INDIA November 1985 elekt©? Volume 3-Number 11 This month's front cover shows the high - definition colour graphics card described in four parts in our October 1985 to January 1986 issues, and some typical examples of the displays obtainable with it. news & views editorial 5 electronics scene - 15 electronics technology electronics & medicine 12 design ideas — clock oscillators 17 viewdata in Britain 50 electronics select • 52 projects high-resolution graphics card — 2 20 Second part of the articles describing a versatile graphics card. sound rotator 26 Electronic version of the mechanical Lesley speaker system , stage lighting Economical way of obtaining a fully controllable stage lighting system. hand-held anemometer 36 Ideal for. among others, wind surfers, glider pilots, amateur meteorologist, and yachtsmen. luxury transistor tester 44 simple sound effects 48 using only two ICS a wide range of sounds can be produced verying from an American police siren to the twittering of birds information corrections 74 new products 65 guide lines switchboard * 69 classified ads 74 advertisers index 74 A selection from our next issue : • sweep generator • stage lighting - 2 • MIDI interface • graphics card - 3 • car radio amplifier selex-6 digi course (chapter-6) 56 constant current sources 59 AC supply 61 resistor combination 52 elektor india november 1 985 1 1 .03 COMTECH TEXES TMNSF0ME1S In both design and construction the reliability of the Transformer is given first consideration. TEXES FOR TRANSFORMERS • Top is Sound Mechanical Construction • Economical and Compact Design • X, Flux and Current Density such as to avoid local loading. • Excellent Performance • Strict Quality Control. M anufacturers of Step up S. Step down Power Transformers (Single Phase S. Three Phase), Ignition and Neon Sign Transformers, Power Supplies and Special Transformers as per Customers Specifications. Regd. Office ; Sabeena Apna Vatan. Opp.Ueonar Bus Depot. Bombay - 400 088. Phone: 551 3681 i simia ? Table Model LD 255 & LO 555 3% And 4% Auto Manual Range Voltage Current Resistance Accuracy Optional 0—1000 Vts Ac/Dc 0—10 A Ac/Dc 0 — 20 Meg 0 ± 5% And 0+ 0.05% Frequency Counter and Thermometer Hand Held LD 155 3% And 4 Vs D.M.M. Auto or Manual Range. Voltage Current Resistance Capacitance Ran- lui! i - I © © © 0-1000 Vts Ac/Dc 0—10 Amps 0—20 Meg. up to 20 MFD Pleace Contact: LOGICS 1 07. 6 Main road, 8 Cross Malleswaram BANGALORE-560003 Phone 366 526 Grams: COMPULOGIK INSTRUMENT-CASE TYPE-T-220 Also available • Thumbwheel switches • Display Bezel for 3*4 digit • Instrument clamps For more information, contact: COMPONENT TECHNOUE 8, Orion. L.P. Road, Andheri (W), Bombay-400 058 Tel: c/o 4224066 1 1.04 elektor india november 1 985 elektGF November 1 985 PUBLISHER: C.R. CHANDARANA EDITOR: SURENDRA IYER EDIT ASSISTANCE: ASHOK DONGRE ADMINISTRATION: J DHAS PRODUCTION: C N. MITHAGARI ADVERTISING & SUBSCRIPTIONS eIeIu Italy Editor 0 FumeyeHi feirtwd 6 Bento UJa « D CiletVM*. X2 1* 1000 Luboa Portugal Editor Jo*9» Goncatve* Ingolek S A Av A ll on VO XMI, 141 Mw-l 16 Speei Frliinr > « Ferrer Copyright ? 1985 Elektuur BV. Netherland Sinclair: any suitors? Why is it that so many clever and inventive entrepreneurs seem to run into trouble so often? Take Sir Clive Sinclair: entrepreneur and inventor par excellence. But, so far, this year has not been a very good one for him. Starting with a divorce, it then confronted him with a declining home com- puter market, troubles with the Advertising Standards Auth- ority about the C5 one-man electric car, and a writ claiming more than £1.5 million from Hoover for alleged non-payment of bills. It all seems rather unfair to a man who has, arguably, con- tributed more to the computer industry than anyone else since the early 1970s. In fact, it is fairly certain that without him there would never have been so many computer en- thusiasts in Britain — or in the rest of the world. Clive Sinclair started his first company in 1962 while still in his early twenties. Ten years later, Sinclair Radionics launched a pocket calculator which was then almost certainly the smallest, best designed, and cheapest in the world. But fierce and growing competition caused the company to be bailed out some years later by the National Enterprise Board. Within a short time, Clive Sinclair had left the company: was it because he could not — or would not — work within the constraints of an industrial undertaking? However, this move was the computer world's gain. Within two years, Sinclair had launched the first-ever computer to sell for under £100: the ZX80. Over 100000 of these were pro- duced. Its successor, the ZX81, was even more successful with sales topping a million worldwide. Even that has been put in the shade by the ZX Spectrum, probably the most successful computer the world will ever see. But then, a decline set in: neither the hand-held flat-screen TV receiver nor the new QL computer caught on. It seems, however, that the C5 electric car — or rather the £12 million spent on its design? — has done the most damage to Sinclair. Sinclair has still plenty of ideas, but he needs money to materialize them. There should be sufficient interest from in- dustrial backers for his wafer-scale chips and his compact data storage device; rather less so for his electric car with a roof. When, last June, Mr Robert Maxwell proposed a £12 million rescue plan, there were sighs of relief in many quarters of the industry. But, alas, when it was learnt that the stock of com- puters amounted to no less than £35 million, Mr Maxwell called off the deal. \ At the time of writing, no other helping hand has been ex- tended. None the less, there are some glimmers of hope for Sir Clive. The recently merged Dixons/Currys Group have placed a £10 million order for computers and miniature TV sets. At the same time, there are signs that the miniature television receiver may take off at last in earnest — particularly in the USA. We hope these may prove to be turning points for Sir Clive, but even so, he will need further financial help. We wish him well: he deserves it! elektor indie november 1985 1 1 .05 n-frS (A MICROPACK pioneers many FIRSTS in micropackaging Electronic Circuits, such ■ Multilayer Boards (Mass laminated) ■ SMOBC/SSC (Solder mask on bare copper/Selective solder coating) PTH Boards for wave soldering ■ CMC Drilling (small holes) ■ Computer Aided Testing ■ Membrane Touch Panels ■ Multilevel circuits ■ Polymer Thick Film hybrids ■ Bonded Heat Sink PCBs ■ Solder coated boards to MIL-P-551 10 C Multilayers approved by Underwriter’s Laboratories (U.L.), U.S.A. Contact MICROPACK and touch tomorrow LlUjJ Micropack MICROPACK LIMITED Raghavendra Complex 418/1, 10th Main, IV Block, 2 Jayanagar, Bangalore-56001 1. -£ Ph. 41232 Grams : 'MICROPACK-Bangalore' ;5 Tlx.: 0845-8429 PAC IN S. Semiconductor Complex Ltd. * Microprocessor Ics/LSI * CMOS/Linear * Watch Modules/Displays * New range of ICs to be announced. 307, Arun Chambers, Tardeo, Bombay 400 034 Telephone: 4946422,4947953,4946553 Cable: AHUJELEC. Telex: OH-76393 GENE. 001 Market Showroom General Sales Agency 4, Yamuna Bldg,TaraTemple Lane, Lamington Road, Bombay 400 007. Tel: 369250, 386178. Almost Everything For Sinclair ZX SPECTRUM Computers • Guaranteed Repairs. • Latest Software Rs. 50/- for two programmes including cassette and manual..Get our free list. • Custom written Software. • Monitor Rs. 2100/- • 80 Column Dot Matrix printers with interface for QL also , • All types of Interfaces . • All types of Peripherials. • Many Accessories. • C-10-C-15 blank cassette for data recording. • Furniture: Simple and Elegant tables specially made for computers. Pioneer Electronics 2, Court House, Dhobitalao, Bombay - 400 002. Ph Office 31 1634 Res; 475693 11.06 elektor mdia november 1 985 CLEANLINESS “ The Invisible Dimension of Quality — DISCOVER WITH KIRLOSKAR CLEAN AIR SYSTEMS. A must for Dustfree, sterile Atmosphere in Electronics, Aeronautics, Engineering, Medical, Chemical and Pharmaceuticals. Clean Bench : Horizontal/ Vertical Flow e Provides class TOO Clean Air in restricted area. • Removes dust particles down to 0.3 micron • 99 97% efficiency e Available in size- 610x610 mm 1220x610 mm. 1830x610 mm e Table Top Models also available Clean Room : • Provides upto class 1 00 Clean Air over large work areas. Air Curtain : • Arrests entrance of dust in doorways • Available in different sizes Air Shower : Fully automatic, electrically operated, direct highly pressurised air through nozzles to remove all dust particles from clothing, footwear and other articles before entry into a clean room. 4528 Bio Hazard Unit with U/V Light/Virus burn unit. • Ideal for safe microbeal research e Available in size-610x610 mm 1220x610 mm. 1830x610 mm e Chemical Benches are also available in above sizes. Portable Electronic Cleaner : For fume clearing and recovery of suspended particles in pharmaceutical industry. Handles 1 5 to 1 80 M Vminute air. Kirloskar Electrodyne : Total Capability In Clean Air Systems • Works with high efficiency at high temperature. • Can be coupled to air conditioning or airhandling systems • Low on power consumption. • Easy to maintain. • Long life For details contact KIRLOSKAR ELECTRODYNE PVT. LTD. 118, General Block, M.I.D.C. Bhosan, Pune 411026 Tel 86121, 86122, 86123. 86124. 84429 Telex: 145-260 KEPL IN Regional Offices at: Ahmedabad Tel 52228 Bangalore Tel. 369556, Calcutta Tel. 433836, New Delhi Tel. 312131. Jaipur Tel. 77734 elektor india november 1985 11.07 11.10 eleklor india n UNLIMITED Electronic and Computer Components The Largest Retail Showroom ON FIRST FLOOR Range of components available: ■ Microprocessor ICs ■ Linear ICs ■ Entertainment ICs ■ CMOS, TTL, LS ICs ■ MC, LF, CA, XR, TLO, 8T series ICs ■ CD, CD45, 74, 74S, 74C, 74LS series ICs ■ 78, 78M, 78L, 78P, 79 and LM series Voltage Regulators ■ All type of T ransistors and Semiconductors ■ A wide range of Resistors and Capacitors ■ Ten Turn Trimpots, Single Turn Trimpots, Cermet Pots. Helipots BFIat Cables ■ FRC Connectors, D Type Connectors ■ Textool ZIF Sockets ■ 1C Sockets ■ A wide range of Switches and Thumbwhell Switches ■ Standard Cabinets and Hardware ■ ASCII Keyboards ■ Floppy Discs ■SparePartsandsoftwareforSPECTRUM. »Seven Segment LED and LCD Displays ■ Crystals from 1MHz to 32.768MHz ■ Optocouplers. UNLIMITED Electronic & Computer Components Neelam Manzil, 1st Floor, 350, Lamington Road, Bombay 400 007 Phone: 369370 E NUMERIC PRINTERS! We Keep Largest Stock At The Most Cheapest Price We Keep Largest Stock At The Most Cheapest Price EPROM • 2716 • 2732 • 2532 • 2764 REGULATOR • 7805 • 7809 • 7812 • 7815 CERAMIC FILTER SFE 5. 5 MB SFE 10, 7 MB TTL/CMOS 74 LSOO-373 4001-45106 SCR/TRIAC 6A-40A 400V-600V OP-AMP • 324 • 339 • 741 • 158 • 258 • 358 TBA-SERIES • 800 • 810 • 820 • 920 TDA-SERIES • 1154 • 1170 • 1190 • 2593 TRANSISTORS • 2N • BL • BD • BF-SERIES MANY MORE SALEABLE ITEMS .... CONTACT: 4 4 #- # t * & t R0CH0R ELECTRONICS ENTERPRISE 5008. BLK 1. ROCMOR ROAD. (ROCHOR CENTRE) SINGAPORE 0718 TEL 2941094 2928364 TELEX RS 26483 ROTRON CABLE ROTRONIC A= NUMERIC PRINTERS] ? elect *onics ■ Epson 460 Mechanism ■ 16 Columns of numeric print ■ BCD TTL input ■ Parallel and/or mux data input ■ Parallel and mux data input can be mixed ■ Microprocessor based ■ Compact and panel mountable ■ Drawout design for paper roll replenishment ■ Low price for its value Also available alphanumeric printers For details contact : ARUN ELECTRONICS PVT. LTD. B Ansa industrial Estate. Saki Vihar Road. Bombay-400 072 Tel. 583354. 587101. 585975 Gram ARUNEl Telex: 11-72236 AOPL-IN I The Motwane A'j s Digit Multimeters DM-450 and 451 are as good as the best in the world. Developed in our inhouse R & D laboratory, these DMMs compare in terms of reliability, efficiency, performance, quality of design and manufacture; or simplicity in operation— with the Prime. The Motwane DM-450/45 1 offer a host of features, the most outstanding of which are: ■ True R.M.S. volt and current measurements over a wide frequency range (Model DM-4501. ■ 1 O microvolt resolution on AC and DC ■ Imposing accuracy— Basic of O.D3°/o of read- ing + 1 digit and low Tempco. ■ B.C.D. output for systems capability. ■ Full protection against inadvertent overload. ■ Sleek plastic casing for the most effective sealing, longer lasting good looks. making the instruments practically unbreakable. ■ Quality that's exclusive, at a price that's not. These 4’/s digit, 5 function. 27 ranges DMMs measure upto 1 OOOV on AC and DC. 2 Amps on AC and DC and 20 Megohm. With careful component selection, sophisticated L.S.I. tech- nology and ultra precision resistors, they guarantee excellent long term stability, quick response time and greatest reliability. Since instruments of this kind are most often used as a secondary standard, recalibration intervals are longer. These DMMs naturally include automatic polarity and zero. Hi Lo ohms, have only two terminals and are human engineered for the simplest, yet most efficient operation. Anticipating your future need to build a system around these instruments, also avai- lable are the following optional accessories: ■ Digital limit comparator for go-no-go checks. ■ Digital printer for hard copy ■ A simple quick mate jig for speedy Q.C tests. The DM-4SO and 451 perf orm better because they are built better. We had you in mind when we designed them. For further details write to: THE MOTWANE MANUFACTURING COMPANY PVT LTD., at Gyan Baug, Nasik Road 422 101 Tel ; 86297/ 86084 Telex: 752- 247 MMPL IN Grams: MOTWANE or Gyan Ghar. Plot 434 A. 14th Road. Khar. Bombay-400 052. Grams: MOTESTEM. Cut rrere and mail — oPiease send literature and Quotation Name Designation Company Address Telephone Telex SELL ADS.MMC. 12.84 elektor India november 1985 1 1.1 1 MOTWANE Electronic instrumentation is saving lives The use of electronic devices and instrumentation in medical care has increased dramatically over the past 20 years and new technology is being introduced at an exciting pace. Elec- tronic instrumentation is used to moni tor the heart, brain and liver, and to measure pulse rate, blood flow, and metabolism of the human body. It is used to determine the size, shape, and orientation of an unborn child, and to locate the presence of tumours, damaged tissue, and broken bones. From the simple blood pressure analyser to the heart pacemaker and whole body scanner, electronic instrumentation has made possible the alleviation of suffering and the pro- longation of human life to an extent that would have caused past gener- ations to gasp in astonishment. Among recent developments has been the pioneering of life-saving surgery by the so called laser scalpel developed by fed into the patient and carefully pos itioned so that its pin-head tip is aligned opposite the treatment site. The important alignment procedure is assisted by passing visible light through the cable and connecting it to a viewing device known as an endoscope. When the cable tip is precisely pos- itioned at the delivery site, the visible light is replaced with a short duration burst of near-infrared light derived from a powerful laser. The laser, its power supply, and all the controls are housed in a floor standing unit that can be used anywhere in the hospital where a three phase electrical supply and mains cooling water are available. Known as Fiberlase-100, the system uses a neodymium doped yttrium alu- minium garnet laser (usually described by the abbreviation Nd-YAG), which produces near-infrared light at a wavelength of 1.06 pm. At this wavelength, the laser light is able to penetrate deep into the tissue and can seal off large, bleeding vessels and Harwell developed ultrasonic inspection system in use at Moorfields Eye Hospital. London displays ultrasonic images of the eye directly on a television screen and allows them to be recorded on a standard video recorder. Electronics and medicine Barr & Stroud. This revolutionary technique was first used in the early 1980s and, in particular, in 1984 at the Oldham and District General Hospital in northwest England, by Naru Hira, senior consultant surgeon. He treated bleeding stomach ulcers and removed tumours all without open surgery and with only a local anaesthetic. Laser Light The Barr Et Stroud technique relies upon a length of fibre optic cable being t . destroy tumour cells. The laser power delivered to the target site is micropro- cessor controlled and can be delivered in pre settable 10 W increments over the range 50 to 100 W. The laser pulse duration can be similarly pre-set over the range 0.1 to 9.9 seconds in in- crements of 0.1 second. Digital displays are a very important feature of the Fibrelase-100. They can be used to indicate the energy de- livered by each laser pulse, the number of pulses delivered, and the accumu- lated energy dose received at the target site. Radioactive Tracers An ancillary facility provided by the system is the ability to deliver carbon dioxide gas at low or high flow rates for clearing blood from the target site. The first application of nuclear radi- ation in medicine took place in 1898, when Marie Curie and her husband Pierre used gamma radiation from naturally occurring radium on tumours in the human body. Since then, techni ques have been developed for produc ing an enormous range of very useful radioactive isotopes of a type that do not exist in nature. These so called artificial radio-isotopes are produced mainly from the neutron irradiation of non-radioactive nuclei in a nuclear reactor or, less frequenctly, from proton irradiation of such nuclei in the target chamber of a particle accelerator such as a cyclotron. Radio-isotopes are widely used in the medical field as tracers, for example, for determining the uptake of iodine by the thyroid gland, the reaction of metabolism of the body to iron and other minerals, and the formation and utilization of fats. Tracers are also widely used to detect the presence of tumours, to measure blood volume in a limb, and to diagnose heart disease. Picture of Heart Many of the radio-isotopes used in Britain — and in many other parts of the world — are manufactured in two nuclear reactors and a powerful cyclotron at the Harwell Laboratory in Oxfordshire. These isotopes, in the This thyroid counting system from Nuclear Enterprises features a dedicated micropro- cessor that provides all the facilities required for operating the system. 11.12 eleklor india november 1985 main, are marketed by Amersham International, which has two packag ing and distribution plants in Britain and exports more than 80 per cent of its products. The Harwell laboratory has recently perfected a new production technique using its variable energy cyclotron for producing the radioactive isotope gold-195m. This new radio-isotope has a half-life of only 30.5 seconds and is used in a technique known as single pass nuclear angiography for investigating the dynamics of the heart. The technique begins with the injec- tion of a small amount of radioactive gold into the bloodstream and it quickly passes through the vessels of the heart. A radiation detector made by Scintronix of Livingston in the Lothian Region is positioned outside the body over the heart, and is used to detect the gamma radiation emitted by the radioactive blood cells. It supplies signals to a special camera, which takes up to 60 pictures a second of the emitted radiation. The equipment uses these to build up an image of the heart for display on a conventional television monitor, and to record the data on a video recorder. The image displayed on the monitor can show the heart beating and privides information about its function and operating efficiency. Repeated Tests It is the short half-life of gold-195m that makes this particular isotope so valu- able. Earlier techniques made use of the technetium-99m radio isotope, but this has a half life of six hours. With a half life of only 30.5 seconds the injected gold disappears from the body within a few minutes and enables the patient to be subjected to repeated tests at short intervals. The ability to do this is valuable in monitoring the performance of the heart after exercise or assessing its response to drugs or other stimuli. The new gold radio-isotope has en- abled a major advance to be made in the development of standard diagnos tic tests for coronary heart disease and is being assessed at St Bartholomew's Hospital, London, and at the Western Infirmary, Glasgow. A British company that specializes in the design, development, and manu- facture of nucleonic instruments is Nuclear Enterprises of Reading. Founded in 1956, this company sup- plies instruments to hospitals, teaching establishments, and nuclear installations all over the world. Ultrasonic Radiation Representative of the company's nucleonic instrumentation in the medical field is its thyroid counting system. This comprises a scintillation- type radiation detector connected to a microprocessor-controlled electronics assembly. A single channel analyser is set to respond to a narrow band of gamma radiation centred on that emit- ted by the iodine radio-isotope con- tained in the patient's thyroid gland. Signals originating in this way are fed to a ratemeter, whose output is con- nected to a chart recorder displaying the relationship between thyroid uptake and time. Among many other products appli- cable to the medical field. Nuclear Enterprises also manufactures the lonex dose/dose-rate meter, which allows precision routine measurements of accumulated gamma-ray dose and dose rate at therapy and patient pro tection levels. Ultrasonic radiation is increasingly used in medicine because it is possible to image soft tissue structure, and it is a safer substitute for X and gamma radiation, particularly in the examina tion of pregnant women and of those parts of the body where there is insuffi- cient contrast in the irradiated tissue to produce a satisfactory radiograph. A good example of the use of ultrasonics is the new data recording and imaging system developed by the Harwell laboratory for the ultrasonic eye scan- ner used at the Moorfields Eye Hospital, London. The system displays ultrasonic images of the eye directly on a standard tele vision monitor and allows them to be recorded on a video cassette recorder. The recording provides the surgeon with a permanent record of the patient's eye condition and allows images to be replayed later during diagnostic discussions. Campaign against noise Noise is a hazard that all too often is over looked by workers and manage ment who could be unaware of its harmful effects until it is too late. That is why Britain's Health and Safety Executive (HSE), the organization that checks safety at work, has launched a major campaign to redure noise levels at the place of work. It is estimated that in the United Kingdom alone almost a million people work in noise levels high enough to create a risk to their ears. Some workers may actually suffer permanent damage. Those particularly at risk include shipbuilders, steel workers, men using pneumatic hammers, riv- eters and, those employed in canning factories. A circular saw produces about 102 dB and a stockman who looks after pigs can be subjected to squeals of 108 dB. This may not sound much of a differ- ence until it is remembered that 1 dB (decibel) is a logarithmic unit of sound intensity, so that a difference of 6 dB represents a sixfold increase in noise. The HSE campaign hopes to increase awareness of the danger of noise and to encourage quieter machines. Noise in machinery can be dampened by enclosure — for instance, chipboard drills can be quietened by 30 dB in this way — or by special mountings that can reduce noise from vibrating machinery by 13 dB in some cases. The HSE has produced a booklet listing 100 examples of quietening tools and equipment. Ear Protectors Another objective is to encourage equipment manufacturers to make quieter machines. This design against noise campaign has received support from the Government and is gaining wider backing all the time. The HSE is also trying to persuade eiektor india november 1 985 11.13 Conducting a noise level survey in a factory. workers that when ear protectors are supplied by management they ought to be worn at all times. Noise related hearing impairment is often slow to develop, but, once it does, can be per- manent. Total or partial loss of hearing can result, or the workers may suffer from tinnitus — a constant ringing in the ears. Whatever the risk, all too often workers leave ear protectors off because they are uncomfortable. In a more general effort to increase noise awareness, the HSE has a mobile exhibition that tours Britain, and it would welcome further legislation to control noise. Computerized artificial hand Doctors and biomechanical re- searchers are producing ingenious devices to replace defective or missing parts of the human body. Prostheses, as they are called, rarely copy the originals absolutely but, with constant advances in associated fields, such as control engineering and electronics, mechanical prostheses are getting better. At the University of Southampton, Professor Jim Nightingale and his team have been concentrating on one part of the body where an efficient pro- sthesis could transform the patient's life. For the last 13 years they have been working on an artificial hand. Now, using computer technology and miniature motors, Professor Night- ingale believes the team will soon have a model to go into clinical trials. This sophisticated artificial hand, con- trolled by sensors and a microcomputer carried by the user, was designed at Southampton University, England. Numerous problems face existing pro- sthetic hands. Many are cumbersome with a limited amount of movement. They are generally controlled by muscle stimulation in the stump and, because of poor processing of this signal, hand movements tend to be delayed and only one move can be made at a time. This is quite difficult for the patient when, for example, he 11.14 elektor india novembar 1 985 wants to pick up a glass. In addition, the user has consciously to provide muscle stimulation all the time to maintain a grip. Professor Nightingale has solved many of these problems with the help of a small personal computer that is carried by the patient. He has also designed special sensors to form part of the artificial hand and constantly feed information back to the computer. More Sensitive Grip This means that all the patient now has to do is to put the new hand near, say, the glass and, through the normal muscles in this arm, instruct the fingers to close. Three tiny motors in the hand control the thumb, forefinger and other fingers. They all move separately so that the hand closes on the glass with the fingers wrapped into the appropriate positions. From then on the patient can forget that he is using a prosthesis because the com- puter takes over. Should the glass slip, special slip sensors cause the hand pressure to be increased. The hand weighs only 450 g but can deliver -a grip force of 2 kg and, with the sensors embedded in the fingers, should be almost as robust as a normal hand. The computer that controls it is about the size of a video tape cassette but Professor Nightingale believes his team might soon be able to fit a smaller one into the hand itself. Unfor- tunately, the battery to power the equipment is still fairly cumbersome. Professor Nightingale's prosthesis is still in the research stage, but the suc- cessful use of sensors and feedback with a computer that is not prepro- grammed offers real hope for more natural mechanical hands. More effective monitoring of drug safety Britain’s Committee of Safety of Medicines ICSM), which monitors the effects of drugs on patients, has initiated a trial to improve the reporting by family doctors of adverse drug reac- tions on patients. The pilot study will involve 300 doctors and will look at the feasibility of using microcomputers of viewdata equipment to make the reports. If the scheme is succesful, there is the likelihood that such systems will ultimately be installed in every family doctor’s surgery. The CSM normally learns about adverse drug reactions by what is known as the yellow card system. The doctor is supposed to fill in a yellow form and post it to the CSM if he has observed some side effect. Many health workers believe this system misses too many cases. If the computerized scheme proves popular, the CSM is confident it will improve A family doctor takes part in a pilot pro- ject in Britain designed to give rapid warning of harmful side effects caused by drugs. The doctor used Prestel viewdata equipment in his surgery to feed the warning into a central data bank. the quality and quantity of reports. Doctors involved in the pilot study key in answers to various questions that they call up on the screen. These ques- tions relate to specific reactions to a drug. The keyed-in information is then sent direct to the CSM by telephone link. Every family doctor has his own security code in the system and the central databanks are further protected by more complex codes so that only authorized people can gain access. The process will speed up the assess- ment of a drug’s safety and effec- tiveness. It should also permit drugs that have proved dangerous to be withdrawn immediately. The CSM even hope to be able to tell doctors that a drug has been banned before the decision is made public through the newspapers. However, the CSM has assured doctors that the computer link will not automatically lead to more drugs than at present being taken off the market. (LPS) dock oscillators desian ideas An oscillator is a circuit that converts direct-current power into alternating- current power, in contrast to a gener ator, which is a device that converts mechanical energy into electrical energy. There are, basically, two categories of ■oscillator that are of interest to the electronics engineer: harmonic and relaxation. The former produce sinu- soidal waveforms and contain at least one active element that supplies power constantly to the passive components, whereas relaxation oscillators produce non-sinusoidal waveforms, such as rectangular pulses. An oscillator is generally an amplifier operating with positive feedback in a manner whereby an output is pro- duced without any input signal. To achieve the desired frequency, every oscillator contains a frequency- determining parj, which may be an LC circuit, a phase-shifting RC network, or a quartz crystal. Clock oscillators are the simplest of crystal oscillator, which may, none the less, have an overall accuracy of 50 p.p.m. These oscillators form the back- bone of the vast majority of digital circuits. The amplification of an amplifier ope- rating with positive feedback is given by A ' = (1 * where A’ is the amplification with feedback; A is the open-loop amplifi- cation; and ft is the portion (expressed as a vulgar or decimal fraction) of the output that is fed back to the input. In designing oscillators, it is necessary to control the feedback in a manner whereby ftA =1 (the Barkhausen cri- terion) is true at only one frequency: the desired frequency. When this cri- terion is met, an output signal exists even when the input signal is zero. The quartz crystal A quartz crystal is cut from a bar of manufactured quartz. If an alternating electric potential is applied across a certain direction (related to Young's modulus) of the crystal, mechanical vibrations result. If the frequency of the applied potential corresponds to a natural frequency of vibration of the crystal, very powerful vibrations are set up, which, in turn, cause an alterna- ting field across the crystal. The mechanical vibrations suffer little from damping and have a sharp resonance peak. The equivalent circuit of a quartz cyrstal is shown in Fig 1. Note that C u r— c 1 i i I Co R2 ; -ii— fa- 1 1 O-I l-o ' p = 2^ZC IMHzI From (2) and (3) it is evident that the parallel resonant frequency is always greater than the series resonant fre- quency, although, since C 0 >>C 1 , they are very close. If the modulus of impedance, IZl, of the crystal is plotted as a function of the frequency, a characteristic curve as shown in Fig. 2 is obtained. When the frequency is increased beyond the parallel resonant frequency, the crystal behaves again as a capacitance (as it did below fj. At about three times the fundamental resonant frequency, there is again a series and a parallel resonan ce: this is called the third harmonic frequency of the crystal. The physical representation of how a crystal can vibrate at three times its fundamental frequency is given in Fig. 3. The thin Fig. 1 (a) Equivalent circuit of quartz crystal, and (b) circuit symbol. L ,■ and /?, are not true electrical com- ponents, but merely serve to illustrate the performance of a crystal vibrating at, or near, its resonance frequency, while C 0 is the electrostatic capaci- tance of the electrodes, and R 2 is the connection resistance, which normally has such a low value that it can be ignored. Typical values for a 100 kHz crystal are L =85 H; C, =0.03 pF; R y =280 Q; and C 0 =3.5 pF. The series-parallel equivalent circuit shows that there is both a series res- onant frequency, f % (zero impedance), and a parallel resonant frequency, f v (infinite impedance). The series reson- ant frequency is '* = ^k ,MHzl ,2) The parallel resonant frequency is Fig. 3 Physical representation of a vibrat- ing crystal. quartz disc is deformed longitudinally, but the deflection at the suspension points remains zero, so that only vibra- tions that are an odd multiple of the fundamental frequency Can take place. Modern crystals are available with fun damental frequencies from a few kHz up to about 30 MHz, third harmonics from 20. . .90 MHz, and fifth harmon- ics from 60. . .150 MHz. Note that the mode of vibration is primarily depen- dent upon the thickness of the plate. Requirements of a clock oscillator Clock oscillators must be reliable, eas- ily reproducible, and simple. An example of such an oscillator is shown in Fig. 4. This operates in the parallel Fig. 2 Modulus ( = magnitude) of impedance plotted against frequency. Fig. 4 Circuit of a simple clock oscillator operating in the parallel mode. mode, which generally means at fun- damental frequency. The dynamic capacitance (C, in Fig. 1) and capaci- tors C, and C 2 form a capacitive div- ider, which means that the inverting amplifier needs a high impedance input and output (whence the current source symbol at its output). The out- put current should not exceed a cer- tain maximum value, otherwise the elektor mdia november 1 985 1 1.17 crystal may become damaged. Also, since some of the energy is converted into heat by /?,, an excessive current may adversely affect the stability. In general, the power supplied to crystals operating in the 2. . .30 MHz range should not exceed 10 mW: a safe value is 1 . . .3 mW. Some special crystals for operation below 2 MHz should not be supplied with more than 100 pW. 5 Fig. 5 Circuit of a series-mode oscillator, which is. however, not very reliable in operation. A series-mode oscillator, Fig. 5, uses a non-inverting amplifier, so that both the. input and the output are low impedance. This type of circuit is not very reliable. For instance, when the Q of the crystal is not very high, the oscillator may work as an astable multivibrator (AMVI. The quality factor, Q of a crystal is n - 2nf J- - 1 ° ~ Rx ~ 2n4C,/7, (5 ' Even fundamental frequency crystals may, in a series mode circuit, prefer to work on the third harmonic. It is, there- fore, advisable that, in general, clock oscillators use parallel-mode circuits. Overtone oscillators Oscillators intended for operation on harmonics of the crystal frequency are called overtone oscillators. The series- mode overtone oscillator of Fig. 6 I Fig, 6 A series-mode oscillator can be made to operate on the third harmonic by the use of a notch filter tuned to the fundamental frequency. needs a notch filter, L-C-C A , tuned to the fundamental crystal frequency, for satisfactory operation. In the parallel- mode overtone oscillator of Fig. 7, however, it is sufficient that either the input or the output of the active 11.18 elektor india november 1 985 Fig. 7 A parallel-mode oscillator can operate on harmonic frequencies (to which the LC circuit is tuned}, but this is rarely done in practical circuits. element is tuned to, or, rather, to just below, the harmonic, because the modulus of impedance must be capacitive. Although the circuit of Fig. 7 is suitable for operation at up to the fifth harmonic, it is rarely used in practice. Practical circuits The most frequently encountered oscillator using NOT gates (inverters) is 8 N1. N2 - 7.7404; Vj74LS04 85116 8 Fig. 8 One of the most common crystal oscillator circuits using TTL gates. shown in Fig. 8. This series-mode cir- cuit is suitable for operation between 1 and 8 MHz. Fine adjustment of the fre- quency is provided by trimmer C v If the required frequency accuracy is not great, the trimmer may be omitted. Capacitor C p prevents operation on a harmonic frequency. The value of /?, is 2k2 for operation below 2 MHz. The value of R 7 is calculated from /? 2 = 3000// : c IQI (6) where f c is the crystal frequency in MHz. A better oscillator using inverter gates is given in Fig. 9; it operates in parallel mode. Resistor R serves to limit the current through the crystal. If fine fre- quency adjustment is not required, C, may be omitted and C B increased to 56 pF. This circuit is suitable for qse between 1 and 30 MHz. Whereas the oscillators in Fig. 8 and 9 Fig. 9 Although this parallel-mode oscil- lator using low power Schottky TTL gates needs fewer components than its series- mode counterpart, it is hardly ever found in practice. are intended for use with low power Schottky - LS - gates, that in Fig. 10 uses high-speed CMOS devices. Relia- ble operation is guaranteed up to Fig. 10 When unbuffered high-speed CMOS gates are used, the parallel-mode oscillator is second to none. 30 MHz. The value of R is computed from /?=10 4 // c — 300 [SI (7) where f c is the crystal frequency in MHz. Since the output impedance of the gates is fairly low, it is necessary to connect a resistor in series with the output to prevent damage to, or destruction of, the crystal. It would be better to use a control system, but that is not easy to realize with gates. If, however, the oscillator uses discrete components, output control may be achieved in a convenient manner — see Fig. 11. The active element in this circuit is a dual-gate MOSFET. The interesting feature of this circuit is that the DC bias is obtained by feedback from the drain to gate G 2 - Diode D, in the feedback loop ensures that the potential at G 2 drops as soon as the signal at the drain exceeds 1.5 V pp . This causes a reduction of the current through T„ which limits the peak out- put of the oscillator. A high drain impedance is obtained by connecting a 10 mH choke, £,, in series with R 3 . Transistor T 2 is not required if the out- put of the MOSFET is AC coupled to a trigger circuit (LS or HC MOS). design ideas The oscillator of Fig. 11 is intended for operation with fundamental frequency crystals in the parallel mode over the frequency range of 0.1. . 30 MHz. It can, however, easily be modified for operation on harmonics. To this end, /?., is replaced by an LC circuit tuned to the desired harmonic — see Fig. 12. When this circuit is required to operate at frequencies above about 45 MHz, a BFR91 should be used in the T 2 pos- ition, and /? 5 should be reduced to about 220 Q. The value of i 2 is deter- mined by i 2 = 724// r c 2 |pHI (8) where f c is the crystal frequency in MHz. A Q value of around 30 is sufficient. The oscillator of Fig. 12 is suitable for operation at frequencies up to 100 MHz. The Z. 2 C„ circuit is tuned to the correct frequency, i.e., the third or fifth harmonic, with the help of a fre- quency counter. If such a counter is not available, connect the detector of Fig. 13 to the output and turn the trim- mer until the meter deflects. Turning the trimmer even further will cause the 0511613 Fig. 13 This detector circuit can be used to tune the L 2 -C x circuit in Fig. 12 if a fre- quency counter is not available. oscillator to switch off. The correct position of C % is about midway between the two points thus found. If this position is far from the centre pos ition, the value of L 2 is incorrect, or the circuit is tuned to the wrong harmonic. Fig. 14 This oscillator for operation on 3rd harmonic frequencies need not be tuned. Another type of third-harmonic oscil- lator is shown in Fig. 14. The resonant frequency of the LC circuit is ^ = 0 . 63/3 I MHz] 19) where / 3 is the frequency of the third harmonic in MHz. This circuit does not need trimming, because at the fundamental frequency it is inductive and can, therefore, not oscillate. At the third harmonic it is capacitive, however, so that oscillating is possible. The value of L is ca culated from Wm* Ub Fig. 11 The best oscillators still use discrete components; this circuit is intended for operation on the fundamental crystal frequency. 12 L =1616/f 3 2 l(jHI (10) where f 3 is the frequency of the third harmonic in MHz. M Fig. 12 A few small modifications make the circuit of Fig. 11 suitable for operation on harmonic frequencies. eleklor India november 1 985 1 1.1 9 high-resolution colour I graphics card - 2 ! the second in a series of articles describing a 512x512 or 512x256 pixel, black & white or colour, graphics card by P Lavigne & D Meyer Figure 4. Internal organization of the graphics display pro cessor. | A graphics display processor GDP) such as the Thomson Types EF9365, EF9366, or EF9367 is designed to generate the video signal and the sync(hronization) signals. To the user, it acts as an intelligent graphics screen controller with a picture scan that can be programmed via an eight-bit microprocessor. Apart from these func- tions, it also contains circuits for writing to the screen memory: a vector generator and a character generator. These enable writing on the screen at high speed (for instance, a 512 pixel diagonal in less than 700 ms) so that the-- host microprocessor is relieved of these basic tasks. The GDP has virtually no effect on the memory addressable by the micropro- cessor since it occupies only sixteen addresses: with the addition of external registers such as scroll, colour, and page switching, to no more than twenty ad- dresses. A further advantage is that the host and GDP memories have completely different cycles. This does not cause prob- lems when the microprocessor bus has to communicate direct with the screen memory since a special timing procedure prevents the disruption of the display. The GDP is programmed by 11 internal registers occupying 16 successive addresses. These registers may also be changed by the character or vector generators while a command is being executed. This means that the user does not have to input a new command or change the contents of any of these registers before the previous instruction has been executed. The GDP’s internal structure is illustrated in Fig. 4, while the pin designations for the Types EF9365, EF9366 and EF9367 are shown in Fig. 5. Only the functions and signals of the GDP that are relevant to the present article will be discussed. Power supply Sc logic levels A 5 V power supply is used, all inputs and outputs of which are TTL compatible. A high input level lies between 2.2 V and 5.0 V; a low level between 0 V and 0.8 V. Nominal current consumption is about 80 mA. Microprocessor bus The input/output (I/O) buffers on lines D 0 . . .D 7 (pins 33. . .26) are enabled by E ; the direction is controlled by R/W. Writing is indicated by a low logic level. The £ signal has the dual tasks of synchronizing and enabling communi- cation via the bus. The address of the register that is to be accessed is a pplie d to lines A 0 . . . A 3 (pins 9 . . . 12). The IRQ signal at pin 13 provides an interrupt request and is programmed by the CTRL1 register. 4 MSLO MSL3 X9 DADO OAM 6 vJ SVMC BLK MW all VB 1 1 .20 elektor india november 1 985 Light pen If a light pen is used, the WHITE on pin 24 forces the video signal to become- white. The signal provided by the light pen is applied to the LPCK (light pen clock) input on pin 21. The GDP then loads the current address into registers XLP and YLP. Sync(hronization) signals The line (horizontal) and field sync pulses (625 lines; 50 Hz) for the video monitor are provided by the SYNC signal on pin 34. All signals outside the display window are suppressed by the blanking signal (BLK on pin 25. The vertical blanking (VB) signal on pin 16 is high during the field retrace. Setting the parameters The format FMAT input must be con- nected as shown. FMAT vertical resolution processor 256 512 EF9365 0 1 EF9366 1 — EF9367 0 1 When the write only WO) signal on pin 23 is high, there is no display and the memories are not refreshed. None the less, the vector and character generators function as normal. Clock and screen memory addressing The general clock CK is input at pin 1; all internal signals are changed at the trail- ing edge of the clock pulse. It is used for multiplexing video memory addresses DAD 0 . . ,DAD 6 . When CK is low. the addresses of the row address strobe RAS lines are output on DAD lines A 0 . . .A 6 . The frequency of the clock is the same as that of the sync pulses. The display and refresh addresses appear on pins DAD 0 . . .DAD 6 in two steps; the maximum memory space is 16 K. Memory select lines MSL 0 . . .MSL 3 provide sig- nals that define a written pixel; they are needed to access a single of the eight bits addressed by DAD 0 . . .DAD 6 . A low level on the ALL line at pin 22 indicates that the operation in progress concerns all the memory banks (i.e., col- lective access). This is, of course, different to the bit-by-bit access provided by the MSL signals. Collective access is normally used for display, refreshing, or erasing. The data written to the memory consists of a single bit output by the display in (DIN) line on pin 15. If this pin is high, it represents a dark pixel on the screen. In a monochrome application, DIN could be the direct data input to the memories. For a colour application, however, DIN must be combined with the RGB inputs. Writing to the screen memory is enabled by the display write DW signal on pin 14. A low memory free MFREE signal at pin 19 indicates that the bit addressed by the microprocessor via a special instruc- tion is available at the output of the appropriate memory. This signal, there- fore, enables communication with the memory address indicated by registers X and Y — which can be programmed by the user — without disrupting the display. It always responds to an external request for access to the screen memory. All these signals will be met again later in this article. Screen memory If a picture consists of VH pixels, and each pixel can have 2* states — where V is the number of usable lines; H is the number of bits in a horizontal line; and b is the number of primary colours — the screen memory must contain VHb bits. Note that a pixel may comprise several bits: usually three, sometimes four. If H is large, the frequency of the video signal is greater than the maximum frequency for reading the memories. When, for instance, 77=512 and the normal line scanning fre- quency of 15.625 kHz is used, pixels appear at 70 ns intervals. A horizontal line is divided into h bytes of n successive bits that are read simultaneously onto the screen and converted by an autonomous (= dot clock + shift register) circuit into the video signal. The memory is, therefore, accessed it times per line. Each video signel b shift registers 5 CK C t V-/ 40 3 V CC DA Ob £ 2 39 3 DAOl DAD4 £ 3 38 3 DAD? DA 03 C 4 37 3 DADO DA 06 C 5 36 3 MSL 1 MSLOC 6 3S 3 MSL3 MSL2 C 7 34 3 SYNC FMAT C 8 33 3 DO AO C 9 32 3oi A1 C 10 EF9367 3, 3 02 A2 C 11 30 3 03 A3 C 12 29 3D4 IRQ C 13 78 3 05 DW C 14 27 3 D6 DIN C IS 26 D D7 VB C 16 25 3 BLK E C 17 24 D MW R/vi C 18 23 3 wo X9 C 19 22 D ALL vss C 20 21 3 LPCK 85080 Figure 5. Pin-outs of the Thomson-EFCIS graphics display processors. Figure 6. Schematic representation of the screen memory. elektor india november 1986 1 1 .21 Figure 7. Outside the display window, writing can take place at all times, except during the three refresh cycles shown. access loads every one of b registers with n bits. The memory then contains Vhb words of n bits as shown in Fig. 6. In the present circuit, the following formats are possible: b theoretically unlimited; in practice 0. . .4 (RGB + 1) Apart from the DAD 0 . . ,DAD 6 signals, used to address the h words of n bits, there are also MSL lines to select a pixel of b bits within the addressed word. Since /7 = 8, there are three MSL lines: MSL 0 . . ,MSL 2 : The MSL 3 line, used in the interlaced scanning mode ( V= 512), enables distinction to be made between odd and even numbered frames. It con- trols the A 7 line of the Type 4164 ICs. Refresh and display As stated earlier, the GDP effects three fundamentally different operations on the memory: display D which arranges the contents of the memory before making it visible on the screen; writing W ; and refreshing R . Outside the window, where the memory is used only for display and refreshing, writing can take place at all times, except during three refresh cycles as shown in Figure 7. The nature of the operation is indicated by the state of the BLK and ALL signals as shown. Figure 8. Correlation between the display and write addresses. h5 h4 X8 X7 n 8 H 512 V 256 or 512 8 9366 9367 (512 x 256) h3 h2 hi h0 V7 V6 V5 "v4 V3 V2 VI VO X6 X5 X4 X3 Y7 Y6 Y5 Y4 Y3 Y2 Yl | Y0 BLK ALL 0 o D 1 1 W 1 R i It will be seen later that there are excep- tions to these situations. 9367 1512 y 512) h5 h4 h3 h2 hi h0 V7 V6 V5 V4 V3 V2 VI VO t X8 X7 X6 X5 X4 X3 Y8 Y7 Y6 Y5 Y4 Y3 Y2 Yl YO Figure 9. Illustrating the discrimination between collective and pixel-by- pixel access. 9a 9366 9367 (512x256) i MSL DAD ALL CK 0 1 2 3 0 1 2 3 4 5 6 0 0 X xj X 1 h5 h4 h3 h2 hi hO ! VO 0 1 X X X 1 V7 V6 V5 V4 V3 V2 i VI b X don't care MSL DAD ALL 1 CK 0 0 1 2 3 0 1 2 3 4 5 6 xo XI X2 1 X8 X7 X6 X5 X4 X3 Y0 tt 1 X0 XI X2 1 Y7 Y6 Y5 Y4 Y3 Y2 Yl] c 9367 1512 x 512) 1 MSL DAD ALL CK 0 1 2 3 0 1 2 3 4 5 6 0 0 XO XI X2 VI h5 h4 h3 h2 hi hO VO 0 1 V7 V6 V5 V4 V3 V2 t 1 0 xo XI X2 Y2 X8 X7 X6 X5 X4 X3 Yl 1 Y8 Y7 Y6 Y5 Y4 Y3 YO Relation between DAD & MSL outputs and x, y coordinates Registers X and Y are twelve-bit read and write registers that contain the coordinates of the next pixel to be written onto the screen memory. They have nothing whatever to do with the video scanning functions but form the write address for the memory. With 2 x 12 bits, this address covers a space of 4096x4096 ( = 2 24 ) pixels. Only the least significant bits LSBsi are needed because the actually memorized image is of lower definition. None the less, the most significant bist MSBs are used, since they enable an image to be generated that is much larger than the actual size of the screen. The nine least significant bits of coor- dinates x and y are called X 0 . . . X 8 and Y 0 . . ,Y 8 respectively. The internal counters generating the screen memory addresses are organized into: ■ six line address bits (A = 64 words of n bits): A 0 ; A,; A 2 ; A 3 ; A 4 ; A 5 ■ nine field address bits (V = 256 or V=512): t, V 0 - Vi; V 2 ; V 3 ; K„; V 5 ; V 6 ; V 7 where t is the LSB that indicates the parity of the frames if V=512; when V=256, t does not exist. The correlation between the display address (bits A, V, and r) and the write address (bits X and Y) is given in Figure 8. The Type EF9365 processor is purposely not included in this figure, although it can 11.22 elektor India november 1 985 support either interlaced or sequential scanning, because it imposes a field resol- ution that is equal to the line resolution. ' The EF9366 provides sequential scanning only (K=256). The EF9367 can handle either 512x256 or 512x512 pixels. In the sequential scanning mode, the EF9366 is interchangeable with the EF9367. Everything dealt with so far has been fairly straightforward, but the procedure for assigning the address bits for display and writing to the processor’s output pins DAD and MSL is rather more com- plicated. So as to simplify matters a little, Fig. 9 discriminates between collective and pixel-by-pixel access. The sequential scanning mode, shared by the EF9366 and EF9367, is, as shown in Fig. 9a, a collec- tive access mode since ALL=0. During the first half of the access cycle, when CK = 0, and the system is in the display mode, lines DAD 0 . . . DAD 6 output horizon- tal address bits DAD 0 . . . DAD S as well as least significant field address bit V 0 on DAD 6 . This is important for the operation of the scroll circuit. During the second half of the access cycle, still in the display mode, the DAD lines provide the remainder of the field address bits. Note that all through this cycle, the logic levels on the MSL lines are irrelevant because operation is in the collective access mode. Pixel-by-pixel access is illustrated in Fig. 9b. The three LSBs of register X appear, therefore, on the MSL lines for selecting one of the eight bits in the word addressed by the DAD lines. Here again, the least significant vertical address bit is found in the first half of the access cycle along with the line address bits. This time, the MSL lines are used to address a single bit in the word addressed by the two bursts of signal issued by the DAD pins. Figure 9c shows the two memory access modes grouped together for an EF9367 in interlace d mo de (collective: ALL = 0; pixel- by-pixel: ALL= 1). The manner of dealing without output Xg of the EF9367 will be discussed later. In the Type EF9366, the memory is refreshed every two display lines (128 refresh cycles), whereas in the EF9367, this refresh occurs every four lines (256 refresh cycles). Dynamic random access memory A knowledge of the basic function of a dynamic random access memory RAM is essential for an understanding of the oper- ation of the present circuit and for seeing the importance of rigid timing of all signals. Each of the Type 4164 RAMs con- tains 2 16 (=65536 = 64 K) one-bit memory cells. The contents of each of these cells must be refreshed once every two milliseconds to prevent their corruption. A refresh is effected simply by a read operation. In principle, sixteen separate address lines are needed to access each of the 65536 cells. Fortunately, the 64 K bits are arranged in a matrix with multiplexed addressing. This means that the address of the matrix line row address on which the cell to be accessed is situated is specified first and then — on the same address line — the address of the column on which the cell is located. In this way, only eight address lines and two extra pins for enable signals are needed. These latter signals indicate when the addresses applied to the RAM are row addresses and when they are column addresses. They are called RAS (row address strobe = line address enable pulse) and CAS (column address strobe = column address enable pulse) respectively. The bar above the signals indicates inverse logic, i.e„ the signals are active when their logic level is low. Not all 65536 cells need to be accessed individually for a refresh: a collective refresh can be used in which only the 256 matrix lines are addressed. Manufacturers of integrated circuits endeavour to keep the 64 K RAM circuits compatible with their 16 K predecessors, which have only 2 14 (128x128 = 16384) cells, with the result that some 64 K memory chips need a re- fresh of only 128 instead of 256 lines. The relevant data sheet indicates this by the note "128 refresh cycles/2 ms", while the diagram of the refresh signals shows "A 7 : don’t care". The timing of the RAS and CAS signals, and the corresponding address signals, is important. Timing diagrams are shown in Figures 10 and 11. During a read or a write operation, the first signal to become active is RAS (pin 4 of a Type 4164). The RAM uses the eight logic levels present on pins A 0 . . .A 7 at that moment to address the relevant matrix line. A shor t time later, while RAS remains active, CAS also becomes active. The IC then sees the levels on address lines A 0 . . . A 7 as the column address, but — in terms of the 64 K memory — these are really address lines A 8 . . .A 15 . It is, therefore, necessary that the bits of address lines A 8 . . ,A 1S are applied to pins A,, . . . A 7 after RAS but before CAS. Note that with dynamic RAMs address signals A 0 .^A 7 need not be present before the RAS or CAS pulses, because / A sr and f ASC may be 0! 10 Figure 10. Timing diagram of the read cycle. • READ CYCLE A. J — 1 ZX X — x - wm /Mm X V»L>0 OAT FI Don » C«.« 85080-10 elektor india november 1985 1 1 .23 • WRITE CYCLE design is defaced. When the movement ceases, all defaced parts of the base have to be reconstructed. This complex oper- ation is carried out by an eight-input NAND, two AND, and two OR gates as shown in Figure 12. Writing to the screen is only possible when the WRITE signal has been enabled and the logic level of the data to be writ- ten D ln has been defined. In the RMW mode, this level is defined not only by the wanted result (pixel bright or dark), but also by the previous state of the element. If the pixel was bright before, it will be quenched; if it was dark, it will be bright. When the pixel appears on the screen, it is examined. If the screen is dark, the pixel appears bright, and if it is light, the 85080-11 Figure 11. Timing diagram of the write cycle. If t he above were a read operation, the logic level of the addressed bit would appear on the D ou , pin of the RAM a few nanoseconds after the address signals had been strobed. If the memory were written to, the WRITE line (pin 3 of a Type 4164) would be activated and the correct logic level applied to the D in pin shortly before CAS became active. For a refresh, the address line must be selected and RAS activated, when not only the timing of the refresh but also that of the RAS signal is important. Before the RAS signal can become active, and remain low during r RAS , it must have been high for at least r RP , i.e., the pre-charge time needed by the IC. Read-modify-write mode One unusual mode of operation is the read-modify-write RMW mode, in which a given address is read from, and written to, in the same access cycle. In Part 1 it was shown that, depending on the logic levels, six colours plus black and white can be produced, and further that a bright pixel has a low logic level, while a dark pixel has a high logic level. Every time an object is moved across a design (base) on the screen, part of that Figure 12. RMW selection in the monochrome mode is carried out by this array of gates. element is dark. The screen is returned to its previous state by redrawing all relevant pixels at the same position, bur with the logic level reversed. To make an object appear on the screen, the state of the relevant pixels is reversed. To revert to the original state, this oper- ation is carried out a second time. The successive inversions negate one another. This could be considered as an exclusive OR XOR function between the pen and the paper; if the paper is white, the pen is black; and where the paper is black, the pen is white. An RMW cycle starts with accessing the video memory for a read and finishes with a write. Between these actions, the addressed bit is modified. The eight-input NAND gate returns the addressed bit with its state inverted, while all other bits are forced high. When the read-modify-write-select RMWS line is high, it indicates that the RMW mode is active. If the line is low, the level on the DIN pin, which is provided by the GDP, is loaded into the screen memory when the data write DW pulse appears. The state of the screen is not taken into account then, since that is only of importance in the RMW mode. When that mode is selected, the combination of RMW S and load LD signals prevent the DW signal from reading the bit output by the NAND gate into the memory: otherwise, this bit would be inverted by the NAND gate and returned to the relevant RAM via DIN. If a bright pixel is required, the DIN line should be low; the corresponding logic level is loaded via D ln . If, on the other hand, the element is to be dark, the DIN line should be high, with the result that the corresponding bit in the memory is and remains high. Operation in the RMW mode can be disrupted if the delay inherent in the NAND, AND and OR gates does not cor- rupt the timing of the RAM. The shorter the f RAC signal, the more time there is for the RMW logic to change t he data a fid input this to D ln before the WRITE pulse arrives. f RAC is the time between the start of RAS and the appearance of data on D out , called ’’access time" on the relevant data sheet. In theory, f RAC should not exceed 150 ns (NEC4164-3; Hitachi 4864-2; 1 1 24 eleklor india november 1 985 Toshiba 4164-3; OKI 3764-15; MOSTEK 4564-15; and others), but practical experience has shown that the RMW cir-' cuit will work quite satisfactorily with access times of up to 300 ns. If colour is included, the RMW circuit is as illustrated in Figure 13, in which the gates specific to this mode are shown shaded. One of the seven gates is shared with another function, which will be reverted to in due course. The two extra gates increase the efficiency of the whole and do not increase the delay when a data bit travels from the output of a memory to its input. The circuit remains the same as for monochrome, except that some colour selection signals are added. These consist of red select RS ; green select GS ; blue select BS ; red write select RWS ; green write select GWS ; and blue write select BWS . Unlike DIN and DW, which are pro- vided by the GDP, these signals are con- trolled by the user via the colour register. If one of RS, GS, or BS is high, the addressed pixel will be dark as far as the corresponding colour is concerned, pro- vided the RWS, GWS, or BWS line is low to allow the memory to be accessed. If an RS, GS, or BS signal is low, and the associ- ated RWS, GWS, or BWS signal is also low, the corresponding pixel will have the rel- evant colour. All possible combinations with and without RMW (in monochrome) are listed in Table 2. 13 RWS lSw$) DW (BWS) memory is needed. The remainder cannot be used for colour, but it need not go to waste, since it is useful for storing several totally independent images that can quite simply — and immediately — be displayed on the screen. In one case, there are four pages available, numbered 0. . .3, depend- ing on the logic levels of multiplexed address signals A 7 and A IS ; in the other case there are two pages, numbered 0 and 1 as shown in Figure 14. Figure 13. RMW selection in the colour mode is car- ried out by this combi- nation of gates. Table 2. (DW 0> RMWS Dout - I DIN pixel 0 X 1 dark 0 X 0 bright 1 0 0 bright 1 0 1 (was dark) dark 1 1 X (was dark) dark (was bright) Memory capacity & image resolution If an image has a horizontal resolution (the number of pixels on a line scanned in the display window) of 512, and a vertical res- olution (number of lines scanned) of 256, the display window contains a total of 131 072 pixels, so that 16 K of memory is needed per colour. An image with a res- olution of 512x512 pixels needs 32 K of memory. Note that in the latter case there are, in fact, two interlaced images, each requiring 16 K. Doubling the vertical resol- ution increases the quality of the image, but also shows up an instability of the interlaced image, which is quite noticeable on static displays like those produced by the present graphics card. It is more visible on some monitors than on others. Whatever resolution is selected, no more than a quarter to a half of the 64 K 64 K x 8 bits 64 K x 8 bits A7 - 1 A15 ■ 1 PAGE 3 16 K 512 X 256 PAGE 1 32 K A7 -0 A15- 1 PAGE 2 16 K 512 x 256 512 x 512 PAGE 1 A7 - 1 A15 ■ 0 16 K 512 x 256 PAGE 0 32 K PAGE 0 512x512 A7 ■ 0 A15 * 0 16 K 512 x 256 85080-14 The use of A 7 and A 15 to switch pages introduces a limitation, namely that the type of RAM used should not require A 7 and A 15 to be refreshed; this prohibits, for instance, the use of the Siemens HYB4164 on the oranhics card. H Part 3 will appear in our December 1985 issue. elektor India november 1985 11.25 sound rotator by T S Norris & M M Bhalsod Audio signals produced by electrophonic instruments, such as electric pianos, organs, and synthesizers, often sound rather artificial and thin, lacking the natural colour of acoustic instruments. Early in electronic organ development, it was realized that a mechanically rotating speaker system (sometimes known as a Lesley speaker system) partially overcame this limitation by providing phase shifts that produce interference, both constructive and destructive, at a number of frequencies in the audio spectrum, which gives vitality and interest to the resulting sound. When two speakers rotate around one another, there exist constantly varying time delays - and therefore phase shifts - in the paths the sound takes to travel from the loudspeakers to the listener. A similar effect can be produced electronically by the use of a delay line with varying time delay. A charge-coupled device consists of an array of MOS capacitors that are coupled, so that charges can be moved through the semicon ductor substrate in a con trolled manner. Although essentially an analogue shift register for use in signal pro- cessing. the CCD can per form a wide variety of electronic functions. The schematic diagram of the sound rotating circuit, which uses a CCD (charge-coupled device) delay line IC, is shown in Fig. 1. The input signal passes along one route to the mixing amplifiers, and along another to the delay line circuit. Filters are required before and after the delay line, because a difference fre- quency, 4=/ s — 4 , is produced in the delay sampling process, where 4 is the delay line clocking frequency; 4 is any input frequency; and 4 is the difference frequency. Delayed signals u del and u^J are then mixed with the original input signal to produce output signals u^ and u odJ' A high-frequency oscillator provides a clocking signal, u dk , to pass charge packets along the delay line. The clocking signal is frequency-modulated by signal u m that is provided by a low-frequency sweep oscillator. Circuit description The circuit diagram of the sound rotator is shown in Fig. 2: ICj is the sweep oscil- lator; IC 2 and FFj form the clocking oscil- lator; IC 4 is the delay line; A, (IC 5 ) is the first, and A 2 is the second, low-pass filter; A 3 and A 5 are the mixers; and A 4 is the phase inverter. The frequency determining components for IC! and IC 2 are R\-C z and R r C s respectively, but the voltage at their pin 5 1 1.26 elektor india november 1985 also influences the output frequency. The frequency of IC, can be set between 0.2 and 6 Hz with P,. The shape of the output at pin 4 is triangular, but this is transform- ed to near sinusoidal by low-pass network -C 3 . The voltage across C 3 is used to frequency-modulate clocking oscillator IC 2 . The output of IC 2 is a rectangular voltage (at pin 3) at a frequency of about 120 kHz. D-type bistable FF , halves the frequency and shapes the signal to a square wave (duty factor=l:l). Delay line IC., is fed with both the Q and the 0 output of FF!; these signals are, of course, in antiphase. Opamp A, (IC 5 ) is a low-pass filter that reduces aliasing caused by the input signal being sampled at the 60 kHz clock- ing frequency. The filter has a cut-off fre- quency of about 15 kHz with a 12 dB roll-off. The filtered signal is applied to the input (pin 5) of the delay line via C 9 . The input pin is also provided with bias voltage preset by P 2 . The output of the delay line is fed to the second low-pass filter, A 2 , via SET BALANCE preset P 3 and C^. Low-pass filter A 2 prevents high-frequency clocking signals from reaching the mixer amplifiers; its characteristics are similar to those of filter A,. The original input signal is applied in equal proportions to the non-inverting inputs of mixers A 3 and A 5 . The level of the delayed signal fed to A 3 via R 22 , and to A 5 via R 2 s an d phase inverter A 4 , is adjusted by SET DELAY SIGNAL LEVEL control P 4 . Construction and testing The TDA1022 delay line should be treated with care, because it is a MOS IC that has no special protection against static elec- tricity. When the construction of the circuit has been completed, a 9 ... 15 V supply should be connected and the operating current measured: this should be around 25 mA, Note that once the preset controls have been set at a certain supply voltage, they may have to be reset if the level of the supply voltage is changed. Check with an oscilloscope that the signal at pin 3 of IC 2 is about 120 kHz. If the measured frequency is very different from this value, change R-, by trial and error: a smaller value for this component causes a higher frequency, and vice versa. At the same time, check that there is slight fre- quency modulation on this signal. Check with an anologue voltmeter — set to the 10 V AC range — connected between pin 4 of IC, and earth, that IC, oscillates. Also, check with P, that the out- put frequency is adjustable. Next, inspect the signal at the wiper of P 3 ; bias control P 2 should be adjusted to give a voltage of 5 ... 6 V at this point. The signal here should show some high- frequency clocking residue. Then, adjust P 3 to minimize the clocking residue on the signal at the output of A 2 . 2 Finally, apply a suitable signal (from a signal generator, organ, or synthesizer) to the input terminals, and adjust P 4 and P s to give the desired "sound rotation" effect to the signal at the output terminals. Finally The sound rotator is suitable for signal levels up to about 1 V r.m.s. and is intended to be installed prior to a power Figure 1. Schematic representation of the prin- ciple of operation of an electronic "Lesley system". Figure 2. The circuit of the sound rotator is based on a charge- coupled device Type TDA1022. elektor mdia november 1 985 1 1 . 27 Figure 3. The printed circuit board for the sound rotator. 3 Parts list Resistors: R, = 18 k R 2 33 k R3;R5;R/;Ri6;R2i 10 k R 4 3k3 R 6 1k5 Ra. Ra; R n ; R 13 . R i r. R is: R 22 : R23iR25- R 29 100 k Rio;R24;R30 Ik R 12 4k7 R 14 = 150 Q R 15 47 k Ri9iR20 - 470 Q Pi 10 k linear potentiometer P 2 10 k multiturn preset potentiometer P 3 - Ik multiturn preset potentiometer P 4 ;Pb 50 k logarithmic potentiometer Capacitors: Ci;C 4 = 1 n C 2 ;Cio 4 h 7; 15 V C 3 10 m; 16 V C 5 = 470 p C6;C9;Cn;Ci4;Ci7; C 2 o C 22 ;C 24 ;C 2 5 1 00 n C?;Ci 5 220 p C 8 ;Ci 6 100 p Ci 2 ;Ci 3 ;Ci 9 ,C 2 3 47 16 V C , 8 220 n Semiconductors: ICi ;IC 2 NE 566 IC 3 4013 IC 4 TDA 1022 (if this is difficult to obtain locally, it may be had from MB Instruments, 31 Ashvale, Cambridge, from whom other parts are also available) IC 5 = TL 081; TL 071; LF 356 IC 6 - TL 084; TL 074 Miscellaneous; 1C sockets as required knobs for potentiometers as required input, output, and battery terminals PCB 85099 amplifier. A PP3 battery can be used to operate the unit if it is, for instance, to be used in a foot-pedal or similar portable enclosure. Our tests show that the effect of the sound rotator is most noticeable with solo instrumental music, and least when the unit is used with popular music received from the radio. In this context, it should be noted that there is not really a "Lesley effect” with monophonic music, but rather a sort of vibrato mixed with phasing. None the less, the sound of small electronic organs, synthesizers, and electric guitars and pianos will benefit from the richness imparted by the sound rotator. N 1 1.28 elektor india november 1 985 stage lighting i ■ i by A Sevriens At this time of the year, thousands of amateur and school dramatic and music societies and groups are preparing for the coming season. A large number of these societies often wish they had a proper stage lighting installation, which permits the smooth changing and blending of lights and colours, the setting up of complete scenes with preset light intensities, and the smooth transition of light of two or more colours from one scene to the next at a controlled rate. The control panel of this type of installation can be located at a convenient place from where the operator can view the scene. Unfortunately, professional installations of this nature cost anything from a couple of thousand pounds upwards. The installation described in this and the next issue can, however, be built for a fraction of that cost. The control panel contains two separately preset light channels. Each lamp of the stage lighting system is controlled by two independent potentiometers, I and II in Fig. 1, which are connected into circuit by a main control switch (S 2 in Fig. 1). This enables the brightness of each lamp to be preset at two different levels, as selected by the two main potentiometers. In prac- tice, this makes it possible for a scene to be enacted, while the light levels for the next scene are being set up. These functions are shown schematically in Figure 1. The upper group in this diagram contains the two main controls, I and II, an audio input with associated electronic circuitry, and a change-over switch S 2 , which permits the desired blend to be selected. In position I, the two channels are preset independently and then mixed. In position I + II, fading of both channels is effected by preset I only. At one end of the travel of this control, channel I is operational, while at the other end, channel II is; at intermediate pos- itions, both channels are operational to a degree which is proportional to the set- ting. This position of S 2 , therefore, allows smooth transition from one to the other channel. The centre group of functional blocks in Fig. 1 represents the individual channels for each lamp. The lamps are connected in groups of three. Each lamp channel is provided with two slide potentiometers, which are fed from main controls I and II. The slide potentiometers enable the brightness of each lamp to be preset at two different levels. Each group of three lamp channels is accommodated on one elektor india november 1985 11.29 1 Figure 1. Block schematic diagram of a six-channel stage lighting installation. More channels may be added as shown. audio o phase 1 phase 2 printed-circuit board (PCB). Even in its smallest set up, the stage lighting control can, therefore, accommodate three spotlights. The bottom group comprises the power stages with triac control. These stages and ;he wiring of the system will be discussed in Part 2 in our December issuS. Suffice it to. say at this stage that each group of three triacs and the zero crossing control is accommodated on two PCBs. Zero crossing control ensures minimal-noise and interference being fed back to the mains, while the three triacs permit each of the three lamps to be connected to one phase of the mains. This in turn means that lamps of up to 5 kW power rating can be used. This is sufficient even for the Royal Opera House, Covent Garden! The power stages are, of course, controlled via opto-isolators, ensuring complete isolation from the mains. Main control The circuit diagram of the main controls and the required power supply is shown in Figure 2. This circuit, with the excep- tion of the mains transformer, is built on one printed-circuit board. The power supply provides a symmetrical IS V supply as well as a 10 V reference voltage (via IC 5 ). Great care has been taken in the design to prevent noise and other interference being fed back to the mains supply or into the control circuit, so that spurious operation of lamps is effec- tively obviated. The system can be audio controlled as 11.30 elektor india november 1985 2 Figure 2. Circuit diagram of the main control board. well as operated normally. The audio input is amplified and processed in opamp A,. Potentiometer which has an internal switch, Sj, can set the input sensi- tivity to about 100 mV. When Pi is turned fully anticlockwise, the audio input is short-circuited and St closes. This results in T, switching off and T 2 conducting. The inverting input of buffer amplifier A 2 is then no longer fed with the rectified audio signal; instead, the 10 V reference voltage is applied to it, which enables normal (manual) operation. Capacitor C 19 limits the base response of the audio circuit; if this limitation is not required, the capacitor may have a lower value or be omitted altogether. Preset P 2 enables setting of the drop-out time. The 10 V reference voltage is provided by IC 5 , which has a good temperature stab- ility. The exact value of the reference voltage is determined by Rg. Should the reference voltage not be exactly 10 V, the value of j? 9 should be altered by trial and error until a value of 10 V is obtained. A higher value of R a causes a higher value of the reference voltage. Opamp A z buffers the reference voltage, which is applied to its inverting input via T 2 and R b . Its output is applied to the two main controls, P 3 and P 4 , which serve channels I and II respectively. The con- trols are buffered by A 3 and A 4 respect- ively. Note that P 3 must be a stereo potentiometer, because in the circuit as drawn with S s set for normal operation, the buffers are fed separately by P 3 and P 4 . Channels 1 and II can, therefore, be mixed or set independent of one another. With elektor india november 1 985 11.31 15 V Figure 3. Circuit diagram of a lamp channel. s 2 j n ,h e other position, the control voltages for Channel I (A 3 ) and Channel II (A 4 ) are both provided by P 3 . The two halves of this potentiometer are in "opposition”, i.e., in one of the extreme positions, Channel I is operational, and in the other, Channel II. The output voltages from A 3 and A 4 are fed to the presets on the lamp channel board(s). Section of aluminium moulding which facili- tates the construction of the housing. Lamp channels The circuit in Figure 3 shows that only one opamp is used for the entire lamp channel. The input voltages at M, and M 2 are supplied direct to presets P, and P 2 respectively. The voltages at the wipers of these presets are combined and applied to the inverting input of IC, via diodes D, and D 2 . The diodes prevent interaction between the two presets and ensure that the higher of the two voltages determines the operation of IC|. The 0.6 V drop across the diodes is countered by that across D 4 in the feedback loop. With full drive, the output of IC, is 10 V. The resulting voltage at terminal C is a measure of the brightness of the lamp connected to this channel. An additional facility is provided by S,, a change-over switch with open centre pos- ition. As drawn, the voltages from the presets determine the operation. When S, is switched to earth, the lamp goes out, since the voltages from the presets are then virtually shortcircuited by With S, connected to the cathode of D 3 , the voltage at terminal C remains at 10 V so that the relevant lamp is held at full brightness. Finally Some thought should be given at this stage to the front panels. Figures 4 and 5 illustrate one possibility for the main and lamp control unit(s) respectively. Note that the position of the switches and poten- tiometers is determined by the printed- circuit boards. The sandwich construction results in a compact unit as shown in Figure 6. 11.32 elektor india november 1985 ILLUMINATOR Figure 4. One possible layout of the front panel for the main control unit Figure 5. One possible layout of the front panel for the lamp channel unit. 85097-4 elektor india november 1 985 Figure 6. The slide poten- tiometers are soldered to the track side of the PCB, which makes it necessary for their terminals to be bent as shown by the arrow. Figure 7. Illustrating how the PCBs are fitted to the front panel. Parts list (Fig. 8) Resistors: Ri = 100 k R 2 = 27 k R 3 = Ik R 4 - 470 Q R 5 = 10 k R 6 = 47 k R; = 22 k R 8 = 1k5 R 9 = 68 Q' Rio - 220 R n - . . R14 = 10 Q Pi = 47 k linear potentiometer with switch P 2 = 1 M preset poten- tiometer P3 100 k stereo slide potentiometer, linear, for PCB fitting P4 = 100 k mono slide potentiometer, linear, for PCB fitting Capacitors: Ci;Cio;Cn;Ci9* = 100 n C 2 ;C 3 = 1 p; 16 V C 4 • • • C7 = 22 n C 8 ;C 9 = 1000 p; 25 V Ci 2 ;Ci 3 10 p; 16 V C14. • Cia = 4p7; 16 V Semiconductors: Di zener diode, 10 V, 400 mW 0 2 = 1N4148 D 3 . . D 6 = 1N4001 T, = BS 170 T 2 = BS 250 ICi;IC 2 = .4558; 1458; TL 082 IC 3 - 7815 IC 4 = 7915 IC 5 - LM317T Miscellaneous: 5 2 single pole change-over switch 53 = double pole mains switch Fi fuse, 500 mA, delayed action fuse holder for Ft; single hole mounting mains chassis socket heat sinks for IC3 and IC4 mains transformer;; secondary 2x15 V; 2.5 VA or 4.5 VA knobs for slide potentiometers PCB 85097-1 Before the construction is started, it is necessary to be quite clear on the size of the required unit. As stated, in its smallest form, the installation offers three lamp channels, for which one main control and one lamp channel PCB are required. Where nine lamp channels are needed, two more relevant PCBs are required. The power supply shown in Figure 2 can cope easily with this number of boards. The housing of the control panel can be kept fairly flat and light as shown in Fig- ure 7. The power stages, which will be discussed in detail next month, are housed in a separate enclosure and con- nected to the control panel by a simple control cable. The wiring in the control panel is not critical, since the boards are interconnected by simple jumpers. It is essential that the potentiometers are of good quality, because "scratchy” ones may result in flickering lamps. M Part 2 will appear in our December 1985 issue. 'see text 11.34 elektor mdia november 1 985 Figure 8. Main control printed-circuit board. Figure 9. Lamp channel printed-circuit board. Parts list (Fig. 9) Resistors: Ri;R2 = 1 k R 3 = 22 Q R 4 ;R5 = 1 M R 6 ;R7 = 10 Q Pi;P 2 = 100 k mono slide potentiometer, linear, for PCB mounting Capacitors: Ci = 47 n C2;C 3 = 4 m 7; 16 V Semiconductors: Di. . D 4 = 1N4148 ICi = 741, LF 356 Miscellaneous: Si = single pole change-over switch with open centre position two knobs for slide potentiometers PCB 85097-2 NOTE: all these parts, except the PCB itself, are required three-fold for each board. elektor mdia november 1985 11.35 Anemometers are in steady demand by yachtsmen, wind surfers, glider pilots, and amateur meteorologists, to name but a few. The digital anemometer presented here can be built for about half the price of commercially available mechanical versions. An anemometer is a device for measuring the velocity of, among others, air move- ment, ie., wind. Its transducer (also called sensor), ie., the device that converts the wind speed into electrical signals, is usually driven by a small windmill, or a set of cups. The latter is used in the pres- ent design — Fig. 8. The sensor, at the left in Fig. 1, generates sixteen pulses for every rotation of the set of cups. These pulses, taken over a time hand-held anemometer by M Huch Fig. 1 Block schematic of the anemometer. determined by the control, are added in a counter. At the end of each counting period, the counter reading is applied to an EPROM; the output data of the EPROM are stored in a display driver. For every counter position, the EPROM has a code in two successive bytes, which determines what will be shown on the LCD display. Via switch S 2 , four ranges of the EPROM can be selected over two free address lines, so that the counter reading may be displayed in four different units; metres per second, m/s; knots; kilometres per hour, km/h, and the Beaufort scale. The conversion between these units is, therefore, not carried out direct by the unit, but rather by the programming of the EPROM. As the anemometer is battery powered, CMOS ICs are used to ensure a low cur- rent drain, while the EPROM is only switched on briefly at the end of a count- ing period to codify the counter reading. Circuit description Counters IC 1 and IC 2 in Fig. 2 control the operation of the circuit. A 4.433 MHz (tele- vision) crystal can be connected direct to the oscillator input of IC,. AND gates N, and N 3 generate the pulses required by the sensor. These pulses have a duty factor of 1:64. 11.36 elektor mdia november 1985 Fig. 3 Timing diagram of the sensor. The LED in the sensor is driven direct by inverter N 7 ; a limiting resistor is not required. When the light barrier is open, trapezium-shaped pulses as shown in Fig. 3 appear at the output of the sensor. These pulses are reshaped into rec- tangular form by inverters N 8 . . .N 10 . The resulting pulses are stretched by monostable MMV 2 in such a way that they become one pulse, whose period is equal to the time the light barrier is open. These pulses (sixteen per revolution) are counted in IC 7 . The measurement period is controlled by counter IC 2 ; the relevant timing diagram is shown in Fig. 4. The duration of a measurement may be set to Vz s, 1 s, or 2 s with link M. This enables the processing circuits to be matched to the transducer. As long as the output of AND gate N 2 is logic low, the signals provided by the sensor are processed and counted. When the output goes high, counting stops, the supply to the EPROM is switched on, and pin A 0 of the EPROM is made logic 0. Gate N 4 then generates a pulse which stores the contents of the memory addressed by the counter in the display driver in such a way that bits 0 ... 3 go to ICg and bits 4. . .7 to IC„. Pin A 0 of the EPROM then goes high, and the resulting pulse from N 5 writes bits 0. . .3 and 4. . .7 into IC 10 and IC l2 respectively. The supply to the EPROM is then switched off, the counter is reset by MMV,, and a new counting period starts. With switch S t closed, new counts are not loaded into the display. The previous information remains, therefore, displayed until the switch is opened. In this way, the EPROM programming can determine which units are to be displayed. 85093 2b The display The readings appear on a 3‘/z digit liquid- crystal display (LCD), which also shows LO BAT when the battery voltage is getting low. This type of display must be controlled with alternating current, since DC control causes damage by electrolysis. Special drivers are, therefore, used to provide an AC control signal. Every other output pulse of the four-bit latches in IC 9 and IC,| is inverted by the clocking signal at pin 2 before the output is fed to the dis- play. As the clocking signal is also applied to the backplane of the display, the segments that are in phase with the backplane (latch = 0) remain invisible, while the elements that are out of phase with the backplane (latch = 1) show up black. IC, 0 and IC U operate in an identical manner, but are decoded for a 7-segment display. The two centre digits of the display are binary coded decimal (BCD). The leading 1, the decimal point after the second digit, and the segments of the last digit can be individually controlled with the Fig. 2 (b) Diagram for use when rotary instead of slide switches are used in the Si and S 2 positions. elektor mdia novembar 1 985 A9 0 - — S~o 1 ° 1 1 1 1 1 O O _ a — 4 N 4 0>o_ N 5 0®_ _R_ J5L 1C 8 0® - A# measuring display __ measuring ' — ►t ' I'lMftr'M to »C» $nd 1C | , (D Tr 1^ Fig. 4 Timing diagram of the processing circuits. appropriate bits. This makes it possible for special symbols to be displayed in the last position. e.g., a ”b” for Beaufort, and a horizontal stroke for the other units. The sensor A thirtytwo-segment disc is rotated across the light barrier by the set of cups that is fitted to the same spindle — see Fig. 5. The thirty-two segments are equally div- ided into clear and opaque ones. A simple amplifier processes the signals provided by the photodiode. The LED is driven by 2 ps pulses provided by N 3 (pin 9). The duty factor of these pulses is 2:117 as shown in Fig. 3. The pulses provided by the light barrier are received by photodiode D 2 . Because of the delay in the diode, the signal at the non-inverting input of comparator 10,3 is sawtooth-shaped. The output of the com- parator is a square wave (duty factor 1:1). Preset P, should be adjusted so that with the light barrier open the signal at pin 6 of IC 13 is as clean a square wave as possible. Fig. 5 Suggested con- struction of the sensor and the carrier for the set of cups (scale = 13:20). cam*r i or Ml of cups light barrier Power supply The unit is intended for operation from four mercuric oxide button cells, whose e.m.f. ranges from S V to 7 V. This is no problem for the CMOS ICs, but does mean that a separate, stable 5.5 V supply is needed for the EPROM. This could, of course, be avoided by the use of a — rather more expensive — CMOS EPROM. Transistors T,. . .T., form a voltage limiter, whose output voltage vs battery voltage characteristic is shown in Fig. 6. The limiting threshold can be set with P 3 . If the base of T 4 goes high, the supply to the EPROM is switched off. Transistor T 5 functions as a battery voltage monitor. Preset P 2 is set so that T s cuts off as soon as the battery voltage drops below 4.5 V i.e., the level at which the elektor mdta november 1 985 1 1 .39 Fig. 6 Limiting characteristic of the supply voltage for EPROM IC 8 . Fig. 7 The double-sided printed-circuit board. EPROM begins to work unreliably. When T 5 does not conduct, pin 15 of IC 9 goes high and LO BAT is shown on the display. Construction The sensor is most conveniently made from an old cassette recorder motor from which all the innards, except the spindle, have been removed. To make the spindle rotate virtually friction-less, a small ball- bearing is added at the lower end as shown in Fig. 5. The sensor circuit is built on a circular (vero) board, which must be made to fit exactly in the motor housing. After the board has been wired up, it should be glued firmly in the housing as shown in 7a 6 5.5 V Uin-Ub 850936 b Parts list Resistors: Ri;R2;R7;R9.Ri2 = 100 k R3;R4.R5;Rii = 10 k R 6 1 M R 8 = 4k7 RiO = Ik R, 3 = 10 M Pi = 100 k multiturn preset ?2 = 50 k preset P 3 = Ik multiturn preset Capacitors: Ci = 22 n C 2 = 47 n C 3 = 10 n C 4 = 10 p; 16 V tantalum C 5 = 1 p; 16 V tantalum C 6 ;C 7 = 22 p Semiconductors: Di = LED; 3 mm; red D 2 - BPW34 T 1 BC 557 B T 2 . . .T 5 - BC547B ICi ; 4060 or 74HC4060 IC 2 .IC 7 = 4040 IC 3 = 4082 IC 4 - 4073 IC 5 * 4049 IC6 = 4538 IC 8 - 27(016 IC 9 MC 12 - 4054 ICio;ICn - 4056 1C , 3 = 741 Miscellaneous: LCD with LO BAT indication, e.g., Hamlin 3901 or 3902; or Hitachi LS007C-C or H1331C-C X, = crystal 4.433 MHz (television!) 4 button cells, mercuric oxide; 1 .5 V each S, = 2-pole; 3-position slide or rotary switch S 2 = 2-pole; 4-position slide or rotary switch PCB 85093 Note that IC 13 ; Re; D 2 ; and . P, are not intended for mounting on the PCB. 11.40 elektor india november 1 985 Fig. 5. Make sure that the preset can be adjusted from outside! The remainder of the sensor is built together as shown in Fig. 5. Make sure that the photodiode and LED are in a straight line. The LED should be prevented from radiating sideways with the aid of some insulating tape or suitable sleeving. The set of cups is made form three ping- pong balls fastened onto a carrier cut from a sheet of PTFE (Teflon) which itself is glued onto the pulley removed from the cassette motor — see Fig. 5 and Fig. 8. The ping-pong balls are cut in half along their seam, so that a reinforcing rim remains. The resulting half balls are best fitted in place with an epoxy resin. The printed-circuit board — Fig. 7 — has been designed to fit nicely in a hand-held case. To keep the completed board as flat as possible, it is best to solder the drivers and the display direct to it. If the pins of the display are too short, they can be lengthened with the cut-off pins of an 1C socket. It is, nevertheless, possible to fit the display and the drivers onto special flat sockets. The unit may be constructed in the hous- ing of a large electric torch, but a more professional case may be built on the lines of Fig. 8. Such a case is made from fibre glass, which is available in kit form from many car accessory shops for DIY car panel repairs. Calibration Before any of the ICs or the EPROM are inserted into their sockets, switch on the supply and set the voltage at pin 24 of the EPROM socket to 5.S V with P 3 . The ICs and the EPROM should now be inserted into their respective sockets (after elaktor India november 1 985 1 1 .41 Fig. 8 Suggested con- struction of the com- plete anemometer. the supply has been switched off). The voltage at pin 24 of the EPROM socket must now be checked with an oscillo- scope and, if necessary, P 3 should be readjusted to give 5.5 V. The supply voltage should now be reduced to 4.5 V, after which P 2 should be adjusted to just cause a LO BAT reading on the display. If the supply voltage is increased slightly, the reading should disappear immediately. Connect the sensor to the processing unit, and an oscilloscope to pin 6 of IC 13 . Turn the segmented disc so that the light bar- rier is open, and adjust P, for a clean trapezium-shaped signal on the oscilloscope. If an oscilloscope is not available, an EPROM programmed as shown in Table 1 may be used. The sensor should then be driven by a small electric motor at about 10 rev/s and P, adjusted until the display shows a steady reading of about 160. The IC 7 count, n, (which is directly pro- portional to the rotary velocity of the segmented disc) and the wind velocity, v, are related by v=nk where k is a constant. If at all possible, calibration of the display should be carried out in a wind tunnel but, unfortunately, this is not available to many. The next best thing is to visit your local meteorological station (consult your local telephone directory or ask your local library) on a number of days. Take a series of measurements, note the time, and then ask the station manager for the actual wind forces measured by his station. 0000 00 0F 10 0F 20 0F 30 0F 40 0F 50 0F 60 0F 70 0F 0010 80 0F 90 0F 00 01 10 01 20 01 30 01 40 01 50 01 0020 60 01 70 01 80 01 90 01 00 02 10 02 20 02 30 02 0030 40 02 50 02 60 02 70 02 80 02 90 02 00 03 10 03 0040 20 03 30 03 40 03 50 03 60 03 70 03 80 03 90 03 0050 00 04 10 04 20 04 30 04 40 04 50 04 60 04 70 04 0060 80 04 90 04 00 05 10 05 20 05 30 05 40 05 50 05 0070 60 05 70 05 80 05 90 05 00 06 10 06 20 06 30 06 0080 40 06 50 06 60 06 70 06 80 06 90 06 00 07 10 07 0090 20 07 30 07 40 07 50 07 60 07 70 07 80 07 90 07 00A0 00 08 10 08 20 08 30 08 40 08 50 08 60 08 70 08 00B0 80 08 90 08 00 09 10 09 20 09 30 09 40 09 50 09 00C0 60 09 70 09 80 09 90 09 01 00 1 1 00 21 00 31 00 0000 41 00 51 00 61 00 71 00 81 00 91 00 01 01 11 01 00E0 21 01 31 01 41 01 51 01 61 01 71 01 81 01 91 01 00F0 01 02 11 02 21 02 31 02 41 02 51 02 61 02 71 02 0100 81 02 91 02 01 03 11 03 21 03 31 03 41 03 51 03 0110 61 03 71 03 81 03 91 03 01 04 1 1 04 21 04 31 04 0120 41 04 51 04 61 04 71 04 81 04 91 04 01 05 1 1 05 0130 21 05 31 05 41 05 51 05 61 05 71 05 81 05 91 05 0140 01 06 11 06 21 06 31 06 41 06 51 06 61 06 71 06 0150 81 06 91 06 01 07 11 07 21 07 31 07 41 07 51 07 0160 61 07 71 07 81 07 91 07 01 08 11 08 21 08 31 08 0170 41 08 51 08 61 08 71 08 81 08 91 08 01 09 11 09 0180 21 09 31 09 41 09 51 09 61 09 71 09 81 09 91 09 0190 00 E0 10 E0 20 E0 30 E0 40 E0 50 E0 60 E0 70 E0 01A0 80 E0 90 E0 00 El 10 El E0 El 30 El 40 El 50 E 1 0160 60 El 70 El 80 El 90 El 00 E 2 10 E 2 20 E2 30 E 2 01C0 40 E2 50 E 2 60 E 2 70 E 2 80 E 2 90 E 2 00 E 3 10 E 3 0100 20 E 3 30 E 3 40 E 3 50 E 3 60 E 3 70 E 3 80 E 3 90 E 3 01E0 00 E4 10 E4 20 E4 30 E4 40 E4 50 E4 60 E4 70 E4 01F0 80 E4 90 E4 00 E 5 10 £5 20 E5 30 E 5 40 E 5 50 E 5 Table 1 Hex dump for the calibration EPROM. 1 1 .42 elektor india november 1 985 Table 2. Beaufort scale Description Wind speed m/s mph knots 0 calm 0 . .0.2 0 ..1 0 .1 1 light air 0.3 . .1.5 1 . .3 1 . 3 2 light breeze 1.6 3.3 4 . .7 4. 6 3 gentle breeze 3.4 . .5.4 8 .12 7. .10 4 moderate breeze 5.5 . .7.9 13 .18 11 . .16 5 fresh breeze 8.0 .10.7 19 24 17. 21 6 strong breeze 10.8 . .13.8 25 . .31 22 27 7 moderate gale 13.9 . .17.1 32 38 28. 33 8 fresh gale 17.2 .20.7 39 . .46 34 40 9 strong gale 20.8 . .24.4 47 . .54 41. 47 10 whole gale 24.5 .28.4 55 63 48. 55 11 storm 28 5 32.6 64 . 75 56 .65 12 hurricane 32.6 + 75 + 65 f Table 2 The Beaufort scale and correlated wind speeds. Plot the results of each series of measurements. The resulting characteristic should be a fairly straight line with a slight drop-off at low wind speeds (this is caused by friction in the sensor). The slope of the characteristic gives the value of k. If the measurements were suspect? calculate the constant k for each and every measure- ment from k-vlrt and from all the results establish a mean value for k. The values thus found, and their relation- ship, enable the EPROM to be pro- grammed in line with Table 3. Note that since S 2 enables the selection of any one of the address ranges 000 . . . IFF; 200. . . 3FF; 400. . .OFF; and 600. . . 7FF, four different codes (m/s; km/h; knots; and the Beaufort figure) may be programmed. M Table 31a). Address As. . Ap Bits ICg.lCll 7 6 5 4 ,3 2 1 0 L— first "1" decimal point after - 2nd symbol -not used _ character at 3rd position "(see Table 3b) lCio;lCi2 7 6 5 4 3 2 1 0, _character at 2nd position (see Table 3b) last position, segments c.e.f -last position, segment a - last position, segment g -last position, segment d Table 31b). binary 0000 0001 0010 0011 0100 0101 0110 0111 1000 1001 1010 1011 1100 i iei 1110 1111 alphanumeric 0 1 2 3 4 5 6 7 8 9 L H P A M a 4 □ DP Table 3 (a) Construction of a 512-byte block in the EPROM; (b) Correlation between binary code and alphanumeric character. elektor India november 1985 1 1 .43 luxury tester A highly important aspect of transistors is their current amplification ratio. This is often indicated by an A, B, or C behind the type number in the case of transistors. Inevitably however (by decree of Murphy himself), this is no longer legible. Using the transistor tester described here, the correct letter can be read from a display. At the same time, it can also be determined whether the transistor is up to scratch or not. R. Storn This particular design first appeared on Elektor's pages in last year's Summer Circuits issue. Readers voted it as one of the most interesting circuits and this article is the elected Elektorised result. The circuit has been slightly modified and a printed circuit board now accompanies it. In technical literature, the current amplification is usually indicated as tipp. For everyday purposes it is not absol- utely necessary to know the precise hpp value, but rather to have a rough idea of its upper and lower limits. The manufac- l c (collector current) and lb (base current). The Luxury Transistor Tester indicates the letter corresponding to the transis- tor's hpp category. Thus an A, B, or C will appear on the seven-segment display. An 'F' will appear, if the transistor is faulty. The circuit has separate connec- tions for NPN and PNP transistors. A switch selects the transistor type. The block diagram Figure 1 shows the block diagram of the transisfor turer used to have no way of precisely determining the current amplification ratio in advance. The best he could do was make a rough estimate, then after the transistors are manufactured, they were selected to meet the required hpp limits. The type number was then printed on the case. Although nowadays this can be determined in advance, the same type numbering is still used. Two transistors with the same type number do not necessarily have the same hFp. That is why industry uses a letter as a suffix to indicate the general hpp value. The letters define the hpp according to the following values: 'A' for an hpp between 140 and 270 'B' for an hpp between 270 and 500 'C' for an hFg of more than 500 The terms hFp and current amplifi- cation ratio describe the ratio between transistor tester. Its operation is quite simple. The voltage across a number of resistors is compared to a reference voltage. Here it is important to know beforehand whether the transistor is NPN or PNP. The switch that selects the transistor group also operates an LED to indicate the position of the switch. This voltage comparison determines the hpp group of the transistor and displays an 'A', 'B', or 'C' whichever is appropriate. If the 'F' on the display does not disappear when the pushbutton is depressed, then the transistor is defec- tive. The layout The complete layout is given in figure 2. Also shown is the parts list. The schmitt triggers in the block diagram consist of three op-amps wired as comparators. 1 Figure 1. The block diagram of the Luxury Transistor Tester. 1 1.44 elektor india novenrtber 1985 The upper half of the schematic, IC1 - IC3, serves to measure NPN transistors. The inverting inputs of the op-amps are connected to a reference voltage. The non-inverting inputs are connected to the collector of the transistor under test (TUT). Resistors are used as voltage dividers here. The base drive current is determined by R1 and RIO. At a certain amplification factor the collector current will also be fixed. Then the three collector resistors will be under a voltage determined by the current amplification ratio and the value of the collector resistor. If the amplification factor is 400 and the base drive is 1 0 pA, then the collector current will be 4 mA. With this amount of current flow, the voltage dropped across the collector resistor R4 (390 ft) will be 1.56 V. Three collector resistors have been included and they all have a certain voltage dropped across them. In the given example, R2 (220 fi) has 0.88 V and R3 ( 1 80 S2) has 0.72 V. As said earlier, R4 has 1.56 V dropped across it. This makes calculating the voltages at the 1C inputs easy. The inverting inputs are all at the same potential (or voltage). The voltage at the TUT's collector will be 9 V - 3.16 V = 5.84 V (the 3.16 V is the sum of the voltages across the resistors that feed the non-inverting inputs and the 9 V is the supply voltage). The reference voltage at the inverting input is 8.02 V which is determined by R5, R6 and R11, R12. In the earlier stated example, therefore, IC3's output will be low along with IC2's. Only ICI's output will be high. This is shown by this simple calculation: 9 V (supply) - 0.88 V (voltage at pin 3) = 8.12 V. 8.12 V is higher than the 8.02 V reference. If S3 is in the NPN position a B will appear on the display. If the output of IC1 werealso to go low, the display would then be C. This would be correct as the voltage drop across the resistors would have risen as well as the current through them. The base current in this circuit being always the same, the higher collector current could only be due to a higher current gain. If, on the other hand, the outputs of IC1 and IC2 were high, only segment d would not light and so an 'A' would appear on the display. Segments a, e and f are always on because they're used in all of the various letters that are displayed. All of the above is on the presumption that the transistor is not faulty, for if it is, then an 'F' will appear. This occurs only when all the 1C outputs are high — when the reference voltage is higher than the collector voltage of the TUT. The display control (the circuit consisting of T1, T2 and T3 along with resistors R 1 5 ... R 1 9, R24 . . . R26 and diodes D3 . . . D5) works quite simply. If the outputs of IC2 and IC3 are low, segments d, b, and c of the display are lit. The common anode of the display is, of course, always connected to the +9 V elektor mdia november 1985 11.45 \ ® 3 Parts List Resistors: R1 ,R10 = 820 k R2.R7 = 220 n R3.R8.R19 = 180 n R4,R9,R13,R14,R18,R20,R21, R22.R23 = 390 n R5.R12 = 1 k R6.R11 = 8k2 R15,R16,R17,R24,R25, R26 = 39 k Capacitors: Cl ,C2 = 1000 /li/1 6 V Semiconductors: IC1 . . . IC6 = 741 (Mini-DIPI T1 ,T3 = BC557B T2 = BC 547B D3.D4.D5 = 1N4148 D6 . . . D9 = 1 N4001 Dpi = LED-Display DL 707 Miscellaneous: Tr - sec. 2 x 6 ... 9 V/50 mA S1.S2 = Digitast and LED S3 - 4 pole two way Verocase 502 (75-3960E) of similar 11.46 elektor india november 198F 4 Figure 4. If the plastic case, that is mentioned in the text, is used, the whole transistor tester can be fitted into one case. Both of the LED's are mounted in the pushbutton (digitastl switches. supply. IC1 controls the three transistors. If ICI's output is high, only T2 will conduct so that segment g is connected to ground. Conversely, if ICI's output is low then only T1 and T3 will conduct with the result that segments b, c, and g are connected to +9 V and these segments go out. A similar situation occurs when S3 is changed to the PNP position. The outputs of IC4, IC5 and IC6 are then connected to the display instead of IC1, IC2 and IC3. Construction In figure 3, both sides of the printed circuit are shown. To make construction as easy as possible, the display and switches have been included on the board. Even the transformer can fit on it if a board mounting type can be found, otherwise a little tinkering may be necessary. The connections between the supply and the circuit itself have been deliberately omitted. This makes it possible to cut off the supply portion of the printed circuit board and mount it above (or anywhere else for that matter) the main board. The entire unit can be mounted in a Verocase type 502 (75 - 3960 E) or similar. Figure 4 shows how this is done. Switch S3 is a 4 pole 2 way and, if desired, can be attached to the printed circuit board. For this, a hole may be drilled into the board and the tumbler can be inserted without any difficulty. If this is done it should be possible to make a slot in the case's lid so that S3 may be operated. The switch's connec- tions must be wired to the circuit board. On the printed circuit the various connections have been marked in the same way as the switch in the parts list. Pushbuttons used to interrupt the base drive of the TUT should be of the digitast type. Below S2 are the connec- tions for the PNP and below SI those for the NPN. The pin assignment code is C = collector, B = base and E = emitter. The 1C op-amps are the popular (and inexpensive) 741 type. There is however one minor disadvantage to this, six IC's are necessary. By avoiding the use of 1C sockets (not needed in this case) costs can be kept to a minimum. It is however advisable to use a socket for the display. The transistors to be tested should, ideally, be connected to the board by means of clip leads. If this proves impossible, a transistor socket may be used, but this has shown disadvantages in practice. H elektor India november 1985 11.47 simple sound effects We have, somewhere in the Elektor laboratories, a sound effects department, although the exact location has yet to be discovered. There was a widely held opinion that it was found during the last Christmas office party but this was eventually discounted because a) the noises were too lifelike and b) it was not possible to simulate them electronically! We usually associate their normal products with the dying shrieks of tortured cats, horrifying howls and a whole assortment of plops, bangs, whistles etc. However on the odd occasion they do produce sounds suitable for publication and, to prove that this department really does exist, here is their latest circuit design. The original design for this rather clever sound effects unit was built into a 19inch rack mounting cabinet which unfortunately tended to overheat to an alarming degree (see ‘workshop heater’ in Elektor number 184) and lacked a little on portability. Further research resulted in the following circuit which uses only two CMOS ICs and is very cheap to build. Despite its modest dimensions it will produce a range of sounds from that of an American police siren to one closely resembling the ‘twittering’ of birds. Sounds simple? As is apparent from the block diagram of the circuit (figure 1), the basic principle is extremely straightforward. The output of a twelve-bit binary counter is converted into an analogue voltage which is used to control a VCO. As the binary output of the counter increases, the control voltage ramps positive, until the counter resets and the voltage falls to zero, whereupon the count resumes and the control voltage once again starts to ramp positive, and so on. The waveform of the control voltage is thus a periodic sawtooth. The VCO produces the actual output signal of the circuit, whose pitch is deter- mined by the instantaneous amplitude of the sawtooth control voltage. An output buffer amplifier ensures that the signal is sufficiently large to produce an audible tone when fed to a loudspeaker. The highly individual nature of the resultant sound is due to an unusual feedback configuration. The output signal of the VCO is not only used as the output of the circuit, but as the clock input of the binary counter. Thus the rate at which the counter steps through each count cycle depends upon the pitch of the output signal. In other words, the higher the sound, the faster it varies in pitch. The result is a repeti- tive beuip-beuip sound which starts each phrase at a low frequency and rises exponentially to a maximum pitch. Circuit diagram The circuit diagram of the sound effects generator is shown in figure 2, and as can be seen, it consists of only a couple of readily-available CMOS ICs and a few assorted resistors and diodes. 1C2 forms the 12-bit binary counter. The binary value of the 8 lowest order bits (i.e. those bits which change state most frequently) is converted into an analogue voltage by means of resistors R 1 . . . R8. The VCO consists of a simple CMOS oscillator (built round N1 and N2) the RC time constant of which is varied by using transistor T1 and a diode bridge as a voltage-controlled resistor. As the control voltage fed to the base of T1 increases, more current is passed through the diodes, with the result that their dynamic resistance falls. The initial frequency of the oscillator is set with the aid of preset potentiometer PI, which is connected in parallel with the diode network. The squarewave output of the VCO is fed both to the clock input of 1C2 and to an output buffer. The latter is formed by four of the remaining in- verters of IC1 connected in parallel. Construction A printed circuit board has been pro- vided for the circuit (see figure 3). As can be seen, due to the low component count, the board can be kept very small. The loudspeaker can be any inexpensive 8 S2 type capable of handling 500 mW. The supply voltage of the circuit can lie between 4.5 and 10 V; at the lowest supply voltage the current consumption of the circuit is only 5 mA, which means that a 4.5 V battery could be used, thereby rendering the circuit port- able. Note that the volume of the output signal is determined by the supply voltage level: the higher the supply voltage the louder the sound. The pitch of the output signal can be adjusted by means of F 1 . Since the pitch directly determines the rate at which the pitch changes, reducing the resistance setting of PI not only in- creases the pitch of the output signal but also causes it to increase more quickly. At the minimum resistance settings of PI the resultant sound some- what resembles that of a chirping bird. The value shown for PI in the diagram (1 Mfi) is chosen to give the maximum adjustment range. However, if desired any value from 10 k to 1 M may be used, with or without fixed series resistors. N 11.48 elektor mdia november 1 985 Figure 1. Block diagram of the simple sound effects generator. The output of a binary counter is converted into an analogue voltage which is used to control a VCO. The output of the VCO forms both the output signal of the circuit proper and the clock signal of the counter. Figure 2. Complete circuit diagram. Two CMOS ICs and a handful of discrete com- ponents are all that is required to produce an interesting range of sound effects. Figure 3. Printed circuit board for the sound effects generator, on which all the com- ponents, with the exception of the loud- speaker, can be mounted. The circuit can be battery-powered if so desired (EPS 79077). IOOOOO N1 ... N6 = IC1 = 4049 Resistors: R1.R9 = 820 k R2 = 470 k R3 = 220 k R4 = 100 k R5 = 47 k R6 = 22 k R7 = 12 k R8 = 5k6 PI = preset potentiometer, 1 M (see text) Capacitors: Cl - 120 n C2 = 100 p/16 V BC 547B 02Ok Semiconductors: IC1 = 4049 IC2 - 4040 T1 = BC547B, BC 107B or equ D1... D4 = DUS 79077 2 Miscellaneous: LS = loudspeaker, 8 fi/500 mW SI = pushbutton 3 RSj- IC2 4040 £ 100k J- 7 RJ r— j elektor India november 1985 1 1 .49 85108 The 1000 viewdata terminals recently ordered by a major British financial ser- vices company were the most impressive example yet of the spread of the medium into everyday use in the United Kingdom. Viewdata has ceased to be a monolothic product, and is now tailored to a series of individual marketplaces such as travel and tourism, stock ordering, financial data, and software downloading where its essential virtues of simplicity, network- ing, and colour recommend it to non- specialist users. Prestel, the viewdata system operated by British Telecom, is the largest of the private viewdata systems. There are other systems too that have always been operated by private companies, such as British Leyland's (BL) car sales system, operated by its computer bureau company Istel. Then there are hybrids of the two, based on private computers but connected to the Prestel network via a gateway computer link. Of the total number of terminals, 41 per cent are now in homes, against 14 per cent in 1982. However, it is a fair assumption that the revenues are still far more biased towards business user. Prestel has had some notable inno- vations in the past 18 months: • Its Mailbox electronic mail service went national, after a trial period in London. Some 40 000 messages are now being sent each week, a figure that has been growing by 12 per cent weekly. • Prestel Farmlink was launched, giving bdth local and national infor- mation, for example on crop problems, weather, market prices and pest warn- ings. Through gateways to other com- puters, the farmer can perform wages calculations and ration formulation analysis. • A nationwide theatre ticket booking service was inaugurated by a leading London ticket agency, Edwards and Edwards. • Prestel Homefinder was launched of the United Kingdom's three million home computer owners, and it will be a test of the recent €250 000 television and press marketing campaign for Prestel telesoftware to see whether this customer base can be substan tially enlarged. These developments are in addition to the Homelink telebanking service whose software has been sold to the Commonwealth Bank of Australia; the Citiservice of stock exchange, com- modity market and other real time prices that has been running for several years; and the extensive travel trade information that has character- ized Prestel since its inception, with 5400 British travel agencies now using the service. Prestel has found a more modest — but commercially more realistic — place than was originally anticipated in the British market for electronic infor- mation systems. But in assessing it, it is necessary to recognize its wider impact: Viewdata in Britain Several major British tour operators have arrangements of that kind and so far, Prestel has 18 such external gateways in operation. New markets So it is a mistake to equate United Kingdom viewdata with Prestel alone, just as it is wrong to condemn as unrealistic the original, over-optimistic forecasts of five years ago for the spread of Prestel to every home. The fact is that in Britain Prestel and viewdata in general have long since turned to different markets and marketing objectives. It is on these that the technology should be judged. Prestel itself continues to grow. At the last count it had 48000 terminals attached to its network and a database of over 330 000 frames of information (that is, screenfuls) on its computers. Despite the evident tribulations of its early years, it has been adding about 10 000 new subscribers a year, and there are now signs of a demand by the private as well as the business user. specifically for estate agents to exchange details of houses for sale. Home Computers Perhaps the most successful single service so far has been Prestel Microcomputing, based on subscrip- tions for which home computer owners can download* games and other software. In association with two telesoftware suppliers, Micronet 800 — the most popular single information provider (IP) out of Prestel's 160 con traded IPs — and Viewfax 258, this service has attracted about 10 000 subscribers. This figure, while proof of the appeal of the service, is still only a small fraction •A downloader is a relatively inexpensive program that converts the data from the com- puter being consulted into a suitable form for the receiving Ihome) computer. Downloaders are commercially available for a number of popular computers, such as the BBC micro, Sinclair Spectrum, Amstrad, C64, Electron, IBM, Apple, etc. Achievements • It has created a more variegated United Kingdom market for viewdata. • It has produced technical standards and standardization that have made it possible for viewdata as a mode of computing to spread to many new areas. • It has triggered interest in view- data in almost all industrialised countries. Many of them, through their national telecommunications auth orities (PTTs), have now adopted it as part of their own systems. • It was and is a test-bed for many new telecommunications services which are now standard parts of the value added services business, such as electronic mail, tele-ordering, gate- ways, and downloading. • It has blazed the trail for British Telecom as an information and value added services company, rather than a passive common carrier. Other countries experience viewdata Fig. 1 Basic set-up for communication with the telephone network. The floppy at the left is only required for downloading software. 11.50 elektor india november 1 985 differently. France has designed its own system, technically incompatible with Prestel, and has made it the cor- nerstone of its telematics pro- gramme. Based on small black and white Minitel terminals, the basic viewdata network is being created as a nationwide elec- tronic telephone directory, replacing paper directories in the home and office. This is an imaginative but expensive approach, because it has to be heavily subsidized by the PTT. About 300 000 Minitels are now in use, and over a million more are on order. The long run objective is that many other electronic information and trans- action services will be offered over the network, and, as in the United Kingdom, there is early evidence of business rather than domestic usage of the system, such as in banking. Substantial Use West Germany has not taken the elec- tronic telephone directory route, but has commissioned IBM to supply a nationwide viewdata system derived originally from Prestel. This has suf- fered technical delays, but has now been formally launched. Forecasts of its growth (as with Prestel in its early years) have varied enor- mously, although the West German PTT, the Bundespost, is saying that it will have one million customers by the end of 1986. Others think it may be a fifth of that figure. But it seems clear that in West Germany, as in France, viewdata will achieve substantial use, thanks to the impetus given to it by the national PTT. The British viewdata marketplace is not so unified. Alongside and inter- mingled with Prestel are the private systems, often operated by major trading companies. Examples are the information networks for dealers oper ated by two major motor manufac- turers, or the stock control system operated by the Debenhams depart- ment store group. Other systems are run by large com- puter bureau houses such as Thorn EMI's subsidiary, Datasolve. This has several tele-ordering services operating through a Prestel gateway, notably Reservision for hotel reservations and Teleordering for the book trade. Drugs and Banks In another area, the Baric computer bureau has been helping to conduct clinical trails of new drugs, by using viewdata as a way of collecting and collating test information from doctors taking part in the trials. Barclays Bank has its own in-house staff training scheme using viewdata screens, and the Bankers Automated Clearing Service (BACS) is using a private system supplied by Rediffusion to cope with the problem of redirecting back to its client banks queries that arise out of the millions of banking transactions processed electronically. The Bank of Scotland, in conjunction with Prestel, has been operating a Home Banking service throughout the UK since last year. This service enables private and business customers to check any aspect of their accounts, pay bills, and transfer money seven days a week, vir- tually round the clock. There is also a special Home & Office Banking Investment Account: a high- interest bank account into which funds can be put that are not required for immediate use. Armchair shopping Club 403, an "armchair shopping ser- vice" operated by Viewtel Services Ltd from Birmingham, could be adapted for use in Australia, New Zealand, Canada and the USA. Potential purchasers from these countries have been among the overseas visitors to Club 403 during its two-year trial with 1000 members. Fol lowing successful completion of the market test, Club 403 is now operating on a commercial basis in conjunction with Prestel. The service is aimed at people who do not have the time or transport to visit the supermarket themselves, and for the elderly or disabled for whom shop- ping is physically impossible. Club 403 is a package of regularly updated news, information and ser- vices available for the consumer at home. Using the latest technology a specially equipped television set can Fig. 2 Through Prestel's Homelink system, customers can order, and pay for. goods in the shops from the comfort of their own homes. access all kinds of local, national, and international news and order and pay for a wide range of goods and services. Club 403 members talk back to the information source to order goods or services by way of a simple alpha numeric keyboard. Services available include: armchair grocer, butcher, greengrocer, tele- florist, telebetting, telebanking, gour- met food service, electronic cookbook, motoring, gardening, education, jobs, library enquiries, and news pages. The most popular service is the "arm- chair grocer" which provides details of almost 10000 products from a local hypermarket. A catalogue is issued to members listing the available products with an individual code number. A member selects the items she wishes to order by keying in the appropriate code. A shopping list of goods required on a regular basis can be stored on the computer and used every time. Goods ordered are con- firmed by products description, price, and quantity on the television screen. The orders are received by the hyper- market, printed out, and sorted into a picking order to speed the collection of goods for packing. The grocery order is then despatched to the member's home. It is fair to observe that in the United States of America, where other styles of computer networking have become firmly established, the European idea of viewdata has not yet caught on. Viewtron, a service begun in Florida, has had to lay off staff after a poor take up. It is also true that Canada has not had much commercial success with its own viewdata design, known as Telidon, which is strong on graphics but very expensive. It would be false, therefore, to pretend that viewdata or any other form of electronic information system will meet all home and business needs. The results of using such a system will also vary from country to country. But what has been shown by Prestel and viewdata generally is that it has a rightful place among the technologies of the electronic information era. (LPS) elektor india november 1985 1 1.51* electronics select The employment of first generation robots without any sensory power has been described as employing deaf, dumb, blind, one legged human operators with their single foot nailed to the ground. Early applications such as spraying and spot welding required relatively low accuracy of arm position- ing, and even machine loading was based on an accurate location point or fixture outside the control of the robot itself. before removing them — although much more slowly, so far, than can be done by a butcher. In the drive towards more automation and the manufacture of high quality products, British companies are look- ing closely at what artificial vision systems can offer. Such is the interest that a new book, Machine Vision by J Hollingum, has just been published by IFS Publications, which is an associate company of the British there is an expanding programme of robotic studies, with over £330 000 received this year in contracts and grants. The main area of work is con- cerned with the integration of sensory feedback into robotic assembly systems. The exploitation of infor- mation obtained from a large number of sensors within an environment allows the work station to cope with an increasingly disorganized state, thus relaxing many of the otherwise stringent and expensive requirements for fixtures. Experience with analysis and implementation of robotic systems has A second generation of robots However, equipment is now being developed to provide such senses as sight, feel, and even smell. In the security field, for example, research is taking place on sensors that can recognize odours given off by explos ives. At London's Imperial College, a robot has been developed that detects the position of bones in sides of bacon Robot Association. Written in con- junction with British Robotic Systems, it reviews the scope for applications and suggests ways for companies to embark on the use of artificial vision. Sensory Feedback In the Department of Electronic Engineering at the University of Hull, shown that a set of well chosen "simple" sensors can provide all the information necessary for the system to determine the success of an attempted assembly. Such a capability is essential for truly flexible auto- mation. The department has developed a camera system of low resolution (up to 256x128 pixels), which is particularly suitable for such purposes as compo- nent inspection and identification. The use of a low cost dynamic random access memory chip as a light sensor provides a means of imaging not nor- mally obtainable by other means. While video cameras normally used for recording are based on charge coupled devices and cost up to £2000, the Hull University camera will cost about £100 — including interface with the robot's computer control. Its low resolution is quite adequate for robot assembly applications and simple quality control inspection. Using a combination of lasers and miniature cameras, the MetaTorch accurately accommodates variations in seam width and direction during robot arc welding oper- ations. Linear Array Processor Formed nearly four years ago, British Robotic Systems has concentrated on the application of vision systems to enhance manufacturing processes. Its first system, the Autoview Viking, is based on a linear array processor developed in conjunction with the National Physical Laboratory. The company has recently developed a more advanced system known as the Viking XA. The original Viking system consists of an industrial television camera, a minicomputer, a monitor, interface hardware, and a command console. Specialized handling equipment and illumination to suit the application is available. By analysing the image pro- duced by the camera, the system can automatically identify the charac- teristic shapes of manufactured parts and their position, orientation, and textural features, so that quality can be confirmed or the next operation in a process initiated. The sensors do this 11.52 elektor mdta november 1 985 by measurement and comparison of key features and such parameters as area, radius, centre of gravity deduced from two dimensional measurements, and surface contours. Recently accomplished tasks include automatic character reading and checking, component measurement, component indentification and orien tation, automatic crack detection and parts retrieval and machine loading from a bin. The Viking XA includes an additional set of powerful software commands incorporating a PROLOG artificial intelligence (All language interpreter, which introduces expert system based "help" functions and Al techniques. The system has full 512x512 pixel res- olution, an extended memory of 256 Kbytes, and an option for real time processing through the use of the lin ear array processor. Robot Arc Welding Having employed first generation robots in paint spraying, spot welding, and similar unpleasant jobs, it would seem most attractive to extend their use to arc welding. However, new problems arise in this application since the seam to be welded may vary slightly in width and direction. Work at Oxford University has resulted in a visual sensor that solves this problem. To exploit the development commer- cially, the university has set up Meta Machines to build and market the MetaTorch arc welding robot, which derives its "sight" from a compact combination of lasers and miniature cameras. These look a short distance ahead of the welding torch to locate the precise position of the joint, and the gap and offset of its edges. The information gathered, after com- plex vision and geometric processing, is passed to the robot arm and welding set. These enable the torch to follow the course of the joint exactly, and automatically adjust the welding con- ditions to secure a strong weld. As a result of this continuous feedback of information, the MetaTorch is able to cope with irregularity in the joint caused by factors such as wear in location pins and fixtures, imperfect components, and the unpredictable springback of pressed sheet metal. The robot can even cope with changes caused by thermal distortion with occur during the course of a weld. Although vision systems may speed up the application of robots to assembly operations, the design of products to suit production can increase robot effi- ciency without resorting to the inclusion of vision sensors. Some years ago, forklift truck manufacturer Lansing engineered a Unimate Puma robot without vision to produce wire harnesses used in the connection of electrical circuits in its trucks. More recently, BP Solar Systems has electronics sel ec t developed an assembly work station for its solar panels, of which it pro- duces 5000 a year. This is also engineered around a Puma robot manufactured by Unimation Europe and operates efficiently without vision. Vacuum Grippers Thought to be the only one of its type, the assembly robot commences its sequence by pneumatically picking up a fragile solar cell disc, using one of two vacuum grippers. It then orien tates the cell with an indexing mark printed on the disc face and solders conductive leads to the cell. The leads are then cut and crimped to allow for subsequent expansion and contrac- tion. After cooling, the cell is placed on a source of intense light and its electrical output is monitored to ensure that it meets the required level of performance. Finally, it is placed in one of thirty-six positions on an assembly table with its leads correctly positioned. When all thirty-six cells have been placed in position, the robot employs two hot air guns to solder the cells together into a continuous circuit. The entire process takes about 50 minutes. Many people regard vision and other senses as the most promising proper- ties in the robot market. Undoubtedly, intelligent robots will help to overcome the problems of the disorganized state A multi-exposure photograph showing movement of the Unimate Puma robot arm dur- ing the assembly of solar panels. The complex nature of the component handling devices engineered by BP Solar Systems can be clearly seen. of an assembly station where compo- nents have been designed for human assembly. In some areas it may be more econ- omic to design the components and the assembled product specifically for handling by a robot without visual or tactile senses. (LPS) by Arthur Fryatt elektor mdia november 1985 11.53 electronics select development. The fifth generation computer will Development Agency and with the The reputation of Professor Michie's soon be with us. But what exactly is backing of Strathclyde University, team has also attracted three United this wondrous device whose develop- which appropriately owes its origins in States organizations, including ment costs over the next 20 years will part to James Watt, the professor and Westinghouse Electric, to sign up run internationally into thousands of his team have set up the Turing under the terms of a recently created millions of pounds? Institute, named after the mathema- "remote affiliate" category. In fact it is not necessarily a computer tician Alan Turing. at all, at least in terms of the present He was responsible for the pioneering generation of work stations, pro- work on computing done at Bletchley cessors, and peripherals. Park during World War II, which con- Hands On Experience It could be a robot able to distinguish tributed to the cracking of the German The training programme for staff between the circuit board it is sup- Enigma codes. Professor Michie was seconded on a long term basis involves | transputer. It is a complete computer on a single chip, and is an essential J element in the fifth generation The thinking computer posed to be picking up and the lunch- time sandwich of the maintenance engineer, inadvertently left lying beside the production line. It could be a com- puter that talks and questions a management decision made when an executive is having an offday. Or it could be an expert in any subject known to man. Basically, though, it is a concept to amplify the intellectual brain power at the command of the worker of the next generation. It will accomplish this by an order of magnitude equivalent to the power that James Watt's steam engine provided for the workers of the first Industrial Revolution. Steam Engine The analogy with one of Scotland's greatest contributions to the modern world is more than coincidental. Just as the steam condenser was the key that unlocked the power of the steam engine, a programming language called Prolog, developed from work undertaken at Edinburgh University, is being used as the kernel artificial intelligence language by a large number of research groups in different parts of the world. Over the past 20 years, against heavy odds and with minimal funding, the pioneering group of artificial intelli- gence computer engineers from Edinburgh has been held together largely through the determination of one man — Professor Donald Michie. Recently, the team moved to a new office complex in Glasgow. Here it aims to establish a European centre for training and development in expert systems, machine induction, com- puter vision, and advanced robotics, which are the core technologies of the fifth generation computer. Code Cracking With funding from the Scottish one of Turing's young colleagues. The Japanese decision to set up a fifth generation computer project in 1982, involving eight major companies and two national laboratories with govern- ment and private funding, gave impetus to the fifth generation con- cept and sparked off the race to con quer the technology. The computer has been around for 40 years, but it is still a crude machine. It cannot understand normal speech or writing. Today's machines do exactly what they are told — no more, no less. The aim is to produce a computer that can acquire knowlegde and use it intelligently. There is now considerable haste in the West to launch programmes to sup- port computer research, and examples are the European Esprit programme and Britain's Alvey project, and various schemes in the United States of America. Private Funding The Turing Institute, apart from rela tively modest financing from the Scot- tish Development Agency, has taken the entrepreneurial approach with private funding and will raise the finance needed for its advanced research and development work by offering its unique knowlegde, its teaching capabilities, and its research facilities to individual companies and institutions through a system of industrial affiliates. Under this subscription system, companies pay for a range of facilities and can send employees to be tutored on artificial intelligence systems for applications in their own companies. During 1983, the institute planned for eight affiliates, all from Europe. They include GMD of Bonn, ICL, two Shell laboratories, Sinclair Research, and Thorn-EMI. The latter recently took over INMOS, the semiconductor company that has developed the formal instruction and hands-on experience. Formal instruction comes under three main headings of inference systems and logic programming, which are the basic underpinnings of the techniques used in the develop- ment of expert systems; computer vision, which takes in the principles of how a robot can be taught to recognize whatever is put before it, such as distinguishing between a cir- cuit board and a sandwich; and expert systems and robot planning. The institute was a pioneer of the use of inductive learning and is a world leader in the field. This involves literally teaching the computer to learn for itself on the basis of its cumulative acquired knowlegde. At the Turing Institute, inductive programming is the basis for expert systems, computer vision, and robitics. Affiliates also have access to extensive library facilities and to the research work of the institute. The institute organizes regular industrial seminars on individual aspects of artificial intelligence. Short courses and individual tutorials are planned to spread expertise in artificial intelligence systems into industry and commerce on an international level. Expert Knowlegde Dr Tim Niblett, director of industrial studies and the executive responsible for the scheme said: "As an institute we are focusing our attention on the synthesis of expert knowlegde. By transferring an expert's knowledge to a computer, the potential impact on society is at least as great as printing in relation to the book. "At the moment it is an extremely diffi- cult craft and the stage we are at now is where computing was 25 years ago. Asking what the problems are is really like asking what the problems are in the study of particle physics. Basically, it is making a computer understand ordinary concepts with all the com- 1 1 .54 elektor india november 1985 plexities of meaning, syntax, and implications. "The first generation of artificial intelligence programs involved a com puter specialist putting all the knowlegde of an expert into the com puter by hand. We are now at the sec- ond generation, where the expert inputs his knowlegde by showing examples of expert decisions. From these examples the artificial intelli- gence system hypothesizes expert rules, which then becomes accessible to any authorized person simply by asking the machine questions. "So the expert knowlegde is then available at all times, without the expert having to be physically pres- ent." The institute has already been directly involved in the first inductive learning expert system that can be used on microcomputers. The Expertease program is being marketed worldwise by an Edinburgh company, Expert Software International, which funded Professor Michie's consultancy company to develop the basic program commercially. Good Potential Dr Niblett added: "Expertease is a second generation product, with good potential in a number of areas such as diagnosis of faults, company pro- cedures and education, but its range is limited and we are now working on very much more sophisticated systems. "Inductive learning is a way of asking the computer to induce rules on the basis of how an expert in a subject displays his skills. Many experts do not know how they make a decision. The Expert System Program is designed to induce the rules by which the expert arrived at a particular decision. "The expert merely puts in statements of fact from his own knowlegde. At its simplest it would go something like this: 'It is raining. I will take an umbrella. It is cold. I will put on a coat'. The computer therefore induces the rule that rain requires an umbrella and cold a coat, and if it is cold and raining take both. Even at that simple level, the expert has not sat down and worked out how he came to the de- cision. It was common sense, but it is very difficult to teach a computer com- mon sense." Present day programs are totally inflex- ible. Given a task, the computer will attempt to pursue it to its conclusion even if the task is impossible. The fifth generation systems will pursue it by working out the best alternative to achieve the result required in the same way that the human mind works. Initially, the Turing Institute will have a staff of around 15, plus seconded staff from affiliate subscribers. One of the priorities is to attract world specialists electronics select in the field — many of whom left Britain in the years when artificial intelligence was a poor relation of computing science. Through its association with Strathclyde Univer- sity, the institute will have access to one of the largest and most advanced computer science departments in the country. The department is also involved in Esprit projects. Next year, the institute hopes to double the number of affiliates. A pro- posal has been put to the Alvey pro- ject directorate to contribute to the training costs of affiliate employees, and the institute will begin promoting the scheme seriously in the United States. Body Of Expertise Dr Niblett added: "The growth of artifical intelligence systems over the next few years will be limited only by the number of people able to utilize the technology. With the affiliate scheme, companies can second one person at a time over a year or two for six months each, building up a body of expertise within the company. "These people will then be able to take some of the 'shell' artificial intelligence programs and develop them for par- ticular applications within a company. There has been tremendous interest in the affiliate scheme despite the fact that we have not really gone out to market it yet. The American response was particularly encouraging. "I would estimate that within five years expert systems will have reached a stage of development in power and cost effectiveness to attract the mass user. Then the market will really begin to take off. At present, the cost of artificial intelligence is high and we are still in the early stages of developing the programming. It is a new style of programming and a new set of problems. "In Britain the demand for home com- puters has been the highest in the world. Once expert systems can be allied to that market there is no know ing in what direction knowlegde-based systems will go. "In the fairly near future, I would expect to see people starting to sell knowlegde bases either for use individually or using a centralized com- puter you can dial up for expert advice. The applications are limitless, but cer- tainly medicine and education are two areas with enormous potential." The Japanese fifth generation project is impressive in its scale and its inte- gration. But the Turing Institute will play its part for Britain and Europe generally. From a Scottish point of view, the institute, combined with projects such as the £3 million Alvey research and development programme at Edinburgh University on the speech driven word processor, gives the Scots the oppor tunity to contribute as much to the second industrial revolution as they did to the first. (LPS) by Maurice Baggott Professor Donald Michie. elektor india november 1985 11.55 Digi-Course Chapter 6 Binary Numbers In the previous chapters of Digi Course we have dealt with only logical problems. The circuits we discussed could solve simple logical problems like the one we discussed in chapter 4, about the cabbage, the goat and the wolf. For the solution of such simple problems, we really don't require electronics. However, for solving complex problems, we certainly need help of electronics. The most frequently used electronic apparatus for such applications is the pocket calculator. The pocket calculator also uses the same logical process that we have seen so far, to solve the problems. For instance, how much is 152132 + 75318 ? The result is 227450, and this is determined by the pocket calculator using the same logical process. HOW this is done is not within the scope of this chapter, and will not be discussed here. What we shall see here is the first and the last step of this calculating process, namely how the numbers from our decimal system are transformed and re- transformed into information which the logical gates can understand. The digital technology can work with only two values or conditions : On/Off, True/False, Black/White, 5V/0V, Logical "1" /LogicaT'O". The 0 and 1 can also be used as genuine numbers. The fact that we have to work only with two numbers, does not create a serious problem because similar to our decimal system, larger numbers can be made up using more number of digits. In the decimal system, any arbitrary number can be represented using the standard stock of 10 digits from 0 to 9. Similarly using the two digits 0 and 1 it is possible to represent any number, though the total number of digits required will be much more compared to the decimal system. This number system consisting only of 0 and 1 is called the binary system. The binary system requires considerably large number of digits, a simple "Two'' requires two digits, and is written as 10. In order to distinguish this binary 10, which corresponds to 2 in the decimal system, the basic number the intended numerical system is also indicated as follows : 102 = 2i0 Three is obviously the total of two and one. 3l0 = 2 10 + 1 10 now the expression on the right can be translated into binary with what has been studied so far : 2 10 = 102 ; 1 10 = !2 The total is : 310 = 1°2 = 12 = "2 1 1 in binary system is thus a 3 in the decimal system and not eleven. With a decimal 3 we are at the end of the binary numbers that can be expressed using two digits. To represent a decimal 4 in the binary system we need three digits. 4iq = 1002 It can be seen from this, that for every additional digit added, the previous digits on the right of the new digit are reset back to zero, exactly as in the decimal system 98. .99. .100. Once a digit is increased, the following numbers continue to build up according to the usual pattern: 5 10 = 4(o + 1 (o = IOO 2 + 1 2 “ 101 2 6(o = 4(o + 2(o = 1 0O2 + 1 02 “ 1 1 02 2 10 = 4 10 + 2 (0 + 1 (o = IOO 2 + 102+l2 = 111j From eight onwards, a fourth digit position becomes necessary. 8(0 ~ 1 0OO2 9io = 1 001 2 10,o = 1010, Beyond nine, the decimal system requires a second digit, whereas the binary system continues with four digits upto 15 (decimal). 1 1 10 - IOII 2 12,0 = IIOO 2 13,o = 1101 5 14,0 - 1110 2 15,o = 1111 2 16,o ~ 10000 2 By now it must have become clear, that the binary system is extremely difficult to handle and would certainly have not acquired any significance if digital electronics could not work so well with binary system. The digital technology also takes care of the translation between binary and decimal system. The decoding is theoretically very simple: since every binary position corresponds to a decimal value, one must add the decimal equivalents of the positions which are occupied by “ 1 " in the binary number. The decimal equivalents of the binary digit positions are : Here we shall now see a simple circuit, which can be connected on the Digilex board, to decode a two digit binary number into its decimal equivalent from 0 to 3. — l4-J> a a 1 >1)0 OUTPUT 1 >>>° — OUTPUT 2 > 1 JO OUTPUT I 3 11.56 elektor india november 1 985 The circuit for encoding a decimal number into its binary equivalent is relatively simple. One such circuit is shown below: A and B are the inputs of the circuit where B represents the right digit and A represents the left digit of the two digit binary number. Complete truth table of the circuit is given below. Table 1 . Outputs |F(1)|G(2)| Inputs While connecting the circuit on the Digilex board, the NAND gates will be used as inverters as usual. OUTPUT OUTPUT OUTPUT OUTPUT The circuit serves to encode the decimal numbers 0 to 3 into their binary equivalents. As we have seen, the first digit (A) must be 1 for decimal 2 and 3, and the second digit (B) must be 1 when decimal 1 and 3 are to be represented. For decimal 0, both A and B must be 0. This circuit can be connected on the Digilex board using NAND gates as shown in figure 4 A 101 101 in binary corresponds to : 32 +0+8 + 4 + 0+1= 45 in decimal. For electronically decoding and adding these equivalents, a complex circuit is necessary as the number of digits goes on increasing. The circuits for decoding binary numbers with more digits are complex and they are never built like this using individual gates. Practical decoders are available in form of integrated circuits (7442, 7445). For testing the function, outputs of the encoding circuit can be connected to the inputs of the decoder. Using two Digilex boards it is possible to transmit decimal numbers 0 to 3 on wires, using only 3 wires (A, B and earth) instead of 5 wires (0, 1 , 2, 3 and earth) that would have been required if decimal numbers were to be transmitted directly. rf'l $4 kf H k-f -J u dL UO ™ 1 ^ I \m« ' MS 1 + p r •\u3 U1 elektor India november 1 985 1 1 .57 designates each of these drawers. These drawer numbers are called addresses of the memory locations. The zero input to the encoder is left unconnected because when numbers 1 to 3 are absent the circuit automatically recognises 0. The decoder circuit shown in figure 5 can decode decimal numbers 0 to 5, using the three inputs for the three digits of the binary numbers 000 to 101 (ABC). In order to make this circuit work on our Digilex board, a small compromise has to be accepted. Here all the LED indicators will glow except for the one which indicates the decimal number being decoded by the circuit. Hexadecimal These 16 bit numbers are very difficult to handle when the programmer writes down the programs to instruct the microprocessor what to do. To simplify this task the Hexadecimal system is used, which is based on numbers 0 to 1 5. All these 1 6 numbers are written using only one position by replacing numbers 10 to 15 by letters A to F. Thus a Hexadecimal A is decimal 10, B is decimal 1 1, C is decimal 12 and so on. OUTPUT OUTPUT Using the Hexadecimal notation, 65536 addresses can now be written using only 4 positions. 0000 in Hex stands for decimal 0 and FFFF in Hex stands for decimal 65536. The Hex (Short form for Hexadecimal) numbers can be directly converted into binary numbers because each position in the Hex number directly corresponds to a 4 bit binary number. Figure 6 given below shows the pin connections of a BCD to decimal decoder chip 7442. OUTPUT Two of the inverters can be substituted by NAND gates with one input unconnected, the third one can be substituted by a NOR gate with one input on Binary Coded Decimal Frequently used in digital electronics is another system of representing decimal numbers. Binary Coded Decimal (BCD) is the compromise between decimal and binary systems. Instead of encoding decimal numbers into binary form completely, they are encoded position by position. BCD coded eleven thus becomes 0001 0001 and not 1011 as would have been the case in binary system. Ninety nine in decimal becomes 1001 1001 in BCD. In the binary system, every position can either be a " 1 " or a " 0 " and in digital logic every possibility of decision making is either true or false (1 or 0). These positions represented by the binary digits 1 and 0 are called Bits (from Binary Digits the letters B and it are taken to form the word Bit). The Bit represents the smallest unit of data that can be recognised by a digital circuit or system. A micro-processor which can process 8 of such bits of data at a time is described as an 8-Bit microprocessor. A decimal number represented in 8 bits can be maximum 256. These 8 bit microprocessors can operate with 65536 different memory locations (storage places where groups of 8 bit data are stored). To address these many locations the binary number required is a 16 bit number. If we imagine the memory chip as having 65536 drawers where data is stored in groups of 8 bits each, the microprocessor will be able to open any one of these drawers by using the 16 bit number which 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 _ 7 _ 3 4_ _5_ 6_ _7_ _8_ 9 _A B c D_ E V, J ! 1 1 1 1 r * i i * 1 \ i ' • * A* 4# 1 s t s l 4 sxji V* \ l ^ t f 1 1 .58 elektor India november 1 985 £ o' o '' o >' 'V'v: *' o'V o V>SJ y' o' . ‘ .'Vnfe/ V'>* ^ 'A ;<•■ s-> i %'^C. % t*’ . > s\ Constant Current Source: | 83639X1 Constant Current Sources A source which supplies constant current at its output, independent of load, is called a constant current source. Constant voltage sources are constructed mostly as complete equipment supplying constant voltage independent of the load. Constant current sources.however, are not constructed as independent equipment. They generally form a part of a larger circuit. Figure 1 shows the block schematic diagram of a constant current source. The source passes a constant current through the load resistor RL. From Ohm's law, we know that the voltage is the product of Resistance and Current passing through it. The voltage at the output of a constant current source thus depends on the load connected to it. The higher is this load resistance, the higher must be the output voltage to maintain a constant current output. Whenever the load resistance changes, the source must modify the voltage output in such a way that the current passing through the load resistance remains constant. Basic Circuit Figure 2 shows the basic circuit of such a source of constant current. The constant current is supplied to the load resistance RL by the transistor. RL may be any apparatus or equipment connected to the source. The diodes D1 and D2 as well as the base — emitter junction of the transistor are all forward biased, and they are conducting . Each of these diodes have a threshold voltage of 0.6V, thus, giving a total of 1.2 volts at the base of the transistor. As the base to emitter voltage is also 0.6V, the voltage on the emitter resistor Re is also 0.6V. Since the three threshold voltages are independent of the load, the voltage across the emitter resistor also remains independent of the load Ub * 9 V \ (BC 547) / 1°’ 0,^1 NPN-Trantiator |C1 ■ P 2 || o;sv D1 , D2, : 1 N 4148 -0 ! 83639X 2 I Ub = 9 V D1,D2: 1N4148 Figure 1: A constant current source automatically adjusts the output voltage in such a manner that the output current remains the same. Figure 2: The circuit of a constant current source using NPN transistor. The emitter resistance determines the source current. Figure 3: For testing the circuit, a linear potentiometer of IK is connected as load. elekior mdia novem&er 1 985 1 1 .59 and remains constant. Consequently, the current flowing through the transistor from collector to emitter must also remain constant, and will be given by the relation 0.6 V lc= _ ■ Red which means that the collector current which is also the load current, is decided by the emitter resistor. For instance a 5 mA current source can be constructed using an emitter resistor of 120 ohms. For testing the circuit, a IK linear potentiometer can be connected as load, in series with a multimeter (with 10 mA DC range). The current reading on the multimeter can be observed to be fairly constant and independent of the potentiometer setting. The circuit of figure 2 requires the load to be connected between the plus pole of the Supply and collector of the transistor. In case the load requirement is such that it needs a connection at the minus pole, then the circuit must be connected as shown in figure 4. The principle of operation remains same, only change being that the transistor used here is PNP-type. Mounting PCBs Many hobbyists find the installation of boards in a cabinet more difficult than the assembly of the circuit itself. Fixing small PCBs inside a cabinet can become much simpler if one can get hold of some double sided adhesive strips. For mounting a small PCB in the cabinet, 8 square pieces of such a strip will be required. 4 of these squares are fixed on the PCB, one at each corner. Small wooden or plastic pieces of about 4.5 mm thickness are then fixed on these four squares. Once again a square of the adhesive tape (double sided) is fixed on these insulating pieces. This provides your PCB with four self adhesive feet. The PCB can now be placed on the bottom plate of the cabinet and pressed firmly in the correct position. The drawing shows how the cross section will look. With this type of simplified assembly, all the drilling is eliminated. The soft adhesive tape also helps in dampening the vibrations of the board if the cabinet is subjected to mechanical vibrations. The circuit can be constructed as shown in figure 5. The source voltage to be used is a 9V battery pack. Figure 5 shows connection diagram for circuit given in figure 3 with NPN transistor. For circuit with a PNP transistor, the connection diagram can be suitably modified. Double Sided Adhesive Tape Figure 4: The circuit of a constant current source using PNP transistor, suitable for loads requiring a connection to the minus pole. Figure 5: The circuit of figure 3 assembled on a selex board. 11.60 elektor india november 1 985 selex AC Supply " Is it true that in the 19th century the DC current was used?'' Yes, that is true. At that time all the power supply was DC. There was no power supply network as we have today. Small power houses supplied electricity to a tew houses in the neighbourhood or to a factory. These mini power plants supplied DC voltage." "Actually I find this much more practical". "Why do you say so?" "Because for all electronic apparatus we require DC voltage, we can store DC voltage more easily in accumulators. Why were these DC power houses not continued?" "Well, I find AC current more practical. The first reason for this is that one does not have to pay attention to its polarity. Don't you remember? the AC current polarity changes every one hundredth of a second..." "Yes,... and on every terminal of the plug we get plus and minus alternately. But even for DC we can get plugs which are designed to fit only the correct way. They are also independent of the aspect of polarity." "What you say is correct, but another important feature of AC current is that it can be transmitted efficiently." "What do you mean by that?" "Transmitting efficiently means transmitting the current from one place to another without much loss of power. In the early days of electrical engineering the power houses had only small capacities. And also the distances over which these currents were transmitted were quite short. At that time it was unimportant, with which type of voltage the electricity was transmitted. However, today when gigantic quantities of current are produced and distributed over the power supply network, one must see to it that this is done in the most efficient manner." "We can use high voltage wires, that is obviousl" "Yes, you are right, but do you know why?" "Well, Yes, the high capacity has high voltage." Resistor In contrast to other electronic products like VCRs, Colour TVs, Computers etc., resistor is a product that is standardised for manufacturing. The International Electrotechnical Commission (IEC) has prepared different value tables, which all the manufacturers follow. This makes the task of a development engineer easy for selecting a particular value for his specific application. The most commonly used are E 6, E 12 and E 24 series values. Series E 12 contains 12 different numbers: 10 12 15, 18, 22, 27, 33, 39, 47, 56, 68, and 82. Multiplying these numbers by 0.10, in, 100, 1000, 1KO, lOkO, 100KO, or 1M XI, a wide range of resistors is obtained and these values are said to be I the E 12 series values. That is wrong. High capacity does not necessarily mean high voltage. The power output is the product of voltage and current. For example take a heater which takes 10 Ampere current at 230 volts. Its power consumption is then 230 V x 10A = 2300 W = 2.3 KW This power must be supplied by the power supply lines. One could also have a heater operating at 23 Volts and taking 100 Amperes, to give same amount of heat output." "But this has nothing to do with the high voltage wiresl” "You will understand this when we apply the same formula to the power generating equipment. Let us say a power plant has 230000 KW capacity. Now if you give this power through the overhead wires at 230 V, 1000000 Amperes of current will flow through these wiresl" "One million amperes? But even the thickest wire will be fused with so much current passing through." "Exactly, that is what will happen. So the only way to avoid this situation is to increase the voltage and reduce the current.” "But how can you increase the voltage?" "That is what makes the AC supply more practical. We have to increase the voltage for the high voltage overhead lines and then subsequently reduce it when it comes to the plug." "That is easy even with DC. Use two resistors as potential dividers!" "You and your potential dividersl To divide 230000 V to 230 V, only one thousandth of the power will be used and remaining power will be uselessly heated away in your potential divider. With AC currents, however, we can use transformers to step up and step down the voltage, without losing too much energy. These transformers unfortunately do not work with DC currents." "And this is what you meant when you said AC current can be transmitted efficiently?" "Right, you have finally understood!" Standards All the resistances are marked with coloured bands, first two bands including the standard values from the series, and the third band of colour indicating the decimal factor (i.e. the number of zeros following the standard value). A fourth ring indicates the tolerance of the component. Tolerance used for general purpose applications is usually ±5%. The resistor colour code was explained in our article on components. The unavoidable tolerances in the manufacturing process is one of the reasons for the peculiar values of the standard serieses. Each value of the E 12 series is about 20% higher than the previous value. This means that the values can deviate upto ±10% from their standard value without much overlapping. elektor India november 1 985 1 1.61 selex Resistor Combinations Parallel Combination The effective total resistance Rg of a parallel combination of two resistors R1 and R2 is calculated according to the following formula: F = HI • R2 J R1 + R2 Rather than calculating the value of the parallel combination every time, we can refer to table 1 and get the effective value directly. The table 1 given here corresponds to the E 12 series of resistance values. The decimal multiplicator is not included in the table, and must be taken care of while referring to the table. Three examples are given here to explain the application of the table: 1. Let us find out the equivalent value of the parallel combination of 47Kfl and 22Kn resistors. If we take out 1 0Kfl as the common factor from these two values, we have 4.7 and 2.2 to be searched in the table. Let us see where the row 2.2 meets column 4.7 and see what appears at the intersection. The value shown here at the intersection is 1 .498, i.e. approximately 1.5. Now by multiplying this by the common factory of 10KH which we had taken out, we get the effective value of the parallel combination of 47Kfl and 22KH as 1 5KO. This happens to be a standard value in the E 12 series and the parallel combination can be substituted directly by a 15KH resistor. 2. A 39KO resistor in parallel with a 8.2KH resistor has a common factor of 1 Kfl. At the intersection of column 39 and row 8.2, we have the value 6.775, which can also be substituted by a standard 6.8K11 resistance. 3. Let us now see how the same table can be used the other way round. Consider a situation where you have a 2200 resistor in the circuit and want to reduce its value to 2000 without removing the resistor from the circuit. The common factor here is 1000. Now searching for a value of 2 in the row 2.2, we find it to be in the column 22. Multiplying 22 by the common factor 100, we have the answer 2.2 K . Thus by soldering an additional resistor of 2.2K in parallel with the 2200 resistor, we can effectively bring down the value of the combination to 2000. Potential Divider A potential divider or voltage divider arrangement consists of two or more resistors in series, connected across the voltage source. Here, we shall see a simple arrangement with two resistors R1 and R2 connected in series and across voltage U -The voltage available at the junction R1 and R2 is U1 which is given by the formula: The value of U1 can be obtained directly by referring to table 2. (For E 12 series resistance values). On the left side of the chart are the distances of R2 from R1 on the E 12 scale and given in the middle of the chart are the ratios corresponding to the resistance values used. Following examples will explain how to use this chart: 1. It is obvious that a voltage divider made of two equal resistors will halve the input voltage U1 = 0.5 V. As the two equal resistances have a distance of zero steps between them on the E 12 scale, we read a ratio of 0.5 against a zero steps between them on the E 12 scale, we read a ratio of 0.5 against a zero in the chart. Table 1. ' i 1.2 1.5 1.8 2.2 IT EH! 3.9 ess 5,6 16^ 8.2 10 12 ■a ■El wm EDI 39 47 68 82 lOteid 023 0,688 EIPFil EKES 0.872 0.891 0,909 mn CHS feu 0.975 0,979 0,982 0,986 0,988 ua r*T- r -Vi (323 0,776 EE23 0,956 0,988 1.020 1,046 1.071 1,090 1,111 ilftfrH 1,148 1.157 1,164 1,170 1,174 1,179 1,182 va BBI [ 'JuZl [yjtiti 0.892 Ik'Ml PE*! 1,137 1,183 DF51 DE23 DE3 UtliJ 1,421 1.434 1,444 1,453 1,460 1 ,467 1.473 PM bb [llTlOl nttn nren BED DEZ3 PEET K53 1,687 1.706 1,720 1,733 UES I’M BS DE3 IlEKil tEEU 1,579 1,662 DEI 1,803 1.859 1,918 1,960 wvo 2,034 2,062 2,083 2,101 EXE1 2,142 EM 1,350 uca 1 1,714 1.821 1,932 FTTH 2,125 2,204 2,288 2.347 WtwJ 2,456 2.495 2,525 2,553 2.575 2,596 2.613 EM ■■■ hb bbi IBB IBB 1.650 1.787 1,938 2,076 2,221 ►**i 2,481 2.588 2,704 BE3 H#] 2,940 3,000 3,042 3,083 3,116 3.147 3.172 EM ^Bl BB IBB IBB E E3 2,131 2.298 2,478 FTTH 023 FTTF] K.i 1 RFIF1 EE3 EXH2 FEin CM BB BH IBB mm BBI 2,350 1 2,555 FTTH FTTrJ H-W:| DE3 DQD cuni OF??l 4.396 E3E23 EM 2.800 H'W'l fcfri 'ti EES3 £ f CSD nfjTl fi-vj 0ES3 p?m EM Finn □23 DEff] EES 5,194 5.431 5,638 5.790 l-H'll'l nrrci [EH LM i Mi 4.871 5,301 5,633 5.973 6,289 6,567 6,775 MS 7,152 7.317 '7,454 1 1 .62 elektor India november 1 985 2. An input voltage is connected across a voltage divider formed by R1 = 1 Kfl and R2 = 2.2Kfl. What is the voltage U1? (across R1). If we go from R1 to R2 on the E 1 2 scale we have to go four steps in the positive direction. Now from our chart, we have a ratio of 0.32 against the distance of 4 steps. This gives us the value of ul as: U1 = 0.32 x 10V = 3.2 V Table 2 3. If the voltage divider in the above example is replaced by another one, with R1 = 1.5KO and R2 = 390H we have to go backwards by 7 steps from R1 to R2. This gives a distance of - 7 steps on the E 12 scale. The ratio against - 7 steps distance is 0.79, which means that the voltage ul across R1 in this case will be Ul = 0.79 x 10V = 7.9V 4. A more practical example is where the ratio is known and we need to find out the combination R1 and R2. There can be a restriction of R1 and R2 so that the voltage source is not loaded very much by the divider. Assume a restriction to be R1 + R2 = 100KO, and the required ratio to be 0.2. From the chart, we observe that the nearest ratio available is 0.21, which can be tolerated. This ratio requires a distance of +7 steps from R1 and R2. From the E 12 scale we have one such combination of 22 Kfland 82 Kflwhich is the closest possible to satisfy the restriction to R 1 + R2 = 100 Kfl. As the deviation is negligible, we can take this as the correct solution. The last example has practical application for designing the voltage dividers using standard resistance values from E 12 series. Even though it is possible to calculate the values of R1 and R2 accurately using the formula forUI, it is practically impossible to obtain those exact values for R1 and R2 in most cases. The values obtained from table 2 will be practically available values, and the tolerances given by these are generally acceptable. The Digilex-PCB is now available! The Digilex-PCB is made from best quality Glass- Epoxy laminate and the tracks are bright tin plated, the track side is also soldermasked after plating. Block schematic layout of components and terminals is printed on the component side. Price: Rs. 85.00 + Maharashtra Sales Tax. Delivery charges extra: Rs. 6.00 Send full amount by DD/MO/PO. Available from: precious 0 ELECTRONICS CORPORATION Chhotani Building, 52-C, Proctor Road, Grant Road (East), Bombay 400 007 elektor india november 1 985 11.63 PCB artwork for main control circuit board and lamp channel circuit board for STAGE LIGHTING system. DIGITAL CONDUCTIVITY METER Century Instruments have introduced their latest model of Digital conducti- vity meter, incorporating cell constant compensation from 0.01 to 1 99. This cell constent can be adjusted with the help of a potentiometer monitoring it on digital display. This facilitates the conductivity measurements of solutions having low conductance values. The range covered is from 0 to 1000 mmhos and is divided into five ranges to get better resolution and accuracy. The same instrument can be used for measurement of resistivity. Automatic temperature compensation can also be provided for direct reading of conduct- ance For further information, write to: Century Instruments Pvt. Ltd. S.C.O. 289, Sector 35-D Chandigarh 160 036. UNIVERSAL BUSH MICROSIGN Universal Bush has unique flexible grommet edging which elimi- nates abrassion of wire insulation passing through cutouts in metal panels. The Universal Bush can be used with panel thickness upto4 9 mm It is very easy to install and ideal for all shapes. It can be easily cut to the required length and fitted without the aid of any special tools. For further information, write to: Microsign Products. “Mehta Terrace" Satyanarayan Road Bhavnagar 364 001. DIGITAL FREQUENCY METER BEACON’ Digital Frequency Meter Model LDFC-8004 is a 4 digit Frequency meter for monitoring Mains Supply Frequency with an accuracy of + 0.01 Hz. It is housed in a compact panel mounting case measuring 168 mm (W) x 87 mm (H) x 168 mm (D). The panel cutout required is 164 mm x 84 mm. The instrument works from 230 V AC supply with + 1 0% variations. It can also be supplied for other supply voltages such as 110 V and 440 V. new pr oducts For further information, write to: Beacon Industrial Electronics Pvt. Ltd. A-13, Shri Ram Industrial Estate, Katrak Road. Wadala Bombay 400 031. DIGITAL COUNTDOWN TIMER ION Electricals have developed a digital countdown timer which can be used with any electrical appliance which requires automatic switching OFF after certain time period. The timer is available with enclosure or without enclosure depending on the customer's requirement. CDT 59 is the OEM version, available without enclo- sure and CDT 59E is for direct application, available with.enclosure.A built in buzzer is also provided to indicate the time out signal. This facility can be used to utilise the timer as an elapsed time reminder. The timer operates directly from 230 V AC mains supply and can tolerate a variation of + 20% For further information, write to: ION Electricals, 307, Owner's Industrial Premises, Gabrial Road, Mahim, Bombay-400 016. UPS SYSTEM FOR TELEX SPECTRON have developed a new UPS system called UPEX for operating telex instrument. The system consists of a solid state inverter, battery charger and an electromagnetic change over relay. UPEX is designed to supply un- interrupted power to the telex instru- ment and accessories. In case of power failure, UPEX switches over the supply on the inverter without loosing even a single letter of the incoming or outgoing message. UPEX can also be used for PBX, PABX, Intercoms, Photo typesetting equipment and Electro medical instruments. For further information, write to: SPECTRON Sales & Serviec Pvt. Ltd. 28, Prasad Chambers, Karve Road, Pune-411 004. ELECTRONC TRAINING KITS Jetking Kits have introduced a wide range of electronic kits. These kits are offered complete with PCB, Enclo- sures, all components and comprehen- sive instruction manuals for teaching the beginners to build the circuit successfully. The manuals are written in an easy to understand manner and extensively illustrated. Guidance on the tools used as well as detailed description of the practical methods of assembling and soldering is given: For further information, write to: Jetking Kits India, 350, Lamington Road, Bombay-400 007. MODULAR COMPUTERS A new MODULAR COMPUTER’ is offered by Arun Electronics, in stan- dard 19" Rack with all PCBs in Eurobus standard. A printer, keyboard and video monitor must be connected to the rack to make the computer functional. The system has 3 slots for expansion by the user The modular computer can also be ' configured as a target system by using the necessary modules, such as Pro- cessor module, I/O module and Memory module. For further information, write to: ARUN Electroncis Pvt. Ltd. B/125, Ansa Industrial Estate, Saki Vihar Road, Bombay-400 072. elektor India nove^iber 1 985 1 1 .65 TEXONIC YOUR DEPENDABLE SOURCE Advertising Rates Our advertisers will be pleased to know that our circulation has now reached 1 7,000 copies with an average readership of 6 readers per copy. Looking at the statistics of rise in circulation, we expect the circulation to reach 20,000 copies by the end of 1 985. Our advertising rates have remained same since the first issue of Elektor India appeared on the scene in May 1 983. We are now finding it difficult to continue with the same rates in view of the escalating costs of eveiy input that goes into publishing Elektor India. Our advertising rates will be increased soon. Advertisers will be intimated through announcement in our comming issues. However, annual contracts will be honoured till the contract period is over. 1 1 .70 elektor mdia november 1 985 classified ads. CONDITIONS OF ACCEPTANCE OF CLASSIFIED ADVERTISEMENTS 1 ) Advertisements are accepted subject to the conditions appearing on our current rate card and on the express understanding that the Advertiser warrants that the advertisement does not contravene any trade act inforce in the country. 2) The Publishers reserve the right to refuse or withdraw any advertisement. 3) Although every care is taken, the Publishers shall not be liable for clerical or printer's errors or their consequences. 4) The Advertiser's full name and address must accompany each advertisement submitted. The prepaid rate for classified advertisement is Rs. 2.00 per word (minimum 24 words). Semi Display panels of 3 cms by 1 column. Rs. 150.00 per panel. All cheques, money orders, etc. to be made payable to Elektor Electronics Pvt. Ltd. Advertisements, together with remittance, should be sent to The Classified Advertisement Manager. For outstation cheques please add Rs. 2.50 Printed Circuit Boards and Kits for the projects published in Elektor India' and Electronics For You' available at SUGI ELECTRONICS, 2, Kamraj high road, Muthamil Nagar, Madras 600075 HURRY! One Chip Radio ICs'ZN 415E Rs 62.00; ZN 416 E Rs 78 00 (All inclusive). Send 50% with order (M O or DO) to: INTELLIGENT ELECTRONICS. 65, SBI Colony, Gopalnagar. Nagpur 440022 Available data sheet and application of any electronic components. Minimum charges Rs 15/- Write to DATA BANK. Plot No. 16. Bldg. No. 3, Flat No. 17, Bhavani Nagar. Marol Maroshi Road, Andheri (East). Bombay - 400 059. corrections Advertisers Index AFCO I. & C. LTD 11.08 APEX 11.70 ARUN ELECTRONICS 11.10 COMPONENT TECHNIQUE .. 11.04 DEVICE ELECTRONICS 11.71 DYNALOG MICRO SYSTEM . 11.76 ELCIAR 11.68 GENERAL ELECTRONICS ... 11.06 IEAP 1 1.72 JETKING 11.68 KEJRIWAL 11.72 KIRLOSKAR ELECTRODYNE . 11.07 LABELLA LABORATORIES . . 11.66 LOGICS 11.04 MICROPACK 11.06 MOTWANE 11.11 OSWAL ELECTRONIC CO. . . 11.08 PADMA 11.66 PIONEER ELECTRONICS 11.06 PLA 1 1 67 PULSECO 11.73 ROCHER ELECTRONICS 11.10 SAINI ELECTRONICS 11.72 SIEMENS '. 11.09 SOLDRON 11.73 SUCHA ASSOCAITES 11.66 TESTICA 1 1 73 TEXES TRANSFORMERS ... 11.04 TEXONIC 11.70 UNLIMITED ELECTRONICS .. 11.10 VASAVI 11-08 VISHA 1 1 02 ZODIAC 11.68 Components are normally available with the following companies: VISHA ELECTRONICS 1 7, Kalpana Building. 349. Lamington Road Bombay - 400 007 Phone: 362650 DYNALOG MICRO SYSTEMS 14, Hanuman Terrace, T ara T emple Lane, Lamington Road Bombay • 400 007. Phone: 353029. 362421 ELECTROKITS 20, Narasingapuram Street (First Floor) Mount Road Madras - 600 002 INTEGRATED ELECTRONICS INSTRUMENTS 8-2- 174 Red Cross Road Secunderabad 500 003 Phone: 72040 1 1 .74 elektor india november 1 985 to POWEI1 For Hig h Quality Switch Mode Power Sup pl y SWITCH MODE POWER SUPPLY FOR IBMPC/XT, APPLE II, COMPATIBLE SYSTEMS AND CUSTOM BUILT SYSTEMS OVERLOAD AND SHORT CIRCUIT PROTECTION FOR ALL OUTPUTS ADDMark R N. No. 3988 1/83 MH/BYW-228 LIC No. 91 Microfriend-ZLC Low Cost Dynalog has done it! E r U3/6L U4ISH A B 10/Rl SV/RH 6 7 R-C R-B 2 3 Smi | RG/YH & Cutting down cost without cutting into features All standard features available on Microfriend-ZLC ■ Z-80 CPU (a 4MHz, supported by Z-80 PIO and CTC. ■ 4K powerful monitor in 2732. ■ 24K EPROM and 8K RAM memory capacity on board. ■ On line EPROM programming facility in Single Step or Block mode for 2716, 2732, 2764 and 27128. ■ All Dynalog standard codes for Block Move, Fill Block, Insert, Delete, Search, Hex to Decimal, Decimal to Hex, Verify, Blank Check etc. ■ RS232C, 20 mA Serial Interface. ■ Audio Cassette Interface. ■ STD compatible Bus signals. ■ Most comprehensive documentation and Courseware 14, Hanuman Terrace, Tara Temple Lane, Lamington Road, Bombay 400 007 Tel: 362421 , 353029 Telex: 01 1 -75614 SEVK IN Gram: ELMADEV1CE Branches and representatives at: Pune. Bangalore. New Delhi. Hyderabad and Chandigarh Printer & Publisher - C. R- Chaoda rena, 2. Kouman. 14th A Road. Khar. Bombay-400052 & Printed et Trupti Offset 1 03. Vasan Udyog Bhevan. Off Tulsi Pipe Road ImwerParel Bombay-400 013