (and several string instruments as well!) have a spectrum consisting of both even and uneven harmonics with a gradually decreasing relative amplitude. To trans- pose an instrument of this type over one octave, it should be sufficient to add one ‘sub-harmonic’ to the original signal. This must be done in such a way that the original fundamental becomes the second harmonic of the new, added, fundamental — with the correct ampli- tude relationship. An advantage of this system (sketched in figure 2) is that the original ‘sound’ is retained to a very large extent. Clarinet. Several woodwinds, including the clarinet, have a spectrum consisting mainly of uneven harmonics with gradually decreasing relative amplitude. The even harmonics are at a much lower, and fairly constant, level of approximately —20 dB. To transpose the sound of this type of instrument it should be possible to use the same basic system as that described for the trombone. The difference is that in this case the new ‘sub-harmonic’ fundamental must be one-third of the frequency of the original fundamental. This means that a divide-by-three stage will have to be used, and that the instrument will be transposed over one octave plus one quint .... this could be a nuisance! ejektor elektor june 1976 — 627 R 1 filter without phase-shift A problem that occurs regularly in control systems using negative feedback is instability at high frequencies. If the system contains a non-minimum phase element , say, the total phase shift at high frequencies can beco??ie 360 ° while the total gain around the loop is still more than unity. If a low-pass filter is added inside the loop to reduce the gain to a safe level at high frequencies , the law of conservation of misery dictates that the point where the phase shift is 360 will also move down to a lower frequency - where the gain is still more than unity , in spite of the additional filter! What is needed is obviously either a completely new design , or else a filter that does not introduce phase shift in the frequency band that matters. The basic principle of a filter of this kind was published in an earlier issue of Elektor: the ‘Frequency dependent resistor' (Elektor 5). The circuit is repeated here (figure 1 ). It should be made clear that this is only a basic block diagram; it is also assumed for the present that the amplifiers have infinite gain and zero phase shift. In this case the total transfer function is: ’- 2 .( 4 - ) ! Vi JC or in which T = R 2 X C The input current is therefore: _ v i ~ v o v i — 2v i _ h * r R i R _v 0 _ R i Ri co 2 t 2 x v; which means that the input impedance is: z; = - 1 = Ric « • 2„2 This is a real resistance — with current and voltage in phase — but increasing with the square of the frequency. If the output is left open, this frequency dependent input resistance can be used to construct a filter without phase shift — the limitation being the fre- quency where the amplifiers start to introduce phase shift. It’s a fundamental law that there must be phase shift some- where! To give an example, if a signal is fed in via a resistor and the output is taken from the input of the frequency depen- dent resistor, the result is a high-pass filter without phase shift at the roll-off point. Literature: Elektor 5, p. 712. COMING SOON The next Elektor is the July/ August ‘Summer Circuits’ issue. It contains over 100 projects and design ideas, from con- trol units for solar heating panels to speech garblers. Some circuits are basic design ideas, such as a monoflop using a single 7400. Others come with p.c. board layouts and are complete functional units, an example is a pulse generator with vari- able pulse width and repetition rate. It should be made clear that this issue is not a review of circuits already pub- lished, nor is it a preview of designs that will be published in the coming year. — dark room timer — rain synthesiser — car clock — kettlestat — current source — battery indicator — antenna amplifier — logic tester — wind machine — min/max temperature indicator — digital contrast — power supply — SSB adapter — wideband frequency doubler — headphone adapter — pulse generator — over 84 other circuits 628 — elektor june 1976 vertical fet's pa: SWISS': WMm C . •Ivlv.v.v « ■ : : -: plete control of the signal from the driving amplifier. However, this is hardly ever the case in practice, since power transistors are not capable of following fast input signal variations, due to the non-linear dif- fusion capacitance between the base and emitter. This capacitance increases as the collector current increases. There exists a certain switching delay which is characterised by the transition fre- quency, fj. In addition, the phase dis- crepancy between the input and output signals, which is caused by charging and discharging this virtual capacitor, can be worsened by the output transistors clip- ping (although this clipping should be prevented in the preceding stage by limiting the driving swing to an ampli- tude that will not drive the final output signal against the power supply voltage). These transition imperfections cause a reduction in efficiency, i.e. an increase in the heat dissipated, so much so that, in some cases, it may be necessary to install a cooling fan. High fp transistors capable of following fast input signal variations at large ampli- tudes are more vulnerable than low t'X types under overload conditions, due to their construction and manufacturing technology. In particular, it is the second breakdown phenomenon, occur- ring at high collector voltages during high heat dissipation, that can lead to their complete destruction. This danger can be avoided by respecting the Io^CE diagram that shows the Safe Operating Area (SOAR), which indicates the safe combinations of collector voltage and current. It can be seen that for high fX types this area is considerably smaller than that for the more robust 2N3055 family, and good protective circuitry is badly needed. Transistors respond inversely with the emitter current, i.e. as the emitter cur- rent rises, the response falls off. A water tap provides a suitable metaphor: between the order to turn off the water and the completion of the action, the water continues to flow, with the volume of the wasted water being dependent on the number of turns that the tap was turned on. Unlike electronic valves and FETs, the base of a bipolar transistor always draws some current, which can be of considerable magnitude in the case of high power transistors. It follows from the above considerations that there is an urgent demand for an alternative high power active semi- conductor with improved characteristics and higher safe ratings for heat dissi- pation, current and voltage. There are, admittedly, improved transistors such as the low emitter concentration (LEC) types, but at the moment they are only suitable for small signal applications. The demand for transistors for large signal applications still remains. How- ever, it appears that it is in this region that the V-FET will be useful. Horizontal FETs Before discussing the new V-FET, it is worth mentioning some aspects of the There is always something new under the rising sun. The Japanese have now developed a semiconductor called Vertical field effect transistor', intended for use in high power output stages. The V-FET performance and basic characteristics are vastly superior to the common bipolar transistors. This article describes the V-FET's construction and operation, along with its application in commercial circuits. Circuit Requirements As unlikely as it may seem, a firm market seems to be developing for heavy, (i.e. 60-1 20 lbs) stereo output amplifiers capable of feeding some 1 50-350 watts to the 8 £2 load of each channel. Obviously, the quality of these ‘audio power houses’ is dependent on the characteristic properties of the active elements used in their output stages. Consider some of the difficulties which are associated with high output power. First, there is the problem of high heat dissipation in the final transis- tors. Then, the transistors themselves are subjected alternately to high col- lector potentials and currents, with the condition becoming more critical as the output power increases. It may even be necessary to arrange the output transis- tors in series-parallel, using the same principle as the coachman who replaces a single horse with a four-in-hand, so that the output is increased although the effort by each horse is less. Further considerations for the output circuit designer are the switching proper- ties of the transistors to be used. In the familiar class B final stage, the two halves alternate between a current- conducting and a current-blocking func- tion which is, ideally, under the com- vertical fet's elektor june 1976 — 629 1 VGS 1 1 1 f VGS = 0 V /"J r I 1 I I VGS : = _ 1 V 1 1 i # I VGS : = _2 V / f J / / VGS = _ 3 V f / W 4 / ► 9425-2 V DS Figure 1. Conventional (horizontal) FET con- struction. The hatched area around the gate indicates the depletion zone caused by the non-conducting state of the pn junction. With a constant Vqs, an increase in the drain cur- rent causes the depletion zone to grow as indicated by the dotted line, until it impinges on the edge of the substrate. At this point further increases in the drain potential Vqs fail to cause an increase in the drain current which may be regarded as saturated. Figure 2. Drain current Iq expressed as a function of Vqs with parameter Vqs for the FET of figure 1. The dotted line shows the knee potential as a function of Vqs- conventional (horizontal) FET, which is only suitable for low power applications such as circuits with high input im- pedance and low noise. Figure 1 shows the construction of an ordinary n-channel FET. A positive potential between drain and source causes elec- trons to flow from source to drain. The gate is made of p-type material, and when a negative potential with respect to the source is applied to it, the pn junction becomes non-conductive. Now the junction is surrounded by a depletion zone (hatched in the figure) which is completely empty of majority charge carriers. Figure 2 shows a family of output curves. For a given Vqs, the drain potential VdS and the drain cur- rent Id increase linearly at first. The current increase extends the depletion zone until the point at which the zone touches the opposite edge of the sub- strate (shown by the dotted line in fig- ure 1). Any further increase in Vqs fails to increase id, so the FET virtually behaves as a constant current source for any higher values of Vds* The output curves also show that more negative values of Vqs cause the ‘knee’ to occur at a lower Vqs so reducing the corre- sponding saturation current. The pos- ition of the knee for varying Vqs is shown by the dotted line. Vertical FETs Both Sony and Yamaha have now devel- oped high power FETs whose properties are very promising. The construction and manufacturing technology of the Sony and Yamaha devices are similar in that both may be considered to be made of a large number of ‘mini-FETs’ work- ing in parallel, but in other respects the two devices are quite different. Sony have developed both p-channel and n-channel V-FETs, whereas Yamaha have only developed an n-channel type. Consequently, the circuits utilising these new devices differ considerably in their design, as will be discussed later. Figures 3 and 4 show the construction of the Yamaha n-channel device. Com- pared with the horizontal junction FET, the current flows vertically. The drain is located at the bottom of the crystal and its mechanical connection to the casing has a very low thermal resistance, which is vital since practically all the heat produced inside the device is developed in the drain and must be led away from there. The channel is made of N" type material into which a grating of P + type material, the gate, has been embedded. In figure 4, each square represents a separate FET, at 5 to 10 micron spacings. The entire chip size is about 5 mm by 5.5 mm and it consists of tens of thousands of FETs, all working in parallel. The output characteristics of these FETs are shown in figure 5, and a major dif- ference from those in figure 2 is im- mediately obvious: there is no knee, or corresponding saturation voltage. Readers once familiar with the output curves of the old faithful triode will no doubt be struck by the resemblance, which has already been used in the publicity given to these new devices, as nostalgia is a great selling point. Those readers will remember the one-upman- ship in the triode-fitted amplifiers against the pentode-equipped counter- parts with their inherent current- saturation effects. However, the com- parison is not really fair, as modern cir- cuits with high negative feedback (made possible by the elimination of the out- put transformer) display hardly any dis- tortion of this kind. On the other hand, it is much easier to trim an amplifier which is inherently devoid of clipping and other nasties, than one which is full of these distortions. V-FETs present other advantages than merely being free of these annoyances. The family of output curves in figure 5 may also be used to describe the oper- 630 — elektor june 1976 vertical fet's 3 , 4 ation of p-channel V-FETs. In this case the drain current consists of a stream of holes (positive charge carriers) flowing vertically from source to drain. Vpg is therefore negative. The gate is made of n-type material and is positively biased with respect to the drain. Figure 6 shows the transfer character- istics for an n-channel V-FET. Again, a close resemblance to the triode transfer curve may be recognised. The curve slopes much more gently near zero drain current than the steep exponential curve peculiar to the conventional bipolar transistor. This property of the V-FET is favourable for rounding the cross-over point in class B power stages. It must be admitted that the slope of the transfer characteristic can be adversely affected by fluctuations in the supply voltage or in the drain-source potential (such as would be caused by a drain load im- pedance). This is due to the negative feedback acting on the drain potential. The equivalent effect with thermionic triodes is the difference in transfer characteristics between the dynamic and the no-load output. Performance Figures The data sheet gives some interesting figures for different types of V-FETs. Particularly impressive are the maxi- mum ratings for the Yamaha 2SK77 type, they are not likely to be equalled by any conventional power transistor. All types permit a very high maximum drain-to-gate potential, which is the highest potential found in a V-FET. All types are completely free from second- ary breakdown effects, since the density of the drain current is the same through- out the channel. This is due to the absence of current crowding and also to the manufacturing technology which permits an extremely low level of con- tamination in the n- (or p-) channel. Although the V-FET is basically a Figures 3 and 4. Construction of the Yamaha developed n-channel V-FET, which may be considered as a large number of FETs working in parallel. The chip size is about 5 by 5.5 mm for the Yamaha 2SK77 type, and about 3 by 3 mm for the Sony 2SK60 and 2SJ18 types. Maximum dissipation rating is mainly depen- dent on the chip surface area, which gives rise to 200 watts for the 2SK77 and 63 watts for the 2SK60 and 2SJ18. Figure 5. Drain current Ip as a function of the drain-to-source potential Vqs with gate- to-source potential Vqs as parameter. These output characteristics for the 2SK77 V-FET show similarity to thermionic triode charac- teristics. Figure 6. Transfer characteristics for the 2SK77 V-FET : drain current Ip as a function of gate-to-source potential Vpg with drain-to- source potential Vpg as parameter. Any load in the drain circuit causes the slope of the dynamic transfer characteristic to drop, with the exception of the curved portion near the cut-off point. These characteristics show that the optimum quiescent current (i.e. a working point where the slope has half the gradient of the full swing slope for a class B stage) should be about 400 mA. Figures 7 and 8. Stripped version (figure 7) and complete circuit diagram for one stereo channel of the Yamaha B-l output amplifier. This circuit clearly resembles the direct- coupled output circuit for thermionic valves and 800 loudspeakers. voltage-controlled device, the gate does, nevertheless, draw some current, as the input impedance is not quite so high as in the horizontal FET. This input cur- rent consists of an inherent leakage cur- rent through the barrier between gate and channel, and a capacitive current caused by the charge and discharge of the source-gate capacitance and the virtual capacitance due to the Miller effect on the gate. Consequently, the currents required to drive the 2SK77 are so high that a source-follower driving stage is needed (see figures 7 and 8), for which the type 2SK75 has been devel- oped (see table). In spite of this, V-FETs offer some important advantages over conventional power transistors. For one thing, the input capacitance is smaller and almost independent of the drain current. For another, the transition speed of all these V-FETs is 5 to 10 times faster and the power switched at these frequencies is 2 or 3 times higher than for the fastest bipolar tran- sistor. High frequency distortion at the cross-over point is practically non- existent, especially with optimum set- ting of the quiescent current in the output stage. An exclusive and very recommendable V-FET property is the negative tempera- ture coefficient of the drain current, i.e. the current decreases as the crystal temperature increases. There is, there- fore, no risk of thermal runaway in class B power stages in contrast to cir- cuits with conventional transistors where, if there is insufficient thermal stabilisation of the standing current, the current rises as the temperature does, the temperature rises as the current does, which ever-increasing circle is only terminated when the transistors give up the ghost. Thanks to the absence of this cumulative effect, V-FETs need no preventative measures against thermal ru naway. vertical fet's elektor june 1976 — 631 < z UJ 3 o' D u UJ □ oc 3 O oo v ds -drain SOURCE VOLTAGE- [VJ 9425 5 TYPICAL TRANSFER CHARACTERISTICS V GS -GATE SOURCE VOLTAGE- (V] 94 25-6 s < a: Q I O V-FET Circuitry in Practice Two commercial circuits featuring V-FETs are discussed here, namely the Yamaha B-I type, which is a separate final stage, and the Sony TA 8650, which is an integrated amplifier con- taining both the pre-amplifier and the output stage. The discussion is confined to the final stages; power supply, filter and protective circuitry are not treated (the Yamaha circuit has in total 39 FETs, 1 13 transistors, 3 LEDs, 64 diodes and 7 zeners!). Both circuits feature class B output amplifiers, and the dynamic transfer characteristics involve a relatively high quiescent current (400 mA for both). This, in combination with the high supply vol- age, results in fairly high quiescent heat dissipation (64 watts per channel in the B-I) in the final stage, but this is no problem for these V-FETs. The tempera- ture protection in the B-I was intended originally to safeguard other com- ponents, for example the expensive computer-quality electrolytic buffer capacitors (rated at 80°C). A condition not found in conventional transistor circuitry, and only revealed on close inspection of figure 5, is that the drain current will rise out of all proportion if power is supplied to the drain without sufficient bias (positive for n-channels, negative for p-channels) on the gate. Even for short durations, an excessive drain current of this magni- tude will endanger not only the V-FET itself, but also the associated series resistors and probably the power unit as well. This condition only occurs at the moment of switching the power on or off. To prevent this, the circuit must be designed such that at switch on, the power is applied first to the pre- amplifying stages, so enabling the gate bias to build up before power is applied to the final stage; conversely, at switch off, the power is removed from the final stage before the pre-amplifier. The requirement for a bias voltage of opposite polarity to the drain voltage calls for more than one stabilised (or unstabilised) power supply; if they were of the same polarity, some of the drain supply which is vital to obtain the full output voltage swing, would be diverted. To avoid this, the driver stage is supplied with a separate power supply. Yamaha B-I Amplifier This stereo output amplifier will con- tinuously deliver 160 watts per channel into an 8 £2 load, with the harmonic and intermodulation distortions remaining well below 0.1%. Figure 8 gives the circuit diagram for one channel of the JTT TR504 TR505 TR518 d y X TR506 @1 TR507 TR512 I I TR513 R, I I JL TR510 X_i X TR511 TR514 -©40V TRa ■©85 V TRb I R -©85V -0200V 9425-7 632 — elektor june 1976 vertical fet's output stage, but the principles of oper- ation can best be explained from the stripped diagram in figure 7. The basic design of the circuit may now be seen to bear some resemblance to the once- familiar, single-ended, push-pull power stage featuring two pentodes (EL 86) feeding an 800 12 loudspeaker. This resemblance is due to the circuit being equipped in all signal handling stages with FETs and the single polarity n-channel V-FETs. In this Yamaha design, the power pair consists of two Darlingtons (TR518, TRa and TR514, TRb) series connected as far as DC is concerned, and with the load connected to the drains of the first pair and the sources of the second. The arrangement requires the two halves of the final stage to be driven in push pull, with the lower half being driven by the voltage between the TR514 gate and the —85 V supply rail, and the upper half by the voltage between the TR5 1 8 gate and the load. Ra and Rg are coupling resistors from the penultimate stage which pro- vides the phase inversion. Bootstrapping makes the drain impedance of the TR512 appreciably higher than that of the TR513. The V-FETs TR518 and TR514 (type 2SK75) in the output stage are arranged as source followers; both are tied to the —200 V rail via constant current cir- cuits. This not only stabilises the quiesc- ent current for these drivers, but also applies the entire output current swing from TR5 1 8 and TR5 14 to the gates of TRa and TRb respectively. The B-I pre-amplifier stages are com- posed of three differential amplifiers in cascade with long tail pairs consisting of constant current circuits. The first stage is an n-type dual FET TR501 . The input signal is applied to the left hand gate while the right hand gate receives a fraction of the output signal, so pro- vidingnegative feedback (since the input and output are in antiphase). A dual FET with both elements on the same chip is necessary here since any differ- ence in the DC characteristics of the two would produce an offset voltage at the output. Each half of the second dif- ferential stage consists of two cascode connected P-FETs (TR504, TR506 and TR505, TR507). The two cascode output FETs (TR506 and TR507) oper- ate in a common gate configuration resulting in the almost complete elimin- ation of capacitive feedback due to the Miller effect. A further advantage is that linearity is guaranteed over a consider- able drain potential swing. This is par- ticularly advantageous in the third differential stage, which is also cascoded and built around the N-FETs TR510, TR512 and TR51 1, TR513. It operates as a phase inverter for the output stage. The details in figure 8 need a little more explanation. Control VR501 is used to set the DC offset in the output stage to zero. Control VR504, which together with R536 forms the Ra resistor of figure 7, adjusts the balance between the upper and lower halves of the out- put stage. The quiescent currents for the output FETs TRa and TRb are adjusted via the constant current circuit TR508 in the third differential stage, whereas the quiescent current for the driven TR518 and TR514 is stabilised by the constant current circuits TR515 and TR517. Variation of the direct currents through TR510, TR512 and TR511. TR513 causes the TRa and TRb bias tc vary, and with it the drain currents. TR508 is fed by a portion of the poten- tial across zener diode D503, and the controls VR502 and VR503 are used tc set the quiescent current. The base and emitter of TR5 1 9 are strapped via series resistors between the positive and nega- tive power rails (±85 V): the collector is connected to the quiescent current cir- cuit. This circuit makes the TRa and TRb quiescent currents practically inde- pendent of power supply (for the final stage) fluctuations. Sony TA 8650 The Sony circuit differs from the Yamaha in the basic design concepts, as it has recourse to both n-channel and p-channel V-FETs. It has therefore beer possible to make the output stages com- pletely complementary, the same as when conventional transistor power stages are being used. The TA 8650. featuring this complementary principle, is capable of delivering 80 watts into each 8 £2 loudspeaker with barely measurable distortion. Figure 9 shows the essentials (to a level suitable for this discussion) of the cir- cuit of one output amplifier. The final stage, which is a class B amplifier, uses three n-channel V-FETs (type 2SK60) in parallel for the positive half (T7 17 . . . T7 1 9 in the figure), and three vertical fet's elektor june 1976 — 633 p-channel (type 2SJ18, T720 . . . T722 in the figure) for the negative half. In this configuration they form a com- plementary source follower feeding the common load. The output stage is powered by ±60 volts. The parallel con- nection enables each half of the output stage to dissipate about 190 watts. The V-FETs in the positive half are biased negatively, while the negative half Figure 9. Simplified circuit diagram of one channel of the output stage in the Sony TA 8650 integrated amplifier. This stage is a complementary source follower with an inherently low internal impedance, even with- out negative feedback. The six output FETs must be selected to have a very small spread in the transfer characteristics between them, since the current distribution in the three parallel outputs in each half and in the two halves themselves must be equal. Table. Characteristics of specific V-FETs 2SK75 n-channel Yamaha 2SK77 n-channel Yamaha 2SK60 n-channel Sony 2SJ18 p-channel Sony maximum dissipation at 25° C Pd 20 W 200 W 63 W 63 W maximum crystal temperature T i 1 50° C 1 50° C 1 20° C 1 20° C maximum drain-gate voltage v DGO 200 V 200 V 170 V -170 V maximum gate-source voltage v GSO -30 V -40 V -30 V to -50 V 30 to 50 V maximum •d 0.5 A drain current 20 A 5 A -5 A maximum •g 10 mA gate current 1 A 0.5 A -0.5 A gain M 40* 7.5** A * # # 4* * * slope internal s 30 mA/V* 1.5 A/V** 0.25 A/V*** 0.25 A/V*** resistance Rj 1k3* 5** 16*** 16*** * with V ds = 80 V and l D = 10 mA ** with V DS = 30 V and l D = 2 A ** with |V DS | = 20 V and |I D | = 1 A has a positive bias. In the no signal state, the potential at the junction of the sources is zero. For this reason the respective gates are connected crosswise across resistor R737. The necessity of providing these gate-to-source bias po- tentials calls for a voltage for the pen- ultimate stage (T711 . . . T7 16) exceed- ing that for the final stage. The required supply used in this case is ±85 volts. The final stage is driven from a low impedance circuit: the emitter followers of T715 and T716. The sum of the drain-to-source bias potentials for the output FETs appears across the resistor R737. The emitter followers T715 and T716 are powered via the constant cur- rent circuits T713 and T714, which are in turn fed via a diode-resistor network from the ±85 V supply rails. Control RT701 is used to set the potential dif- ference across R737 and thereby the quiescent current for the output stage. This complex circuitry eliminates the effect of power supply fluctuations upon this quiescent current. N.B. The six output FETs are selected for accurate equality of |Vqs| at a constant drain current, which is impera- tive for uniform current and dissipation distribution among the six power FETs. Unfortunately, this means that a failure of any FET in the final stage will mean replacing all six. The driver stage T715, T716 is sym- metrically controlled via resistors R735 and R736, by the pre-amplifier. The latter is composed of three differential stages, the first of which is designed as an n-type dual FET T601 operating on 634 — elektor june 1976 vertical fet's led light show the difference between the input signal and an inverse feedback signal derived from the output circuit via the potential divider of R608 and R616. This first stage feeds the amplified signal to the second stage, T602, T603. The T607 collector circuit in the third of these dif- ferential stages includes a current mirror T604 fed from the collector of the other transistor T606 in the third stage. This circuitry guarantees symmetrical control of the driver and hence the output stage. Conclusion The object of this article has been to provide our readers with some insight into the properties and possible appli- cations of these new beasts. The high quality of the commercial circuitry described goes without question, especially as regards the absence of cross-over distortion (although this has been achieved at the cost of a quiescent dissipation that equals the maximum output power of some small conven- tional power amplifiers!) Despite the impressive figures the rela- tionship between quality and price is not in all cases more favourable than that for equipment using conventional bipolar transistors with a few FETs thrown in. In this respect, it is worth noting that Sony offers two interesting alternatives in the shape of amplifiers TA 4650 and TA 5650, which are con- siderably lower in price and have an output power only 2 or 3 dB down. H LEDs are now finding applications in more and more varied fields and here is a unit to introduce them to the world of discos and the trendy scene. Small is beautiful, as they say, and the LED light show can brighten up the front panel of an amplifier, tuner or tape recorder. The light show described here is fairly inexpensive and simple to build, and is distinguishable from conventional light shows, not only by its use of LEDs (light-emitting diodes), but also by its filtering system and the lack of a (hazardous) high voltage. It is intended as a decorative addition to an audio set- up by installing the LED light show in the front panel (say) of the amplifier or tuner, perhaps. This will give an attract- ive (even if somewhat miniaturised) visual rendering of the music; the dis- play is enhanced by the use of three different colours. Design philosophy Consider the requirements of a light show: its task is to accentuate audio information with a visual display. The audio information will be from various sources (e.g. musical instruments), each of which has a characteristic frequency spectrum and amplitude. Rather than attempting to respond to the complete frequency spectra, the light show is designed to select a number of fre- quencies which are representative of the incoming signal. As for the problem of varying amplitudes, the light show uses dynamic compression of the audio sig- nal. If the visual information possessed the same dynamic range as the music, the difference between minimum and maximum light intensity would be annoying. Another consideration is that visual selectivity is much less than acoustic selectivity, so the frequency bands chosen for visual representation are deliberately restricted. It is gener- ally accepted that audio frequencies are divided into three groups for dis- play: low, medium and high. However the boundaries which lie between these groups are not so well defined. Many applications leave overlaps between these groups, but in this case it has been found that leaving gaps gives a better solution (see figure 1). Figure 2 is a block diagram of the light show. The compressor which is used tc reduce the dynamic range of the audic signal can be of a much lower quality than one for use in an audio circuit. The only restriction is that the harmonic distortion produced by (say) the middle channel should not be visible on the high channel. The characteristic of the compressor should not be such as tc amplify noise or hum which would then be displayed during quieter passages Even so, the visual effect is still im- paired by peaks, but since the response of the amplifiers A1 ... A3 is unimport- ant, they are designed to limit on a fairly low input signal to mask non- linearities in the compressor. The filters feeding A1 ... A3 are low-pass, band pass and high-pass, respectively. In the circuit described here, use is made ol double-T filters. This facilitates the early-limiting design of the amplifiers The double-T filters are of simple design and give good selectivity, bul they have the disadvantage of beinfc peaked. This can be remedied b> damping the filters, bearing in mine ed light show elektor june 1976 — 635 the restriction, still, that the frequency Dands must not overlap. [n designing the amplifiers A1 ... A3 it nust be remembered that the pro- gramme material upon which this type }f device is most commonly used con- :ains a predominance of the low fre- quencies. The gains of A1 ... A3 must therefore be such as to overcome this disparity. rhe point at which the light show is ;onnected into the audio circuit can /ary between a pre-amplifier and a 100 W output amplifier. The com- oressor should not only be resistant to the very high levels which may well occur, but must also continue to func- tion normally when they do. In addition, the input impedance must be suf- ficiently high so that the circuit to which it is connected is not loaded by it. [n most light shows, the loads of the amplifiers A1 ... A3 are driven by triacs or thyristors, either via an iso- ating transformer or an optocoupler. Mo relationship between light intensity and amplitude can be obtained, since the condition controlling the lamp state ts binary, i.e. the lamp is on if the ampli- fier output is above the trigger voltage, and off if it is not. How the relationship is obtained in this particular design is explained shortly. The Circuit Figure 3 gives the circuit diagram for the light show. The compressor is formed by the circuit around IC1 (741), which is connected as a non-inverting amplifier whose gain may be varied from 20 dB to 60 dB by means of PI. When the emitter voltage of T1 becomes greater than +6.5 V D1 and D2 start conducting, thus reducing the input signal and hence the output signal until equilibrium is restored. Since the time constants in the control circuit are fairly large, it takes some time for equilibrium to be restored. Asymmetri- :al compressors require a fairly slow control rate because otherwise motor- boating (low frequency oscillations) could easily occur. rhe output of the compressor stage is fed to the three double-T filters, as »hown. These filters have enough output current to drive the LEDs through 220 £2 resistors, allowing the relation- ship between amplitude and light ntensity to be retained. Construction and Operating rhe circuit should preferably be given a netal housing, and a screened lead should be used between the input socket and the compressor input, with :he screening connected to the housing it the input socket. The connections Tom the filters to the LEDs can be made with ordinary connecting wire. \fter construction and careful inspec- :ion of the completed circuit, the light show is set up in the following way: idjust PI to give maximum gain (slider igainst R4) and then P2 . . . P4 (sliders igainst C5). Connect the compressor input to the loudspeaker or to the pre- amplifier output so that the audio signal is fed to the light show (it is better to avoid classical and choral music while setting up). Now adjust PI . . . P4 so that the light display gives the best (subjective) match with the music. The setting is a matter of personal taste and it may be found that taste demands different settings for different types of music. If instabilities occur in any of the channels, a lower value must be selected for the corresponding resistor, R 13, R 19 or R25. Figure 1. Two ways of dividing frequencies into groups or channels. The light show gives better optical effects if the second method of leaving gaps between the groups is em- ployed. Figure 2. Block diagram of the light show which at this level of detail appears the same as a conventional one, consisting of com- pressor, channel filters and channel amplifiers. A printed circuit board and component layout are shown in figure 4. Compres- sor % % 2 % 3 9403-2 636 — elektor june 1976 led light shov 01,02,04,06,08 = 1 N4148 ct a i 33k C2 560n lOOn D1 02 R3 C3 C4 100/i 10V C7 '470k 47/J JOV / 9325-1 < 2 > digibell elektor june 1976 — 639 numbers and converted into binary code (since this is what the counter will be programmed with). This results in the following table: Note Decimal Binary c' 90 0101 1010 b 96 01 100000 a 108 01 101100 g 120 01 1 11000 f 135 1000011 1 e 144 10010000 d 160 10100000 c 180 10110100 B 192 1 1000000 A 216 11011000 G 240 1 1 1 10000 It is evident, from the above table, that if the programmable counter is set to count to 90 and is fed with a clock fre- quency 90 times that of c' then the out- put frequency will be c\ If it is set to count to 180 and is fed with the same clock frequency then the output will be c, one octave below c'. I bis is how each note is synthesised from a single clock frequency. Since each note bears a fixed frequency ratio to all the other notes it is obvious that the only tuning necessary is to adjust the clock fre- quency until the melody is in the re- quired key. The Westminster Chime uses only the notes G, c, d and e in the sequence e, c, d, G, G,d, e, c, so at first sight it appears that division ratios of 240, 1 80, 1 60 and 144 are required. By coincidence how- ever, it happens that these numbers are all divisible by four, so the division ratios can be reduced to 60, 45, 40 and 36. This means that the programmable counter can be of shorter length, and that the programming is simplified. Figure 1. Circuit diagram for Digibell. Figure 2. Truth table for the counter program. Figure 3. Timing Diagram which shows se- quencing of the Digibell. In addition to getting the notes right it is also important to achieve the correct tempo. T he first three notes each have a duration of one crotchet, while the fourth note has a duration of one minim (two crotchets). This is followed by a rest of one minim duration. The fifth, sixth and seventh notes are each of one crotchet duration, while the final note has a duration of one minim. The total duration of the tune is thus 1 1 crotchets. When designing the circuit that per- forms the sequencing this must be taken into account. The Circuit The circuit of the Digibell is given in figure 1. The clock pulse generator consists of two NAND gates N 1 and N2. They are connected as an astable multi- vibrator whose frequency and duty- cycle can be adjusted with PI and P2. The programmable counter consists of two presettable up-down counters type 74193. These are connected to count down from a preset number to zero, the number being loaded into the data inputs Al-Dl, A2-D2 before the start of each count. The operational sequence for the presettable counter is as follows: initially the borrow output of IC6 is low. This lakes the load inputs of IC5 and IC6 low, so the data on the inputs A1-D2 is loaded and the count com- mences. During the count the borrow output is high, but when the count reaches zero it again goes low, the data is reloaded, the count recommences and so on. Since the borrow output is low for only a small proportion of each count the output waveform is very asymmetric and is not suitable for use as an audio tone. For this reason FF1 is connected to the borrow output and produces a square wave with a 1:1 mark-space ratio (50% duty-cycle) at half the fre- quency (i.e. one octave below) the bor- row output. To produce the Westminster Chime melody the programming numbers cor- responding to the four required notes must be fed to the data inputs of the presettable counter in the correct se- quence. This is controlled by a second counter (type 7490), and a 7442 BCD- to-decimal decoder. When the bell-push SI is pressed the Q output of FF2 goes high, enabling the astable multivibrator N6/N7, which feeds clock pulses at about a 2 Hz rate into the A input of IC2. These are counted by the 7490 and the BCD out- puts of the 7490 are decoded by the 7442. The 10 outputs of the 7442 go low in turn, at each step feeding a different number into the data inputs of IC5 and IC6 via the encoder compris- ing N3, N4 and D3 to D6. (Note that the 7442 has active low outputs, i.e. out- puts are normally high and go low when enabled by the appropriate input code). On the tenth clock pulse the D output of IC2 goes low, clocking FF2 back to its original state (Q output low) until the next time the bellpush is pressed. The correct tempo of the melody is achieved in the following manner. If each note were sustained until the next clock pulse occurred then the notes would simply slur into one another without a pause. This is avoided, and the rest in the middle of the tune is included, by means of N5 and diodes Dl, D2, D7 and D8. At the start of the sequence the 0 out- put of IC3 i low so FF1 is held in the COUNT IC3 OUTPUTS IC5, IC6 DATA INPUTS BINARY DECIMAL NOTE 01 23456789 D2C2 B2 A2D1 Cl B1 A1 0 0111111111 0 0 1 0 1 1 0 0 44* — 1 1011111111 0 0 1 0 0 1 0 0 36 e 2 1101111111 0 0 1 0 1 1 0 1 45 c 3 1110111111 0 0 1 0 1 0 0 0 40 d 4 1111011111 0 0 1 1 1 1 0 0 60 G 5 1111101111 0 0 1 0 1 1 0 0 44* — 6 1111110111 0 0 1 1 1 1 0 0 60 G 7 1111111011 0 0 1 0 1 0 0 0 40 d 8 1111111101 0 0 1 0 0 1 0 0 36 e 9 1111111110 0 0 1 0 1 1 0 1 45 c * Not important as output of FF1 disabled during rest 3 clock 1 1 1 i i i i l 1 1 1 “i r “1 1 — 1 1 ~l 1 — 1 clear FF1 T L J 1 J FF1 1 1 L _J L_ _l | J L_ 1 1 1 1 1 1 Q FF1 mm 1^1 r j | l.y.w.vl Kw.v.vJ 1 .•#•••••••• vl r.v.v.v.l — J | |V.V.V.V.V.V.V.V| r-y v.wv.vi • -J mXil 9325-3 640 — elektor june 1976 digibef clear state and there is no output. During notes 1, 2 and 3, pin 12 of N5 is high and pin 13 is switched alternately high and low by the out- put of N6. The output of N5 thus gates the J input of FF1 so that there is an output only when the clock pulse (output of N6) is low. The first three notes thus have a duration of half a clock pulse. On the fourth note pin 12 of N5 goes low, so the output remains high what- ever N6 does. The J input of FF1 is thus high and the fourth note has a duration of one clock pulse. On the fifth step output 5 of the 7442 goes low, so FF1 is held in the clear state via D1 and there is no output. This is the rest. The next three notes are all of half a clock pulse duration, but on the final note pin 13 of N5 is held low, via D7, the output of N5 holds the J input of FF1 high and this note has a duration of one clock pulse. To make the sequence of operations clearer, a truth table for the counter programming and a timing diagram for the sequencing are given in figures 2 and 3. Figure 4 shows a p.c. board and component layout for the Digibell. For use as a doorbell the output of FF 1 must be amplified to a level suitable to drive a loudspeaker. The output /I Parts list for figure 1 Resistors: R1 ,R2 = 2k7 R3 = 68 k R4 = 15 k R5,R6 = 1 50 12 PI ,P2 = 2k2 adjustable Capacitors: Cl ,C2 = 1 n5 C3,C4 = 1 00 m/6 V Semi-conductors: D1 . . . D8 = DUG IC1JC4 = 7400 IC2 = 7490 IC3 = 7442 IC5JC6 = 74193 IC7 = 7473 SI = Push button switch Figure 4. p.c. board (EPS 9325) and com- ponent layout. Figure 5. A possible amplifier for use with the Digibell. 9325-5 attenuator R3/R4 may or may not bt required, depending on the sensitivity of the amplifier used, or they may bt replaced by a potentiometer of betweer 10 k and 100 k to provide a volume control. market elektor june 1976 — 641 Divide-by-Four Gigahertz counter from Motorola Semiconductors Communications engineers will be interested in a new integrated circuit from Motorola known as the MCI 699 dividc-by-four gigahertz counter. This is a very high speed device for prescalcr applications. The clock input requires an a.c.- coupled driving signal of 160 mVpp amplitude (typical). A sine-wave signal is acceptable for frequencies from 50 MHz to 1.2 GHz. Below 50 MHz wave- shaping is recommended. With pulses which have good rise and fall times (in the order of 1 to 2 nsec), the MC1699 has no lower limit on clock frequency. The clock toggles two divide-by- two stages and the complemen- tary outputs (50% duty cycle) are taken from the second stage. DIVIDE -BY- FOUR GIGAHERTZ COUNTER V'm + 2V d.c. Vout COUNT FREQUENCY TEST CIRCUIT Ins/div OPERATING CHARACTERISTICS AT 1-2GHz ins/div OPERATING CHARACTERISTICS AT 1-5GHz The MCI 699 includes clock enable and reset inputs both compatible with MECL1 1 1 volt- age levels. The reset operates only when either the clock or the enable is high and provides in- creased flexibility for counter and time measurement require- ments. The MCI 699 is supplied in a flat ceramic package (F Suffix) for compact assemblies and soon will be available also in a DIL ceramic package for easier mounting. Motorola B. V. Emmalaan 41, UTRECHT, The Netherlands Microwave Power Transistor Motorola have just announced a state-of-the-art microwave power transistor, the MRF 835. Charac- teristics specified at 870 MHz using a 12.5 V DC supply arc 15 watts output power, 7.0 dB minimum gain and 50% ef- ficiency. High gain, high power transistors actually consist of a few hundred transistors in parallel. Current to and from these doped regions is carried by metal conductor stripes deposited on top of the die. These stripes must withstand high current densities. In order to achieve good prefomiance at frequencies as high as 950 MHz Motorola make these ‘fingers’ extremely narrow. However, in conventional designs, aluminium in narrow lines with high current density migrates, thus causing early failure of the device. In the MRF 835 Motorola have over- come these problems by using a gold metallisation system - because the automatic weight of gold is higher than that of alu- minium the migration is greatly reduced and the mean-time- between-failure figure is increased 1,000 to 10,000 times. The MRF 835 is designed specifi- cally for mobile radio applications at frequencies around 900 MHz and incorporates Motorola’s ‘Controlled Q’ built-in matching network which ensures broadband performance. Motorola B. V. Emmalaan 41, UTRECHT The Netherlands High current digital clock AMI Microsystems have intro- duced a new digital clock module which offers a high current out- put for direct drive of large LED displays as used in clocks, clock radios, and timers/ elapsed-time counters. Designated S1998A, it provides more than 8 mA per segment, and can be directly substituted for the industry-standard S1998 in high current, low voltage display applications. The 1998A directly interfaces with both solid-state LED displays, and fluorescent/ gas discharge displays. The time- keeping function operates on 50 Hz or 60 Hz inputs, and the display output is available with either AM/PM indicators or 24-hour format options. Other outputs include timed radio turn-off, and radio/alarm enable. A power failure indication is provided to inform the user of an incorrect time display. The S1998A also incorporates a presettable 59-minute count- down timer, an alarm with snooze feature, and unlimited snooze repeat. Clock input noise rejection cir- cuitry eliminates the need for external filtering of the line fre- quency input. Reset-to-zero circuitry is included for timer/ elapsed time applications, and blanking control also allows the use of several circuits in parallel with a single display for multiple event timing. The S1998A, which can operate from power supplies between 8 and 26 V, is pin compatible with the S1998, MM5316, EA5316, and FCI3817. AMI Microsystems Ltd. 108 A Commercial Road, Swindon, Wiltshire, England. Low profile 1C socket Molex have announced the avail- ability of a range of low profile dual-in-line integrated circuit sockets. Designated the 6197 series, they consist of a 94 V-0 black poly- ester polarized housing containing either 14 or 16 discrete pin sockets. These utilize side-wiping contacts, which offer significantly greater contact surface than the more conventional edge-bearing type. Contacts are of 60Cu 30Zn cartridge brass, and are available in a variety of finishes including tin plate and gold over nickel. Optional contact materials such as phosphor bronze, and alternative special contact finishes, are also available. Molex Electronics Ltd., 1 Holder Road, Aldershot, Hants GUI 2 4RH. Multi-family logic probe Designed to simplify and speed logic circuit testing, this new S 125 model 545 A logic probe from Hewlett-Packard indicates digital states and pulses in both high level (CMOS) logic and low level (TTL) logic. An unambigu- ous single lamp indicator displays high or low level or detects bad level and open circuit conditions. CMOS and TTL operation is selected with a slide switch. CMOS logic threshold levels are variable and set automatically. Now, nearly all positive logic up to +18 volts dc can be sensed using one probe. These families include: TTL, DTL, RTL, CMOS, HTL, HiNIL, NMOS and MOS. Another feature of the model 545A is a built-in pulse memory which, along with the display, will catch intermittent pulses. When a logic change occurs, the indicator lamp turns on and remains lighted until the memory is reset. Pulse stretching is provided so the operator can sec fast pulses as short as 10 nanoseconds with the blinking display. Pulse trains to a frequency of 80 MHz are detected in TTL logic, and to 40 MHz in CMOS logic. Light and rugged, this hand-held model 545A is fully protected against voltage overload. Power required for TTL operation is 4.5 to 15 volts DC, and for CMOS operation is 3 to 18 volts DC. To use the 545A, the operator connects the probe to the circuit’s highest level power supply, sets the slide switch to the appropri- ate logic family, then probes. Open, pulsing, or stuck nodes and gates are quickly detected. Hewlett Packard, P.O. Box 349, CH-1217 Meyrin 1 Geneva, Switzerland 642 — elektor june 1976 market Figure 1. Block diagram of an Echo Sounder. The ultrasonic transducer T operates both as transmitter and as receiver. Figure 2. The circuit of an Echo Sounder constructed around LM1812. The transmission fre- quency for underwater distance measurement is about 200 kHz. Figure 3. This block diagram is intended to complement figure 2 to make clear at which stages the external components are connec- ted to the 1C. Figure 4. A Sodar equipment with the LM1812 for operation in air. It operates with a transmitting frequency of 40 kHz. The ef- ficiency is improved by the stage enclosed on the right of the figure, the purpose of which is to extend the internally produced 1 jus pulse to 5 ps. LM1812 Universal Ultrasonic Transceiver The National Semiconductor LM1812 IC consists of a 12 W ultrasonic transmitter, a selective receiver and indicator drive cir- cuitry on a single chip. It is available in an 18 pin D1L plastic package. The circuit was devel- oped primarily for use as an underwater echo sounder (Sonar). As well as measuring water depth it can also be used to locate the position of shoals of fish and other immersed or sunken objects. The IC may also be used with ultrasonic transducers in air, which opens up a completely different range of applications. For example, it is possible to measure the level of corrosive liquids where a sensor cannot be dipped in the fluid. Other possibilities include collision warning systems and intruder alarms (ultrasonic radar -Sodar). Finally, it is also possible to trans- mit data over the ultrasonic link so that communication or remote control systems are possible, both in air and underwater. One appli- cation would be the remote control of model submarines, which is virtually impossible by any other means. In order to understand the oper- ation of the circuit a block diagram of an echo sounder is given in figure 1. The transmitter (block A) feeds the ultrasonic transducer T with a burst of 1 ps pulses for the duration of the transmitting phase (about 800 ps). In under water applications the pulse repetition frequency is about 200 kHz. During this period the receiver (blocks C to F) is switched off to avoid damage by the high transmitter power. This is indicated symboli- cally by switch S, but is of course accomplished electronically within the IC. The ultrasonic pulses applied to the transducer are coupled into the water and spread out as spherical wavefronts. When the sound waves strike the bottom or some other objects they are ultimately received by the trans- ducer and converted back into electrical impulses. These are amplified and detected by the receiver which is now activated. The time duration between transmission and reception of the burst of pulses is proportional to the distance between the trans- ducer and the reflecting object. In the common type of commer- cially available echo sounder the timing is carried out electro- mechanically using a constant speed motor (M). The motor has a disc attached to its shaft the periphery of which are mounted a small neon lamp and a magnet, 180° out of phase. Once every Legend for figure 1 A Transmitter, power stage B Modulator C Receiver D Surge Value Detector E Pulse Sequence Detector F Integrator G Indicator Driver H Indicator I Control, Keying Ratio K Control, Transmitter M d.c. Motor S Transmit/Receive Switch T Ultrasonic-Transducer Pt Gain Control P 2 Interference Suppression Control Legend for figure 3 A Transmitter, power stage B Modulator C a Receiver, 1st. stage Cb Receiver, 2 nd. stage D Surge Value Detector E Pulse Sequence Detector F Integrator G Indicator Driver I Control, Keying Ratio S Transmit/Receive Switch T Ultrasonic-Transducer Pi Gain Control P 2 Interference Suppression Control a = Transmit Pulse Sequence b= Measurement Range market elektor june 1976 — 643 mHRKeTmHRKeTmm he* mn PF F Mum m mnujpiip) revolution the magnet induces a pulse in a pickup coil (block K). This pulse triggers the transmit- ting sequence by controlling the modulator (B). The ultrasonic wave sent out by the transducer is of course subject to the inverse square law, and since it must traverse the distance between the transducer and the reflecting object twice it is atten- uated very severely by the time it returns to the transducer. Energy absorption by the reflecting object and the efficiency of the transducer must also be taken into account, and it is evident that the receiver must be very sensitive. The receiver consists of a multistage amplifier (block C), with gain control provided by the potentiometer PI. This is followed by a threshold gate (block D), which allows only sig- nal exceeding a certain amplitude to pass. This ensures that noise and delayed echoes from very distant objects cannot cause a spurious indication. The signal is further processed in blocks E (pulse sequence detector) and F (integrator). These together form a pulse width detector, which performs two functions. It is first determined whether a valid echo has been received by ensuring that the received signal is of the same duration as the transmitted signal. If more than five consecutive pulses in the sequence are missing then no indication of depth is given. This ensures that any interference pulses strong enough to pass the threshold gate will still not cause a spurious indication. The degree of interference suppression may be adjusted by P2. If a valid echo has been received then the output driver (block G) is triggered and produces an out- put pulse that is stepped up by a transformer to strike the neon. The angle through which the motor shaft (and hence the neon) has turned before the neon strikes depends on the distance between the transducer and the reflecting object. The rotating disc is mounted behind a circular scale with transparent divisions marked off in fathoms, feet or metres. Figure 2 shows a practical circuit for an echo sounder using the LM1812, while figure 3 shows how the external components associated with figure 2 tie in with the internal circuitry of the LM1812. The pulse induced in the pickup coil L3 by the rotating magnet activates the transmitter. The 200 kHz sinewave oscillator is tuned by LI and C3 (it also tunes the receiver gain stages. See figure 3). The output of this oscillator is amplified and limited MOD to provide a 200 kHz squarewave that is used to trigger a mono- stable that produces the 1 /is pulses. After amplification by the transmitter output stage the pulses appear across the trans- ducer T. L2, C2 and the capacitance of the screened transducer lead form a parallel resonant circuit that can be tuned to the transmitting fre- quency by adjusting the core of L2. In this circuit an LED D1 is used rather than a neon. If additional receiver gain is required then an extra stage of amplifi- cation may be connected between blocks C a and Cb- This system relics for its accuracy on the speed stability of the d.c. motor. Slip rings must be used to make connection to the neon unless a rotating transformer is used, and the system inevitably suffers from mechanical wear. However, it is cheap and provides an easily interpreted analogue readout, and echo sounders using these principles are popular with small boat owners. A completely electronic system can be constructed by replacing the motor arrangement with an electronic digital counter. The transmitter can be activated by a rectangular pulse of about 1 ms duration. This also opens a gate to the clock input of the counter, which is fed by pulses from a 644 — elektor june 1976 market Figure 5. Additional circuit for triggering an acoustic alarm in anti-collision devices or burglar alarm systems. Figure 6. Pinout of the LM1812. stable clock pulse generator. The combination of R2 and D1 in figure 2 can be replaced by a 5k6 resistor, and when the returning unltrasonic signal is received a negative going pulse (from about +V b down to 1 V) is available at pin 14. This can be used to store the count and to reset the counter ready for the next count. Simply by choosing the appropriate frequency for the clock pulse generator the out- put of the counter can be made to read in fathoms, feet, metres or any required units of length. Underwater Communications Information can be amplitude- modulated onto the carrier by feeding a low-frequency (e.g. audio) signal into the modulator input pin 8, rather than simply switching the transmitter on and off with digital signals. The received signal must be taken out of the receiver at a point before the signal processing stages (since after signal processing it is no longer an amplitude modu- lated carrier). A convenient point to take off the signal is at the LC circuit L1/C3, connected to pin 1, which tunes both the transmitter and the receiver. The signal can then be fed into a high input impedance buffer stage followed by an AM detector and audio amplification stages. It is also possible to use other modulation techniques that arc not so wasteful of carrier power as AM, notably Frequency Modulation (FM) and Pulse Width Modulation (PWM). Notes for Tables I and II Note 1 : If the 1C is operated at higher temperatures, then load derating should be allowed for. This should refer to a chip temperature of +125°C and a thermal resistance of +167°C/W. This applies to an 1C soldered into a printed circuit board with stationary surrounding air. Because the system is being used for pulsed operation, the heat resulting from the dissipation in the enclosure is only slight. Note 2: During the measure- ment of the sensitivity, an attenuator of 500:1 was fitted, to ensure reliable measurements at higher input levels. Note 3: The 'Modulator Threshold Voltage' is the volt- age which has to be applied to pin 8 to bring the system into the 'Transmit' condition. The current flowing in pin 8 should be limited to 1 ... 10 mA. LM1B12 Pulse sequence detector Interference limiter Display control Ground Display output +Vb, transmitter +Vb Key sequence limiter Ground Table I ABSOLUTE MAXIMUM RATINGS Supply voltage +V b (Connections 12, 6 and 4) Dissipation (Note 1) Operating Temperature Range T amb (Ambient) Storage Temperature Range Maximum Lead Temperature during soldering (Solder duration 60 s max.) 0°C -65° C 18 V= 600 mW . . +70° C . +150°C +300° C Pin Function ^ext (min) Vmax (Instantaneous value V s ) 'max 1 LC-Circuit 30 V 2 Input 2nd. stage 18 V= 50 mA 3 Output 1 st. stage 18 V= 4 Input 1st. stage 50 mA 6 Transmitter Output 36 V 1 A (Transmitter 'Off') for 1 fis 7 Modulator Output 75 k 18 V 8 Modulator 50 mA 9 Pulse Width, passage through zero 7 V 11 Key Sequence Limitation 50 mA 12 +v b 18 V= 13 +V b -T ransmitter 18 V= 14 Display Output 25 V= 1 A (if 'Off') for 1 ms 16 external Display Control 2 M 18 V 17 Interference Limiter 50 mA 18 Pulse Sequence Detector 50 mA Table II (+V b ~ 12 V, T amb ~ +25 C) Parameter Conditions Min. Typical Max. Unit Sensitivity Note 2 200 600 /jV (V ss ) Transmitter (V sat ) rl = io n 1.3 3.0 V Transmitter Leakage Pin 6 = 32 V 0.01 1.0 mA Current Pin 8= 0 V Modulator Threshold Voltage Note 3 0.55 0.7 0.9 V ^° ^ 0 & 0 ? O 9 A® O FIRST BOOK OF TRANSISTOR EQUIVALENTS & SUBSTITUTES A handbook containing 73 pages of transistor equivalents and substitutes including British, German, Dutch, USA, Japanese and European manufacture 40 p O SECOND BOOK OF TRANSISTOR EQUIVALENTS & SUBSTITUTES Complementary to the first book with over 56,000 additional equivalents and substitutes covering virtually all known types at time of publishing 95 p O HANDBOOK OF INTEGRATED CIRCUITS (IC's) EQUIVALENTS & SUBSTITUTES 115 pages of most available types of integrated circuits which have equivalents or substitutes 75 p O PRACTICAL ELECTRONIC SCIENCE PROJECTS 12 construction projects including Electroscope, Gieger radiation counter. Digital clock etc. with circuit diagram & parts lists 75 p o PRACTICAL STEREO AND QUADRAPHONY HANDBOOK Stereo recording, 10 plus 10 stereo amplifier, Quadrasonic reproduction etc. are detailed in its 90 pages 75 p If you do not wish to spoil your copy of Elektor please make a note of the books you require with your name and address and send it to us with'your remittance. further paperbacks now available O PRACTICAL ELECTRONIC PROJECTS A selection of construction projects with circuit diagrams and parts lists including a simple to build, portable radio receiver, 21 watt amplifier for guitars and a simple tester for Digital ICs ... O 79 ELECTRONIC NOVELTY CIRCUITS Circuits include Burglar alarms. Reaction tester game. Wide range frequency doubler. Simple capacitance and Transistor tester. Intercom and signalling device etc. etc. Interchangeabilities and equivalents of solid state devices used are also given O WORLDS SHORT MEDIUM & LONG WAVE FM & TV BROAD- CASTING STATIONS LISTING 91 pages of lists giving station, country, KC/S and KW ratings O THE COMPLETE CAR RADIO MANUAL Includes the selections of a radio and its installation, aerials, suppression and fault finding 90 pages Q MODERN TAPE RECORDING HANDBOOK Covers microphones, sound effects, tunes and editing, fault finding etc. contained in its 91 pages 75 p O DIODE CHARACTERISTICS EQUIVALENTS & SUBSTITUTES European. USA, Asian, Japanese, USSR and Military service types etc. including rectifiers and zener diodes. 25000 entries with index in 9 languages, over 1 50 pages 95 p O RADIO TV AND ELECTRONICS DATA BOOK Over 90 pages of facts and figures including wavelengths and frequency tables. International morse code, the Q code. Resist- ance wire data and many others 60 p O RADIO TV INDUSTRIAL & TRANSMITTING TUBE & VALVE EQUIVALENTS Handbook of equivalents, British, European and USA desig- nations of valves, CRT's and tubes. New system pro-electron CRT's tubes etc. etc 60 p Please send me (tick box) I enclose remittance of for the books and 20 p post and packing. Name . Address 75 p 75 p 60 p 75 p Postcode 650 — elektor june 1976 advertisement 15 TUP I5TUN JC c Uj C r 5 iV i r ] w □ R 11 r | WTO 7T\ r \ \ CAj CO L ' h W WBT • alwr 1 ‘J • a 'iivw 7 * ' jg ■ lyi \ k . O W B ^ ~a ^ • • •• : ..I- • - • - i A advertisement elektor june 1976 — 651 cylinde andteisu^ ^Tp.ctronjcs IC5 1 Q 1/2 7473 Clock Clear K _L_ |2 I 3 ELEKTOR BACK ISSUES ARE STILL AVAILABLE This is a selection from the contents of the various issues: number 1: — tup-tun-dug*dus — equa amplifier — mos clock — distortion meter — tap sensor — electronic loudspeaker — steam whistle number 2: — minidrum — universal display — dil led probe — tv sound — big ben — modulation systems — how to gyrate number 3: — tap preamp — pll systems — fido — time machine — compressor — disc preamp — a/d converter — led displays number 4: — tup-tun tester — interference suppression in cars — thief suppression in cars — supplies for cars — cybernetic beetle — the moth — quadro in practice number 5: 'Summer Circuits' issue, with over 100 circuits: amplifiers, generators, dividers, universal frequency reference, im- proved 7-segment display, receivers, power supplies, rhythm generators, measuring equipment, etc. Prices for single copies: 1 to 4 and 6 to 8 — number 5 — from number 9 — (prices include p & p) U.K. and overseas surface mail 50 p 85 p 55 p overseas air mail 00220 038 — Axe 1 IC4 74121 . s' ■ • •• •V y. ffl ■ m ■ ■ ■■■■ lab and leisur* m ..'a " 4C6 ,. r LPI186 Varactor front end £5.38 LP1185 I.F. Strip £6.22 LP1400 Multiplex decoder £5.62 VAT IN< Z Pin Socket*, pin* in ttripi of 100. Ju*f *nip off what you need. 65p per strip Dual in line 71 8 pin I3p, 74 pin 26p 14 pin 15p 28 pin 3 Op. 16 pin I5p 36 pin 39 p. TOS 8 pin 3 Ip . KJpin 35p. FREE CIRCUITS Our 8 Page A4 Audio I.C. Booklet it tupplied FREE with purchote* of Linear I.C.’t worth £1, or more (35p . if told alone) Contain* circuiti and pin connection* 15 Amplifier* ?50 mW to 20 watt* 5 Audio Pre-Amplifier* Tape Pre-Amplifier Power Driver Instrument Amplifier (Bifer) Generol Purpote Mini Amplifier D.C . Controlled Gain Control Micro-mini radio (2 I.C'*) PROTO -DARLINGTON Vceo 25v 2N5777 Vcbo 25v Vebo 8v II 750 mA „ /// Pd 200 mW J- ) - Hfe 2500 >/J 35 p - LIGNT NEW LED Linear Cursor* each device contain* 10 light cmj fting diode* in a 20 pin dual- in-line package. Ideal for jolid Jtate analogue meter* or dial* Type 101 RED £2.26.* Complete wi th leaflet . 74 TTL 7400 7401 7402 7403 7404 7408 7409 7410 7413 7417 7420 7427 7430 7432 7437 7441 7442 7445 7447 7448 744 7A 7470 7472 7473 7474 7475 7476 748? 7485 7486 7489 7490 7491 7492 7493 7495 74100 74107 74121 74122 74141 74145 74154 74174 74180 74181 74192 74193 74196 24 25-99 100- Up* I2p * 10p • Up* 1 2p * lOp • Up* 1 2p * lOp • 15p* i?!p* lOp • 16p* 13 P * Up • 16p* I3 P * lip • I6p* 13p* lip • I6p* 13p* lip • ?9p* 24p* ?0p • 27p* 22 : P * 20p • 16p* 1 3p * lip • 27 P * 22-p* I8p • 16p* 13p * lip • 27p* 22;p* I8p • ?7p* 22;P* I8p • 75p* 62p* 50p • 65p* 55p* 43p • 85p* 71p* 57p • 81p* 75p* 65p • 75 P * 62p* 50p • 95p* 83 P * 67p 30p* 25p* ?0p • 25p* 21p* I7p • 30p* 25p* ?0p • 32 P * 26p* 21 P • 47p* 39p* 3 1 p • 3?p* 26p* 21p • 75 P * 6?p* 50p • .30 * £1.09 * 87p • 32p* 26p* ?lp • 49p* 65p* 57 P * 45p* 67p* £1.08 * 35p* 34p* 47p* 78p* 68p* £1 . 6 ? * £1 .00 * £1 .06 * £3.20 * £1.35 * £1.35 * £1.64 * 40p* 55p* 46p* 40p* 55p* 89p* ?8p* 28p* 39p* 63p* 58p* £1.48 * 83p* 88p* £2.50 * £1.14 * £1.14 * £1.34 * £1 3?p * * 36p * 3?p * 45p * 72p * ??P * 23p * 3 1 p * 53p * 48p * 86p * 67p * 7 1 p * .90 * 90p * 90p * 99p * 555 <8 pin diplV 55p 555 (TO-991T 8lp ' 556 (14 pin dip' £1.29 703 (RF 'IF Amp' 709 (8 pin dipt 709 (TO-99) 709 (M pin dip) 710 (8 pin dip' 710 (TO-99) 710 (14 pin dip) 71 KTO-99) 711 (14 pin dip) 720 (A. M. Radio 723 (TO-99) 723 (14 pin dip 741 (8 pin dip 741 (TO-99) 741 (U pin dip) 747 M4 pin dip) 748 (8 pin dip) 748 (T0-99> 748 (14 pin dip) 753 IF.M. Ijt. I 75491 75492 68p 38p 45p 39p 39p 45p 44p 51 P 44p ,) £1.76 £1 .09 55p 30p 43p 36p £1.04 42p 46p 49p F) £1.08 88 P El. 10 ' AY-1-0?12 AY -I -5051 AY-5-1224 AY-5-3500 AY-5-3507 AY-5-4007 Regulator* 100 mA 78I05V/C ( T 0-92' 60p * 78LI2WC (T0-92) 60p * 78L15WC (T0-92) 60p * Regulator* 100mA 78L05AWC (TBA625A) o mi < < < o • • • • u u O u BCD encoded digital switch Reading 0 to 9. Suitable for digital clock alorm setting DVM input Scaling etc. I to 9. £1 .49 each * 0.33 LI front x DL 707 Series 0.33" Montanto MAN 50770/8(^3600 0.33, Xciton XAN 70 Serie* KVjntanto MAN 4000 Serie* Litronix DL747 'Jumbo' Series NOTE: 0.43, 0.63 J ■/ -/ */ all £1 .82* each (Red only) on £1 .82* Each (Red, green, yellow, Orange) / .// . oil £1 .49* Each (Red G reen, yellow) \/ J y y * oil £2.3?* Each (Red, Green, Yellow, Oronge) :// all £242* Each (Red onfy) Common Cofhode a* Common A node (Red only) MAN 4000 Serie* pin out* are 14 pin dil the tame at MAN 50; 70 ond 80 series. LIGHT EMITTING DIODES tout'd Free *nop~on plastic retainer 0.125" 0.16" dio Jen* (T1L209) dia . lent 0 . 2 " dia. lent (MLED 650) 1* 10* 100* 1* 10* 100* 1* 10* 100* Red I6p 1 5p 13p * 27 P 24p 22p* 18p I6p Up * Green 27p 24 p 22p* 33p 30p 27 P * 30p 27p 25p * Oronge 27p 24 P 22p* 33p 30p 27 P * 30p 27p 25p * Yellow 34p 3 1 p 29p* 35p 32p 29p* 35p 33p 3 Op * Our Bulk Buying Power enable* a repeat of our Special Offer: 0.337, (LIT 707) 90p.*(inc. VAT) 0.63 (LIT 747) £1.99 .*(inc. VAT) OPTO ISOLATORS I L 1 4N25 or T1L116 6 pin induttry standard packoge 2. 5KV isolation £1.00,* MAINS TRANSFORMERS Secondaries 6-0-6V 9-0-SV 1 ?-0-l 7V 20-0-20V 24-0- 24V 28-0- 28V 100 mA 100 mA 100 mA I A 500 mA 1 A 97p. 97p. 97p £4.55 £2.48 £6.18 0-9-17 V 0-1 2-15- ?0-?4-30V 0-24-30-40-48-60 V 0-19-25-33-40-50 V lA £2.70 * __IA £1.95 * £2.48. * £5.18 * £3.40 * __2A_. £2.27 £4.15. £7.02 £4.53 4A_ £2.64 * p.p. on transformer* 10% of price, min. 20p t M6800 has taken the gamble out of microprocessors Seven reasons why you'll always win with M68Q0. 1 . Programming language. So easilv learned that it makes your transition to MPU's that much easier. 2. Unlike competitive ranges, the M6800 family is capable of further development while still maintaining upward compatibility. Example: The M6900 series is now being defined to meet defined customers' requirements. 3. Very efficient programme code. Wide instruction repertoire, including seven addressing modes. 4. Sub-function devices already available. 5. Single power rail. 5 volt. 6. Interfaces easily with TTL and CMOS. 7. Second sourced by AMI across Europe. Heres the M6800 family today: — MC6800 MC6820 Microprocessor. Peripheral Interface Adapter. MCM6810 Static RAM. MCM6830 ROM. Asynchronous Communications 4 / Interface Adapter. Low Speed Modem. MC6850 MC6860 Alternative N-Channel Si Gate RAMs for large systems: — MCM 38102 IK x 1 MCM6814' 4Kx 1 MCM 68 15 4K x 1 Static 16-pin. Dynamic 16-pin. Dynamic 22-pin. Recent new devices include: — Dynamic Memory Refresh Controller. %7 v MCM681 12A 256 x 4 Static RAM, 16-pin. MCM 683 17 16K Static ROM, 24-pin. An 8K x 1 erasable and electrically reprogrammable ROM (MCM 68708) was introduced in the first quarter of 1976. And there's more to come! MOTOROLA Semiconductors Motorola Ltd. Semiconductor Products Division York House, Empire Way, Wembley. Tel: 01-902 8836.