elektcf Volume 3 - Number 12 4 4 news & views editorial electronics scene electronics technology Artificial intelligence: a myth or a moneymaker? Capacitors & resistors: some recent developments . Fantasia on a MIDI theme Batteries: a danger to the environment? Computers aid survival of rare plants Moves towards a cashless society projects A star for Christmas stage lighting — 2 sweep generator minivolt overload protection for electric drills inverter play ball with Elektor! graphics card — 3 Indus information new products guide lines This month's cover shows a typical test set- up for the sweep Generator described on pp. 12-26 & 12-32. Also in the picture are the Function Generator and the uP-Controlled Frequency Meter described in the January and February 1985 issues of Elektor India respectively. switchboard classified ads advertisers index Selex — 7 Digi-Course II Audible Continuity Tester Potentiometers Polarity Tester 12.05 12.13 12.22 12.24 12.34 12.39 12.47 12-48 12-15 12-16 12-26 12.33 12-40 12.41 12-42 12-50 12-56 12-67 12-71 12-78 12-78 12-59 12-61 12-63 12-65 12-03 elekte? December 1985 r Publisher: C.R. Chandarana Editor: Surendra Iyer Editorial Assistance: Ashok Oongre Production: C.N. Mithagari Advtg & Admin: J. Dhas To all our readers: thank you! Editor: Len Seymour Elektuur 6.V. Peler Treckpoelstraat 2-4 6191 VK Beck ihe Netherlands Editor: PEL Kersemakers Elektor sari Route Nationals: Le Seau; B.P. 53 59270 Bailleul - France Editors: D R S Meyer; G C P Raedersdorf Elektor Verlag GmbH 5133 Gangell Postlach 1150 West Germany Editor: E J A Krempelsauer Elektor EPE Elektor JOE Via Rosellini 12 20124 Milano Italy Editor: D Fumagalli Ferreira & 8ento Lda. R.D. Estefania. 32 1° 1000 Lisboa - Portugal Editor: Jorge Gonpalves Ingclek S.A. Av. Alfonso XIII. 141 Madrid 16 Spain Editor: A M Ferrer In this, our last message to you in 1985, we want to express our heartfelt thanks to all of you who have loyally supported us throughout the past year. We never forget that it is essen- tially you who keep this wonderful magazine of ours going. We, for our part, will continue to do our utmost to give you the best we can from the fascinating world of electronics. There will always be something of interest for all of you, whether you work professionally in electronics, or are merely interested in it for your leisure hours: projects for home con- struction; articles of an informative nature on the electronics of today and tomorrow; news and views; and, of course, such regulars as New Products and New Literature. Our par- ticular aim is to offer a better balance between articles deal- ing with construction projects and those of a descriptive and informative nature. None the less, construction projects will continue to form the nucleus of the magazine. Furthermore, we are planning a Readers Letters column to enable you to express an opinion, or tell your fellow readers about an interesting or unusual aspect in the field of elec- tronics, or to exchange ideas with us or other readers. Again, we will be grateful for your support in this. There has, unfortunately, been one aspect of our services that has suffered through the restructuring of the editorial depart- ment and other changes that have taken place during the past twelve months: our response to your letters. Some of you must at times have felt a sense of utter frustration, if not of downright anger. For this, we apologize. At the same time, we would assure you that we are slowly coming to grips with the backlog, and hope to be back to normal by the beginning of the new year. Finally, to round off this message of goodwill, all of us at Elektor India wish you a Merry Christmas and a Happy New Year. led At: ti Ollset bay -400 013 Copyright Elektor B.V Goodbye ’85 Welcome ’86 Dynalog prepares to say goodbye to 1985, an year of great achievements, and to welcome 1986 ... with products to match new technologies. Dynalog wishes you a Merry Christmas and a Happy New Year DMS-Super80 The most powerful single board computer created by Dynalog's High-Tech R & D team. The DMS-Super80 satisfies all application needs where a high performance-low cost CP/M system is required. DMS-Super 80 is based on Z80A (4 MHz) CPU and operates with CP/M 3.0 Operating System. CP/M 3.0 is upward compatible with CP/M 2.2 and gives you access to the widest range of application software written for CP/M. Complete with on board Floppy Disk Controller, CRT Controller, Serial/Parallel Interfaces, Z80 CTC, 256 K DRAM and fully buffered STD Bus, the DMS-Super80 can be expanded into a fullfledged computer system. It is ideally suited for Industrial, Commercial, Educational and Development applications. MICROFRIEND -ILC Continuing the Dynalog tradition of cutting down cost without cutting into features, one more product joins the MICROFRIEMD Series. MICROFRIEND-ILC is a low cost training and development system based on 8085A CPC at 3 MHz. It has all the standard features of an 8085A system available on board. 4 K bytes of powerful Monitor Firmware resides in 2732 EPROM and supports all Dynalog standard codes, functions and interfaces. An added attraction is the on board - on line EPROM programmer with 28 pin ZIF socket. For information, write or call : Dynalog Micro-Systems 14, Hanuman Terrace, Tara Temple Lane, Lamington Road, Bombay 400 007 Tel: 362421, 353029 Telex: 01 1-75614 SEVK IN Gram: ELMADEVICE Branches and representatives at: Pune, Bangalore. New Delhi. Hyderabad and Chandigarh 1C testing at your finger tips ® APLAB 4026 Microprocessor based Automatic 1C Tester now available at amazingly low price. Now you can test your 1C. at fingertips, with APtAB 4026. It must be in your possession, if you are in R & D establishment, laboratory, electronic industries. ■ No reference 1C or Data Book required. ■ Search mode to identify group of unknown 1C. ■ Fast test mode for inward inspection and shop floor testing. ■ Single ., stepping mode for studying operation of digital ICs. ■ Zero-insertion-force DIL 1C socket to allow easy insertion and removal of the test 1C. APPLIED ELECTRONICS LTD. Aplab House. A-5/6. Wagle Industrial Estate. Thane 400 604. India. Phone: 59 18 61/2/3, 50 73 18. Telex: 71979 APEL IN. Cable Aplabthana NOW THE LATEST BOOKS-KITS ARE AVAILABLE WITH US DATA BOOKS : snjnotics te * a ^ s ™“„ ments SEMICONDUCTORS NATIONAL Q ^owoLA nc/i INTERSIL FAIRCHILD ZILOG - GE' - ANALOG DEVICES SIEMENS -INTEL SGS. ATES SILICONIX - TELEDYNE COOK BOOKS OSAORNE BOOKS SYBEX BOOKS - TAB BOOKS SAMS BOOKS - TOWERS BOOKS AND APPLE COMPUTER BOOKS PLEASE WRITE FOR DETAILS ELTEK BOOKS-N-KITS 6, RITCHIE STREET. 1ST FLOOR, M OUNT ROAD, MA DRAS-600 002 We also stock ELEKTOR INDIA kits & back issues. The Ultimate Personal Computer It will assist your wife in managing the domestic It will streamline the accounting system in your office Highly reliable and Inexpensive Sophisticated yet simple to use Touch typing (IBM type I Facility to use Audio Cassettes. Floppy Disk Drives and Printers Number of peripherals can be added Complete Software Packages available Does not require airconditioning SAVLA ELECTRONICS PVT. LTD. ' 6.. Shakuntal Shopping Centre Opp La Gajjar Bunglow. Ambawadi Ahmedabad 380 006 Phone 402860 It will remind you of your best friend's birthday LOGICS Authorised distributors for HCL Also available Thumbwheel Switches, IC’S, Transistors, Diodes. Tantalum Capacitors etc. ‘fairiwitl jg^corpor/ TRADING ^CORPORATION 69/75, New Bardan Lane, 1st floor, vadgadi, Bombay-400 003. Phone: 323333 329090, 329190 Cable: MATASHITLA Table Model LD 255 & LD 555 3Vz And 4Va Auto Manual Range Voltage Current Resistance Accuracy Optional 0—1000 Vts Ac/Dc 0—10 A Ac/Dc 0—20 Meg 0 ± 5°/o And 0± 0.05°/o Frequency Counter and Thermometer FROM LOGICS DIGITAL MULTIMETEF 12-10 eleklor in CLEANLINESS ” The Invisible Dimension of Quality —I 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 1 00 Clean Airin e Removes dust particles down to 0.3 e 99.97% efficiency e Available in size- 610x610 mm 1220x610 mm. 1830x610 mm e Table Top Models also available 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 Bio Hazard Unit with U/V Light/Virus burn unit. e Ideal for safe microbeal research e Available in size -610x610 mm 1220x610 mm, 1830x610 mm Chemical Benches are also available Portable Electronic Cleaner : • Works with high efficiency at high For fume clearing and recovery of temperature suspended particles in pharmaceutical • Can be coupled to air conditioning or industry Handles 1 5 to 1 80 M’/minute airhandling systems air • Low on power consumption. Kirloskar Electrodyne : Total Capability In Clean Air Systems • Easy to maintain # Long life KIRLOSKAR ELECTRODYNE PVT. LTD. 118, General Block. M.I.D.C. Bhosari. Pune 41 1 026 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 »s 12-11 electronics scene EEPROMS Excel Microelectronics Inc., believed [ to be the pioneering company in making electrically erasable memory chips, is considering a proposal to set up a subsidiary in India, according to reports. Mr. B.K. Marya, an India-born American citizen, who is president of the company, is quoted as saying that he is interested in India primarily for two reasons namely cheap labour and highly qualified technical talent, the Indian subsidiary will be entrusted with the job of silicon wafer fabriction which can be finally converted into "Electrically Erasable and Programmeable Read Only Memory" (EEPROM) chips. Since the Indian market was not yet ready for using the EEPROM chips, the company would have 100 per cent buy back arrangement. POWER ELECTRONICS A national seminar on the role of electronics in the power sector, held recently at Delhi, has adopted a set of recommendations which have been submitted to the government for consideration. The seminar organised by the National Council for Power Utilities and "Urja", a journal devoted to energy, called upon all power generating and distributing organisations to integrate electroics in their power distribution and monitoring systems and in their communication networks. The seminar opined that the power sector should go in for digital communiction and low-cost satellite terminals to be used at power stations and at large grid substations to transmit data. An action plan for the implementation of an integrated communication system for power stations management should be drawn up at th earliest, the seminar concluded. For adopting computerised and microprocessor-based distribution systems by the power utilities in a phased manner, it has been suggested that one 210 MW unit and one 500 MW unit should be earmarked as units on which new electronics and computer applications could be tried out for their usefulness under Indian conditions. TCIL DIVERSIFIES Telecommunications Consultants India Ltd. (TCIL), an organisation under the Union ministry of communications, is poised for making feasibility studies for the introduction of cellular mobile radio communication system, view data, yellow page service and office automation. The organisation is currently carrying out a feasibility study on digital transmission, telemetry and telecontrol for oil pipeline and gas pipeline system. It is also setting up a computer network on an experimental basis for exchange of information among computers in three cities - Delhi, Bombay and Madras. TCIL, which has so far been operating in conventional fields like switching, transmission and satellite communication, has recorded the highest net worth growth rate among all public sector undertaking, according to Mr. M.P. Shykla. chairman and managing director of the company. Mole saves British Telecom £10 million a year A £100 device that could save about £10 million a year in maintenance costs has won its inventor first prize in British Telecom's New Ideas Com- petition. Ernie Huggins, a 59 year old assistant executive engineer, worked for two years in his spare time to conceive the Mole, an electronic locator which can pinpoint faults in underground cables simply and accurately. It reduces the number of holes that have to be dug for each fault from five to two, saving money and reducing disruption to the public. There are already 3000 Moles in use by. British Telecom throughout the UK, and licences are being arranged for its manufacture and sale throughout the Runner-up was Jed Isbell, a 29 year old manager from London, who devised a testing system that will make British Telecom's Packet SwitchStream net- work even more reliable. Highly commended were Perry Beb- bington and Peter Mosely, both techni- cal officers from Nottingham, for an idea that enhances the compatibility between the Monarch and Herald elec- tronic switchboard systems. Database helps choose computers A computerized database, which allows any company considering com- puterization to have a specification of their business requirement to be fed into the system and matched against existing available systems, is being offered to computer consultants throughout the world under licensing agreements. The database, Computerscan, is a fourth generation system that is able to give reasoned, intelligent answers about information stored in its memory. Its was developed by Atlas Computer Consultancy (UK) Limited of Preston, Lancs, and has been tried successfully by many companies in Britain. The database has details of over 3000 suppliers, incl. 60 000 application soft- ware packages. This information greatly exceeds that which a computer consultant could analyse during research. In addition to licensing agreements, the company has undertaken a period of investment to enable a worldwide consultancy service to be offered from its headquarters. Assistance has already been given to the UN in Geneva, as well as to organizations in New Zealand, Nigeria, Dubai, and Australia. (LPS) Thick film in cars The new Granada Scorpio, the latest model from the Ford Motor Company, owes one of its smallest, but most vital parts to Swindon-based BICC-CITEC. This company has designed and manufactured the thick-film fuel tank sender resistive element which indicates not only the amount of petrol in the tank, but also enables the miles per gallon ratio to be calculated by the in-car computer. The element consists of a thin ceramic tile on which the resistive track is screen-printed with a specially devised cermet ink that is capable of with- standing both the corrosive environ- ment of blended petroleum and the constant track wear caused by the wiper. BICC-CITEC has been developing sender elements for Ford for three years, beginning with the launch of the Sierra, and currently produces around 250000 elements a year for the Granada, Sierra, and top-of-the-range versions of the Orion and Escort models. 12-13 electronics scene ; CHIPRIGHT ACT Any invention needs a protection ,from illegal imitations and the interest of the inventor is safeguarded by what is known as the Patent law. In the computer industry, the piracy of computer software programmes threw up a new challenge as the product sought to be protected here is not a design of a watch or a television but an intellectual thought. Legal provisions have been evolved in countries like the USA to check the software Of late, semiconductor integrated chips have also become a victim of intellecutal plagiarism and the original designer is left with little profits as his designs are proliferating illgally in the hands of ace copiers. The ICs play such an important role in various fields that experts call it the "crude oil of electronic industry" to signify its economic potential. The major part of development cost of new ICs goes for the enormous amount of time and effort spent on designing the circuit layout. As the degree of integration of thousands of elements together on a tiny semiconductor substrate progresses, the developmental costs will also increase. At the same time, if competitors copy the circuits without incurring the developmental costs, the return for the original designer will dwindle and also hamper his interest in the work. Until recently, neither the Copyright Act nor the Patent Act adequately protected the original designs of the ICs. In late 1983, the Japan-US working group on high technology industries observed that "both governments should recognise that some form of protection to semiconductor producers for their intellectual property is desirable to provide the necessary incentives for them to develop new semiconductor products" The group concluded that "both governments should take their own appropriate steps to discourage the unfair copying of semiconductor products and the manufacturing and distribution of unfairly copied products" The United States enacted a new Semiconductor Chip Protection Act in 1984. Japan, the second largest supplier of ICs in the world, has also followed suit. A few months ago. the Japanese ministry of international trade and industry prepared a bill and the Act concerning the Circuit I ayout of Semiconductor Integrated Circuits was passed in the Diet. The Act will come into effect on a date to be announced by the Japanese government. The Japanese Act, though similar to the American Act, has three differences: Protection is extended to all persons, regardless of nationality, whereas the US law is based on "reciprocity"; protection begins on the date of registration under the provisions of the Act and not from the date of first commercial exploitation; and infringement can result in criminal punishment. Under the Japanese Act, the subject of protection is called "circuit layouts" which is described as "mask work" in the US law. The exclusive rights acquired by the creators include manufacturing, transfers, leases, exhibitions and imports of any product incorporating the original layouts. The right holder can demand an injunction and compensatory damages for any infringement which is punishable by imprisonment up to three years or a fine not exceeding one million yens. These rights are not absolute Independent development, which is different from imitation, of an identical layout will be granted a similar copyright. The Act provides protection for ten years. RAJIV'S KEYBOARD A local area computer network has been designed and installed in the office of the prime minister, Mr. Rajiv Gandhi, which is being operated by the prime minister himself. As a consequence, all senior officers from the level of additional secretaries to the level of section officers are now learning to use computers programming. Dr. N. Sheshagiri. additional secretary to the department of electronics, revealed this information in his key- note address to the two-day national conference on computers for productivity and quality organised by the Pune Chapter of Computer Society of India. The proposed installation of supercomputers at New Delhi, Pune, Hyderabad and Bubhaneshwar will help eleiminate the "adhocism" rampart in various Union ministries, particularly in formulation of policies and programmes. Dr. Seshagiri said. With the supercomputer net work, it would be possible to have information in analytical forms which can be called on the video screen with the touch of a button. The sypercomputer in New Delhi, which is ready to come on line any moment, will serve the entire northern region. Supercomputers in Pune and Bhubhaneshwar would be ready by March, 1986 and the one at Hyderabad by May. 1 986, according to Dr, Seshagiri The supercomputers would be connected to 430 small computers, one in each district headquarters, through out the country via earth stations and the satellies INSAT-1B and INSAT-1C. The entire country is expected to be covered by this network by 1987. This network will provide the data base for the central and state governments. This network information collected by various agencies like the National Sample Survey orgaisalion Central Statistical Organisation, departments of rural development, agriculture education, health an so on, would be analysed at each hierarchical level and stored for retrieval in the supercomputer network. India was in a position to design complex civilian and mechanical parts, off-shore drilling platforms, launching pads for rockets and even shipbuilding with the aid of software procured from Norway, it was pointed out. Mr Arun Firodia. managing director of Kinetic Honda Motor Company, who inaugurated the conference stress the indispensability of computers for modern industry. Computer techniques could be used to design fuel efficient automobile engines, he added, for an example. SPENCER COMPUTERS Spencer and company, Madras, better known for their soda, having diversified into a variety of fields like pharmaceuticals, shrimp export, department stores and so on, have now entered the computer field. The company has taken up the sole selling agency of the French computer firm Honeywell Bull. The company proposes take up assembly and manufacutre of small computers with commercial applications and process controls and negotiations were in progress with three American firms for a joint venture in this arena. Simultaneously. Spencer would Venture into a project for the development and export of computer software. 12-14 elektor i by D Folger The time of the year has arrived again (doesn't it come quicker and quicker?) when most of us are frantically racking our brains trying to decide on suitable presents. If you are in that position, this little Christmas Star may solve one of your problems. It is not a complicated circuit, as its appearance is obviously of far greater im- portance than its technical ingenuity. It sometimes appears as if there are light- emitting diodes all around us, and certainly they have gradually replaced filament bulbs and neon lamps in all sorts of applications. Since these tiny components are now available in five different colours (red; amber; yellow; green; and blue), they are eminently suitable for use in ornamental lighting. When LEDs are arranged to light in flashes, rather than constantly, they can be made into very attractive, eye-catching ornaments. The circuit consists essentially of six relax- ation oscillators as shown in Fig. 1. Because the feedback resistors have a different value in each of the oscillators, the width and rate of the generated pulses will vary. Each LED lights for a time dependent upon the width of the relevant pulse, while the frequency at which it will be switched on and off is deter- mined by the relevant pulse rate. The circuit is powered by a 9 V Type PP3 battery or equivalent. The average current drawn amounts to 20 mA. Construction An example of how the Christmas Star may be constructed is shown in the photograph. It is recommended to use an IC socket and not to solder direct onto the IC pins. The component terminals may be stiffened by encasing them in araldite or simply with some suitable sleeving. An on-off switch was considered unnecessary, since it is a simple matter to connect and disconnect the supply with the battery clip. The colour of the LEDs used is left to your own imagina- tion and taste. M by A Sevriens The power stages for the stage lighting form a completely independent unit with voltage- controlled inputs. For a number of reasons, it should be contained in its own enclosure. The control cable from the control panel can then easily be up to 100 feet long, so that the power unit may be installed close to the lighting, i.e„ between the mains supply and the flood lights, spot lights, or whatever other lights that may be used. 9 “'" Block schematic To ensure first-class performance, filters are provided between the mains supply and all electronic circuits as shown in Fig.l. This obviates the possibility of any mains inter- ference reaching the electronic circuits, and also of any feedback from these circuits onto the mains supply. The mains filters are, of course, additional to the decoupling already provided in the electronic circuits. Mains unit I provides the power needed for the control of the triac stages which are shown here in simplified form. Mains unit II provides the power for the con- trol circuits and isolates the control panel from the mains. The zero crossing detector and curve shaper ensure that the triacs are fired in a manner which ensures that the brightness of the lighting varies in linear proportion to the setting of the relevant slide potentiometer at the control panel. The voltages provided by the potentiometers are applied to inputs C, . . ,C 3 . The role of the comparators will be dis- cussed later. The opto-isolators isolate the control signals from the gates of the triacs. Circuit description The mains voltage is applied to the mains transformer via a 15 A mains filter. This allows a maximum total dissipation of 3600 W, i.e., 1200 W per lamp channel. None the less, more power may be provided: this will be reverted to later. Mains unit I of Fig.l consists of Tr,. D„ D 2 , and C,. The voltage across C, is about 10 V, and the maximum permissible current is S00 mA, which is more than adequate. It is important to note that the negative terminal of <7, is connected to the mains! Zero crossing The control panel provides the power stages with direct voltages, on the basis of 2-16 elekta 12-17 function Kx) = sin 2 -y. In addition, there is no linear relation between the power applied to a lamp and its brightness. The empirical relation between the brightness and the phase gating angle is shown in Fig.4. The peculiar shape of this curve is also found in Fig.3 (6). The circuit around T 5 , T 6 , and T 7 serves to delay the gating in such a way that the re- lation between the voltages at inputs C,, C 2 , and C 3 and the brightness of the associated lamps is linear. This happens as follows. At each pulse generated by MMV 2 , T 5 con- ducts and short-circuits C 19 . The potential at the junction of P 2 and C| 9 is then + 15 V. At precisely the moment the mains passes through zero, the pulse from MMV 2 ceases. T 5 switches off immediately and C, 9 begins to charge via P 2 and 7? 28 : the potential 3 4 across P 2 drops. As soon as the voltage across C l9 reaches the 0.6 V threshold of the base-emitter junction of T 6 , the capaci- tor is charged more rapidly by the current through T 6 , T 7 , and H 2e . The potential across C, 9 continues to rise, while that across P 2 drops. When the voltage across P 2 has fallen to 1 V, the current through T 7 has become very small, and the voltage drop begins to slow down. The entire process, with component values as shown, lasts about 10 ms. After that period, a new pulse from MMV 2 starts a fresh discharge-charge cycle. The resulting "bent” sawtooth signal is buffered by IC 10 , and then used by com- parators IC, . . . IC 6 as a reference voltage. Phase gating IC 4 . . . IC 6 compare the input voltages at C,, C 2 , and C 3 with that provided by IC 10 — Fig.3 (6). If the latter is larger than the input voltage, the outputs of the comparators are logic 1 and the LEDs in the opto-isolators remain out. If the situation at the com- parators were reversed, their outputs would be 0, and the LEDs would light, Fig,3 (7) shows the output level of the comparators when the potential at the C-inputs is 5 V. This corresponds to the slide poten- tiometers set at exactly the centre of their travel, and results — in this case — to a duty factor of 50 per cent. These logic levels are inverted by the transistors in the opto- isolators and then fed to transistors T„ T 2 , and T 3 . The resulting pulses at. the gates of the triacs are shown in Fig.3 (8). The curve in Fig.3 (9) shows that phase gating takes place during alternate half cycles of the mains voltage. Construction The power stages have been divided over two printed-circuit boards: Fig.5 shows that for mains unit II. the zero crossing detector, and the curve shaper, incl. IC 10 , while that in Fig.6 contains the comparators, opto- isolators, and trigger circuits for the triacs. The triacs themselves are fitted on small boards that are cut off the board in Fig,6 along the dotted lines. Connecting wires should be stranded with a diameter of not less than 1.5 mm. It is essential that the power sockets and mains switch are rated at 240 V, 15 A, in the 3600 W (3-channel) ver- sion — see Fig.8. It is, of course, possible to use other con- figurations than that in Fig.8. For instance, Fig.9 shows a rather larger unit with nine channels (three per mains phase) each of 1200 W rating, for operation from a three- phase mains supply. This version must, of course, be fitted with a suitably rated, three- way mains switch. Another configuration is to build the version of Fig.8 twice and connect the two in paral- lel to the normal single-phase mains supply. Provided they are cooled adequately, Type TIC226 triacs may be used, and these should be protected by 6 A fuses. Other- wise it is advisable to use Type TIC263 devices: these have the advantage of being 0 - 11-0 J oHHo Z (TIC226) , (TIC263C 2x6 v,o.: 2x15 V;0 8 able to withstand the surge when a 1000 W will be difficult to obtain: it is best to make or 1200 W lamp burns out. These types them yourself. They should be wound like a should definitely be used when the con- coil in a loudspeaker cross-over network figuration of Fig.9 is used with only two with enamelled copper wire of not less than channels per mains phase i.e.. six channels 2 mm 2 cross-sectional area, with a power handling of 1800 W per channel. If three boards as shown in Fig.6 are used. and each is fitted with only one comparator. Setting up one opto-isolator, and one triac. a three- Before any setting up is attempted, carefully phase installation with three channels — check all wiring, because on the one hand, each rated at 3300 W — is obtained. The m ains voltages are present at several places TIC263 triacs must then be protected by in , he circuit and on the other some of the 15 A fuses. Suitable chokes for this version components are not cheap. Once you are sure everything is in order, connect the mains to the power stages and check with a multimeter (not digital!) set to the 300 V AC range whether there is AC present between the three C-inputs and the — earthed! — enclosure. If the me-er indicates 240 V, there is a fault in the wiring or Tr 2 is defect; if it shows between 0 and 30 V, everything is all right. Such a small voltage is normally caused by random quiescent currents, which require no further attention. Connect the outputs of the control panel to the inputs of the power stages by a suitable cable. Load each channel with a flood light, spot light, or whatever other lamp may be required. Adjust the controls for each channel in such a way that the relevant lamp just lights. Next, adjust P, in the power unit so that the lamp(s) flickers in a regular rhythm. Then adjust P, so that the lamp(s) just cease to flicker: this is the correct setting for P,. Finally, adjust the control panel for maxi- mum brightness of the lamp(s); adjust P 2 in the power unit initially so that the lanip(s) -begins to dim, and then turn this preset in the opposite direction until the brightness no longer increases. K Artificial intelligence: a myth or a moneymaker? by Professor W B Heginbolham. D.Sc. The term artificial intelligence (All means different things to different people. Its interpretation largely depends on the background interests of the individual, but if one takes an unbiased viewpoint categories can still be established. The first activity area can be identified by the notion that in some way com- puters can be made to duplicate the human intellect. Their use for evaluating highly complex theories in physics, chemistry, and cosmology is in no way unusual these days; perhaps computer based studies to evaluate theories of neurophysiology and psychological phenomena will eventu- ally lead to a better understanding of the human central nervous system. It is interesting to speculate what pro- gress the alchemists of old would have made with the assistance of modern computing methods. After all, the practical benefit of alchemy was a spin-off from its main theme with the inadvertent establishment of the basis of modern chemical science. A similar by-product could be derived from any attempt to create synthetic humanoid responses. The main restriction in the develop- ment of practical Al until now has lain in the vast amounts of computing required, with the consequent high costs and slow responses. This has led to a disappointing level of ability to solve real everyday problems, particu- larly those of industrial production, both on the shop floor and in functions such as production control. With the development of very large scale inte- grated (VLSI) systems, however, these restrictions will be significantly reduced in future. Designs Outstrip Reality For instance, the number of intercon- nected paths that can theoretically be concentrated on one chip can only be visualized by making an analogy with the entire street systems of the two American states of California and Nevada. Chip design techniques have outstripped practical capacities to actually generate this number of inter- connections. Maybe Al will beget more Al as the method is set to work to solve its own physical problems. Future applications will be profoundly influenced by the development of expert systems, sometimes referred to as knowledge based systems. The idea behind these is to enshrine the experience of a human operator as a set of rules so there is no need to work everything out from scratch. Theoretically, this reduces the problem to more manageable dimensions, but as yet such methods are in their in- fancy and can handle only around 500 to 1000 facts. There are less than ten such systems currently working in practice worldwide and the majority are concerned with medical diag- nostics. To apply Al successfully and economi cally, it is necessary to identify areas of 2 activity that fit the technology available. The use of artificial vision remains the most active area of appli- Bin Picking Problem The United Kingdom is one country that has never lagged behind in research into the concept of artificial vision and the amount of effort being put into this activity by universities, research organisations — both govern- ment sponsored and independent - and private industry is quite signifi- cant. This is illustrated by the following account. In-process handling and machine feeding still cause difficulties, particularly for variable batch pro- duction and semi-ordered production. There have been numerous attempts by researchers worldwide to solve the bin picking problem so as to enable machines to select individual compo- nents from parts stored loosely in bins or containers, so imitating human However, the fully generalized three dimensional bin picking problem has far too many degrees of freedom for economic solutions to be applied at this time. Therefore, stack picking, where the parts are not completely randomly positioned but are in a par- tially ordered state, is a much more practicable situation. British Robotic Systems Ltd (BRSL) has implemented practical stack pick- ing for loading /unloading a flexible manufacturing system (FMS) turning cell. Initial investigation to prove the system was carried out with a BRSL Autoview Viking system. The vision o X Gripper for handli processing algorithm consists of several stages — preprocessing, feature location, back end processing, adding intelligence, and final checking. A Matter of Processing It has been found sufficient to use a 128x128 pixel matrix for viewing the components. The preprocessing stage takes in a set of grey level images and reduces them to a binary image suitable for next stage processing. A high-pass filter is applied to the original image of the parts to improve feature extraction. By this means, much of the noise is removed from the image scene, concentrating analytical effort on to essential information pro- Feature extraction uses simple diameter measurement, which, when found to be within known limits, ident- ifies the component. The back end processing determines the centres of the parts within a cluster for a single layer and records successes as hits. Positional determination accurate to ±17 mm for a 1 m square bin stack can be achieved. By the use of an extra module, estimates of the centre of the area of parts can be improved to ±4 mm if required. This latter figure is sufficient to enable a robot mounted gripper to retrieve a component. Another example of the application of Al can be found in the vision guided cutting machine developed by Westland Helicopters' advanced manufacturing development depart- ment. Sheet metal components were the subject of this development, whereby computer numerical control (CNC) routing machines cut out flat profiles, but left them as nests within the parent sheet until a later stage. This facilitated easy in-process hand- ling to the degreasing and deburring operations. Removing Tags Parts were finally separated by a detagging operation. Artificial intelli- gence, in the form of machine vision, was used to control an automated detagging machine so that accurate separation of components from the flimsy and sometimes distorted parent sheet was achieved. The vision system used a DEC11/23 computer and was interfaced to a Siemens CNC machine controller. Vision transducing was by 384 x 576 picture element cameras, supplied by BRSL and arranged to view the tag area through an Olympus Borescope. Illumination of the viewed areas was by a 10 mm fibre optic light guidance system fed from a 150 W bulb. By angling the light source appropriately, well defined sharp edges were created so as to aid reliability of image transducing. Tags, approximately 3 mm wide, varied according to the size of the cutting tool used. It was necessary to remove tags without running below the edge of the general line of the component and not to leave a protrusion of more than 0.1778 mm. Tag cutting was car- ried out by a slitting saw system. Other practical applications of artificial intelligence are found in process con- trol activities, and this theme will be continuously expanded as and when appropriate processes are capable of being understood and represented accurately by computation. High Intensity Energy Beam A good example of an appropriate pro- cess is plasma welding, and the follow- ing is a typical example of a British university industrial project. This is aimed at investigating the possibilities for Al controlled plasma welding with integral arc guidance and is being car- ried out at Coventry (Lanchester) Polytechnic. Plasma is a high temperature region of ionized gas stimulated by an electric arc, which can be focused to form a high intensity energy beam. Because of this feature, the process is much less sensitive to variations in arc length than the gas tungsten arc system. The experimental system was based on a Unimation PUMA robot, connec- ted to a British Oxygen Company Sabre arc micro plasma unit with a maximum current output of 15 A. By the application of a magnetic field, an electric arc can be deflected. The direction or arc movement is depen- The electromagnetic arc deflector fitted to a plasma welding torch. dent on the polarity of the magnetic field and the arc polarity. This provides a system of weld control that has virtu- ally no lateral inertia. Essential pro- cesses normally carried out by human beings, such as weaving, can therefore be carried out very effectively by a pro- grammed variation in magnetic field intensity and polarity. Superimposed sinusoidal arc weaving at 2 Hz can improve weld conditions quite signifi- ' Great Potential This concept is laying the foundation for a highly interactive welding sytem capable of being controlled by Al to produce high quality welds automati- cally. The relative ease of welding arc directional control will also assist in automatic seam tracking to follow weld lines that are not well deter- mined. This feature will improve the prospect of achieving small batch weld auto- mation by reducing dependence on expensive welding jigs. The system also has great potential because it can be applied successfully to the welding of materials like aluminium, which cause difficulties even for human manipulative intelligence. The handling of non-rigid sheets of compliant material poses a problem for automated handling. Assembly of aerospace structures from composites requiring lay-up of carbon fibre profiles is a major problem. The University of Hull has developed a sensory gripper that provides visual feedback from two gripper mounted linear array cameras, permitting accurate alignment and lay-up of car- bon fibre composites. It has always been the hope that artificial intelli- gence would provide the key to auto- mation of the production of structural composites. While it is possible to rap- idly program a robot arm with a move- ment sequence to deal with small batches of structural composites, the actual act of laying-up poses severe mechanical problems in terms of how one transports and positions compliant materials. Visual Sensing The gripper especially designed for this has six suction cups on the under- side face. These are connected through rubber tubing to a vacuum pump to provide the means of sup- porting pre-cut profiles. To cater for the case when all vacuum cups are not in contact with the material, this grip- per can operate five cups open to the atmosphere while the single remaining one holds on to a flat surface. Visual sensing is provided to monitor profile position and ensure a quan- titative check on the final jointing of lay-up sheets. This is achieved by two 256 element charge couple device 12-23 (CCD) linear array cameras, mounted on opposite ends of the gripper. The available resolution of this system is ten pixels per millimetre, considered adequate for the application under consideration. A single line of pixels provides all the necessary information to determine edge position of the profile. A mirror is used in conjunction with each camera to allow two slits, each 25 mm wide, to be viewed by high power light emitting diodes. The gripper is used mounted on the arm of a PUMA robot and ad- dresses pre-cut composite profiles that are pre-stacked but have some pos- itional variation. This is a prototype development but illustrates the problems of applying Al to deal with problems of variability at the workplace. Knowledge Based System The University of Edinburgh is particularly active in developing robot programming languages, and notably the robot assembly programming tech- nique (RAPT) system. This is knowl- edge based and an illustration of the power of combining knowledge bases with artificial intelligence. The RAPT modelling system is an object level robot programming arrangement in which the parts to be handled and the robot work station are described in terms of their surface features. These can be plain or spherical faces, cylindrical shafts or holes. Straight edges and vertices are represented as very small diameter cylindrical and spherical features respectively. An assembly program is defined in terms of a sequence of distinct requirements defined by the program- mer to meet a progression of workplace needs. Therefore, the RAPT system provides for planned assembly relationships for components and monitors the difference between planned and actual positions by apply- ing machine vision. This is done by: ■ Specifying which features of a par ticular object are to be sought, with which camera. This is the look command. ■ Describing in broad terms the maxi- mum uncertainty expected in the position of an object being viewed. This is the tolerance command. ■ Specifying limitations on uncertain- ties in terms of spatial relationships. This is the inviolate command. Languages Enhanced By using this combination of RAPT and visual verification, it has been shown how vision can be used to enhance object level robot program- ming languages. By the addition of such a knowledge based solid modell- ing system, a powerful programming aid is created because no assumption about positions of cameras or about the manner of presentation of objects to the vision systems is required. This information is naturally included in the object level program as it is written. (LPS) Capacitors and resistors: some recent developments Despite the ever growing influence of integrated circuit technology, demand for the ubiquitous resistor and capaci- tor shows no signs of fading. In the United Kingdom, for example, the capacitor market last year was worth in the region of E130 million, and is expected to grow substantially in 1985. A little more than 10 per cent of this was accounted for by the tantalum capacitor, which is likely to be in increasing demand in the next few years especially in chip form for sur- face mounting assemblies. Tantalum is a material with many remarkable properties as far as capaci- tor manufacture is concerned. Unfor- tunately, it is also very expensive. An interesting statistic that illustrates the capacitor's importance is that more than half the tantalum mined is used in the manufacture of such devices. Production of tantalum capacitors in Britain, and recently in much of Europe, is dominated by STC. This has come, about as a result of STC's acquisition in 1983 of the tantalum capacitor operations of SEL at Bislohe, in West Germany, and of Union Car- bide at Aycliffe. With these two plants complementing its existing tantalum facilities at Paignton, Devon, STC hopes to capture a substantial slice of the world market for tantalum capaci- tors, which in 1984 stood at about £500 million, with almost half in the United States of America. Expanded Production STC also believes it is now in a much better position to build up its present share of the £45 million world market for surface mounted chip devices, and to expand its production of dipped, metal case, and moulded tantalums. Representative of the wide range of tantalum capacitors from STC are the TAP and TAG series. These are resin dipped, radial lead types approved to IECQ 300-201 GB0001 and BS CECC 30-201-027. The capacitance range extends from 0.1 to 680 pF at operating voltages of 3 to 50 V. The TAA series features hermetically sealed, axial lead types with additional approval to MIL-C-39003. Their capacitance range runs from 0.1 to 330 pF at operating voltages of 6.3 to 63 V. Then there is the TAQ series of chip-type tantalums in seven different sizes, all suitable for direct bonding and auto-insertion. These have a capacitance range of 0.47 to 100 pF at operating voltages of 6.3 to 35 V. A particularly interesting type of tanta- lum from STC is the CA series. These are etched foil capacitors in a hermeti- cally sealed silver case and approved to DEF 5134. They are available in both polar and non-polar form with axial leads, and have a capacitance range extending from 0.68 to 1500 pF. Operating voltage runs from 6 to 160 V. Coupling, Filtering, Suppression The ceramic based capacitor is one of the oldest and most reliable types, and is still widely used for such appli- cations as coupling, filtering, sup- pression of radio frequency inter- ference, and the protection of sensitive semiconductor devices. The United Kingdom market for ceramic capaci- tors was worth more than £30 million in 1984, and a substantial proportion was supplied by STC from its two manufacturing plants in eastern England. Its Norwich plant is a totally integrated facility devoted entirely to the manufacture of multi-layer ceramic capacitors. It undertakes the entire operation, from the development of raw materials to the shipment of fin- ished products. An interesting example of ceramic capacitor technology is the FLT radio frequency interference filter manufac- tured by Oxley Developments. The filter is in effect a n-section assembly of two ceramic tubular capacitors interconnected by a special ferrite material that ensures good insertion loss figures at load currents of up to 10 A. The filters are available in two case styles and with two capacitance values. One case has an M5 fixing thread, while the other has a 12UNEF thread. Minimum capacitance values are either 1500 or 5000 pF, giving minimum insertion loss figures of 45 and 50 dB respectively over the fre- quency range 200 MHz to 1GHz. The working voltage is 350 V dc for both capacitor types over an operating tem- perature range of — 55°C to 85°C. Oxley Developments also manufac- tures an impressive range of ceramic discoidal lead-through and tubular chassis mounting, high voltage, filter capacitors, ranging in value from 22 to 10 000 pF at working voltages of 50 to 500 V dc. All are manufactured in Britain according to 8S9000 and DEF- STAN 05-21 requirements. The manufacture of mica capacitors began in Britain more than 60 years ago. Although the original manufac- turing technique of using alternate layers of mica and tinned copper foil has long since been replaced by the silvered mica method, the mica capaci- tor is still in great demand for appli- cations demanding exceptional re- liability. Low Power Factor Apart from their outstanding reliability, mica capacitors are physically rugged and exhibit excellent performance at high frequencies and in pulse appli- cations. They also have a very low power factor and close tolerance of capacitance. Typical applications include oscillators, logic and trans- mission circuits, and pulse forming networks. A particularly valuable prop- erty of the mica capacitor, which has been of importance only since the introduction of nuclear power, space flight and guided missiles, is its relative insensitivity to radiation originating from nuclear reactors and outer space. STC manufactures an impressive range of silvered mica capacitors, as does MPE-Dubilier. The STC capacitors range in value from 4 to 100 000 pF, at voltage ratings of up to 400 V and in various case styles. MPE-Dubilier makes a specialized range of capacitors suitable for smoothing 12 kV, 70 kHz power supplies deliver- ing 22 A continuously, and for use in pulse forming networks operated at up to 21 kV. The United Kingdom resistor market for 1984 was worth about €65 million and is not expected to increase very much this year. This is because manufacturers are now able to design out discrete resistors, which means fewer are required for new products. A good example of this is the domestic colour television receiver. In 1981 it would have contained more than 200 discrete resistors, while today the same type of receiver has fewer than 150. On the other hand, the impressive improvements made in resistor manufacturing technology during the past decade or so mean that very high quality resistors are available to designers at prices that make them competitive with traditionally cheaper alternatives with much inferior per- formance. This applies particularly to the metal film types, which are largely replacing the moulded carbon and car- bon film resistors and are being used increasingly for surface mounting application. This says much for the advances made in manufacturing technology when it is remembered that the metal film resistor has been around for at least 30 years. Competitively Priced Conductive plastics have been available for more than 20 years, but it is only in the past five years or so that the technology has developed to the stage where such materials can be used for the manufacture of com- petitively priced potentiometers. The basic constituents of a conductive plastics potentiometer track are car- bon and resins, together with modify- ing agents that ease the processing and improve the performance. Con- ductive plastics tracks enable manufacturers to produce inexpensive, long life potentiometers featuring high accuracy, low wear, low operating tor- que, and high resolution. Another advantage is that tracks can be pro- duced to virtually any shape of resist- Representative of the latest technology in this field is the 11SB potentiometer manufactured by Fer- ranti. In fact, this particular poten- tiometer is available both as a conductive plastics and a wire wound component, and if necessary multi- gang versions can be constructed using a mixture of the two types. The 11SB is used extensively in appli- cations where the height from the mounting surface is critical. (LPSI 12-25 1§| audio JgJ sweep generator This sweep generator is intended primarily for use with the Function Generator described in the January 1985 issue of Elektor Electronics. but can also be used with other types. It varies the output frequency over a predetermined range of values so that the two units together can be used to investigate the behaviour and frequency response of, for instance, an electronic filter or amplifier. by P Theunissen abled and replaced by an external one. This external sawtooth oscilltor also drives the VCO in the function generator. When the sawtooth signal is zero, the VCO frequency is low, and the electron beam is at the left of the screen. When the level of the sawtooth signal rises, the VCO frequency increases, and the electron beam is deflected to the right. In this way. the (horizontal) x-axis is produced on the screen. The vertically varying quantity is displayed by the vertical time base (y- axis). If, for instance, the output of the function gener- ator is applied to the input of a filter, and the output of the filter to the y-input of the oscilloscope, the screen will display the fre- quency vs voltage response of the filter. The period of the sawtooth generator in Fig.l may be varied between 100 ms and When the sweep generator is used with a function generator other than that described in the January 1985 issue of Elektor Electronics, it should be noted that the VCO < voltage-controlled oscillator > must be capable of operation over an input voltage range of 0.1 V to 10 V, corresponding to a frequency ratio of 1:100. If not. a suitable level adapter should be used. Block schematic The generator provides the signals that are necessary to display, for instance, the fre- quency response of a filter on an oscillo- scope. The fundamental requirement of a sweep generator is a sawtooth or ramp oscil- lator. Since the oscilloscope during wob- bulating operates in the x-y mode, its internal horizontal time base must be dis- 12-26 ele 10 s, so that even for low frequencies a suffi- ciently long-duration sawtooth is available. Before the sawtooth signal is applied to the VCO input of the function generator, its stop and start frequencies are preset by P 3 and P, respectively. Changing the zero level of the sawtooth with P 4 affects the minimum VCO voltage and, therefore, the minimum frequency. Preset P 3 alters the peak value of the sawtooth, which determines the maxi- mum VCO level and, consequently, the upper frequency limit. The frequency limits are easily preset with the aid of a frequency meter. An erroneous setting is indicated by an LED. With S, in position a, P 3 is short-circuited, and the sawtooth signal is switched off. A direct voltage, Uy co corresponding to the lowest required frequency, i.e., start fre- quency, is then set with P 4 . The start fre- quency may be read from a frequency meter connected to the SYNC OUTPUT socket on the function generator. The upper frequency limit is set in a similar way, but with the aid of P 3 and with S, in position b. The CONTROL ERROR LED lights when the level at the VCO input is higher than the maximum allowed 10 V. Moreover, a facility is provided to cause the sawtooth to provide a logarithmic instead of a linear horizontal time base. The difference between the two becomes clear from Photographs 1, 2, and 3. Switch S 2 enables switching between the two modes. It is important to note that the setting of the fre- quency limits is valid for one of these modes only. Therefore, in practice, the time base mode should be selected before the fre- quency limits are set. Preset P 6 serves to set the MARKER FRE- Quency. This is required because, although frequencies between the limits can easily be read on the oscilloscope screen when a linear time base is used, this is not so simple (£> 12-27 Fig. 3 Dc jble-sided Resistors: R i ; R4; R14; R is; R20; R22.' R29;R32iR33= 1 k R 2 = 220 k R 3 = 3k3 R5 = 470 Q R 6 = 5k6 R7;Ri3.'Rt7.R3o= 1 M Rie=l5k Rt9 = 4k7 R 2 t = 1k8 R 23 ;R 2 4 = 270 Q R 2 5iR3i = 2k2 Rm - 1k5 R34 220 Q Pi Ik preset P 2 = 100 k stereo logar potentiometer P3; R4 - 10 k logarithm! potentiometer P5iPto 50 k preset P6 = 10 k multi-turn Semiconductors: Dt = 1N4148 D 2 ;D 3 LED; 5 n D4 - LED; 5 mm; Tl = BF256A T 2 - BC557B (optional - see 1 1C 1 7555 IC2.IC3.IC4 = TL 08 IC 5 - 3046 IC6= LM317T IC7 = 7905 ly DIN sockets fo mounting ocket for panel iting (Z-output: enclosure205x 140 x 75 it e g. Vero 075-0141 ID front panel foil 85103F - PCB 85103" with a logarithmic time base. In the latter case, it is necessary to produce individual scale divisions with the aid of P 6 . This is done by setting P 6 in a manner whereby the direct voltage at its wiper is equal to the VCO voltage. A pulse is then generated which holds the ramp for a short time: this produces a bright vertical line on the screen of the oscilloscope as shown in Photographs 1, 2, and 3. This line indicates the frequency that corresponds with the DC level at the wiper of P 6 . With S : in position d, the marker frequency may be read from the fre- quency meter. The marker frequency is, of course, also produced with a linear time Note that no sweeping takes place with S, in positions a, b, and d! Circuit description The sawtooth/ramp oscillator consists of A,, A 2 , A 4. T 2, and ic t — see Fig-2. Opamp A, and T 2 form a voltage-controlled current source that charges capacitor C 2 with a cur- rent of 0.4S. . .45 pA, depending on the set- ting of potentiometer P 2b . Timer IC, has been connected in such a way that when the voltage between pins 2 and 6 has reached a ' level of 5 V, C z discharges. When the potential across pins 2 and 6 has fallen to 0 V, a new charge-discharge cycle com- mences. The sawtooth signal is applied to frequency limit setting potentiometers P 3 and P 4 via buffer A 2 . The voltages at the wipers of the presets are combined in A 3 ; this stage also ensures that under normal conditions the VCO is driven correctly. The minimum drive level for the VCO is set by P 5 to about 100 mV for a linear time base. The signal for the VCO input of the function generator is taken from the wiper of-S lb . With this switch in position d, the direct voltage set with P 6 (which determines the marker frequency) is applied to the VCO. Differential amplifier A 9 compares this voltage with the sawtooth signal and, if these levels are the same, switches the out- put to —15 V. The leading edge of this pulse is shaped by CyltyP^ and then used to switch on field-effect transistor T,. This causes the ramp to be sustained for as long as the pulse lasts, lb ensure that the pulse duration corresponds to the charging period of C 2 , it is preset by in direct pro- portion with the charging period. The CONTROL ERROR LED lights when comparator A l0 is unbalanced, i.e., when the VCO drive is too high (>10 V). The 10 V level is preset with P 7 . When this level is exceeded, the comparator toggles which causes pulse stretcher A u to switch on D 3 . At low sweep frequencies this LED flickers. Because the oscilloscope operates in the x-y mode, the flyback is faintly visible. If the oscilloscope has a Z-input, this small de- ficiency is easily eliminated by connecting the Z-output of the sweep generator to this input. It may be necessary to invert the out- put or adapt its level as appropriate. The linear-to-logarithmic converter circuit consists of A 6 , A 7 , T 3 , and T 4 . Although this circuit is in principle temperature compen- sated, the compensation is not sufficient for the present purposes. To remedy this, additional temperature compensation is provided by A e , T 5 , T 6 , and T 7 . Transistors T 3 ...T 7 are contained in IC 5 , a transistor array Type 3046. Transistor T s functions as a temperature sensor with a sensitivity of —2 mV/°C. The difference between its base-emitter voltage (about 0.6 V) and the drop across 7? 20 is amplified by A s . The amplified voltage drives current sources T 6 and T 7 . When U BE is greater than i/ R20 , a large current flows through T 6 and T 7 , which heats the chip. When the temperature of the chip reaches 1 2-28 ele the value preset by P 9 , the current through T e and T 7 decreases. In this way, a balanced condition is reached in which the dissi- pation in T 6 and T 7 is of a level that keeps the temperature of the chip within narrow When IC 6 has attained its correct operating temperature, the LOG SWEEP NOT READY LED goes out. The voltage at the output of A 8 is then between —5 V and 0 V. Construction Before work is started on the printed-circuit board of Fig.3, the following preparations should be made in the function generator. The ± IS V, earth, and VCO lines should be taken to the sweep generator. The + IS V line is tapped at the cathode of D 7 ; the —IS V from the anode of D 8 ; earth from the central pin of IC 4 ; and VCO from pin c of the VCO input socket: the connection between pins a and c of this socket must be cut. The socket can then still be used as input for an external drive voltage, because if a plug is inserted, the drive from the sweep generator is automatically switched off. The four lines are best terminated in a 5-pin DIN socket that may be fitted in the sides of the board: C12; Cg; ICs(2x); P,; P 5 ; R 3 ,; R32; R25; Rs; R27; R11; R12; R22; P10; C5; R15; R23; C7; Ti; Cs (2x1: C 4 ; p 6 : P2; Ris: P 4 ; solder pin +15 V. A short length of bare positions: pin 4 of IC2; pin 1 of IC3; pin 2 of IC3: pin 4 of IC3; pin 4 of IC 4 ; pin 8 of IC5. X-AXtS terminal, S2a. S2b- and S2M must be fined at the track side of the board. rear panel of the function generator. It is also necessary to drill some extra ventilation holes near the IS V voltage regulator in lid and bottom panel of the function generator to ensure sufficient cooling with the in- creased dissipation. The sweep generator may, like the function generator, be housed in a 205 x 140 x 75 mm verobox. Fit a 5-pin DIN socket, similar to that in the function generator, to the rear panel. Where appropriate, also fit a BNC socket for the Z- or Z-output. Finally, remove the two fixing nuts from the lid and bottom panel (these are located roughly where P a and P 10 will be situated). The comers of the printed-circuit board should be rounded with a file to make them fit snugly into the guides provided in the enclosure. The front panel should be drilled in accord- ance with Fig.4. The holes should be prop- erly deburred to ensure that the self-ad- hesive front panel foil fits smoothly onto the panel. Next, glue the three LEDs into place; fit the SWEEP switch, and then the OUTPUT socket. Where a mains on/off switch and a second output socket are required, appropriate holes should, of course, be drilled at the same time as the others, otherwise the foil may be damaged! Only when all this preparatory work has been done should the printed-circuit board be started, but do not yet fit £7, or link A-B. Calibration Once all the connections between the sweep generator and function generator are made, the temperature-compensating cir- cuit must be set. Connect a digital voltmeter between pins 12 and 13 of A 8 and adjust P 9 for a reading of about 60 mV; note that pin 12 is more positive than pin 13. Switch off the mains, and solder wire link A-B on the PCB in place. Switch on the mains again, when after a short time the LOG SWEEP NOT READY LED should go out. Turn P 2 . anticlockwise. Next, connect the X-AXIS output of the sweep generator to the Y- INPUT of an oscilloscope as shown in Fig.6. Set the oscilloscope to DC and 50 mV/div. Set the frequency range on the function generator to 1 kHz. Connect a frequency meter to the SYNC output of the function generator. On the sweep generator, set S, to position a and S (SWEEP) to LINear. Turn first P 4 (START) and then P s fully anticlock- wise. Finally, turn P 4 clockwise until the fre- quency meter reads 1 kHz. Set Sj (SWEEP) to LOGarithmic and P 4 (START) fully anticlockwise. Switch off the sweep generator, and turn P 8 fully anticlockwise. Switch on the sweep gener- ator and turn P 8 slowly clockwise until the ! 2-31 frequency meter reads exactly 1 kHz. Set S, to position b and turn P 4 (START) fully anticlockwise and P 3 fully clockwise. Then adjust P ]0 to obtain a drive voltage, U V co' °f about 11 V. Finally, turn P 3 (STOP) fully anticlockwise, and set S2 (SWEEP) to LIN. and S, to pos- ition b. Turn P 3 (STOP) to obtain a reading on the frequency meter of about 102 kHz, and then adjust P 7 so that the CONTROL ERROR LED just does not light. If the fre- quency is increased slightly, the LED should light. This completes the calibration. Capacitor C, can now be soldered in place. Finally A typical set-up for measuring a frequency response is shown in Fig.S. If primarily a logarithmic time base is required, linear types of potentiometer may be used in the P 3 and P„ positions: this makes setting the frequency somewhat simpler. With the sweep generator connec- ted to it. the function generator should not be operated in the 10 kHz range, because the VCO voltage would tend to drive it up to 1 MHz, whereas its maximum frequency is only 100 kHz. The photographs give some idea of what the sweep generator can do. A clearly defined frequency response as, for instance, plotted on paper can only be obtained on the screen if the voltage at the Y-input of the oscilloscope is rectified and converted to its logarithmic value. In view of the extra expense, these facilities have, however, not been provided in the present cost-effective sweep generator. N DOGDOm? A.M. Bosschaert In this age of digital things (?) it is nice to dress up the front panel of home-brew equipment with digital readouts. This handsome little digital voltmeter is easy to build and is inexpensive. In the form presented here, it is a single range unit, intended for use with power supplies •Since this DVM was designed to replace conventional moving coil panel meters, on such things as power supplies, the price was one of the main design con- siderations. The unit uses only 6 tran- sistors and three C-MOS integrated cir- cuits: one 1C contains four Schmitt triggers, the other ICs each house a decade counter with built-in BCD-to- seven segment decoder/drivers. It fea- tures 2'h digit readout and has an acceptable accuracy. Basic operation Conversion of the positive input voltage into a quantity that can be digitally displayed is accomplished by converting the input voltage to a current. This cur- rent is then used to control a variable frequency oscillator; the output fre- quency of this oscillator is a linear function of the input voltage. This frequency is counted by the counter unit and displayed. The counter unit is controlled by a free running oscillator that determines the gate time for the counter stages and resets them just before the start of each count. It also blanks (turns off) the readout during the brief count cycle. Circuit operation At first glance, the input circuit may appear to be confusing, but if one breaks it into smaller units it is much easier to understand. The heart of the DVM is a current- dependent oscillator. This oscillator consists of the following parts: gate Nl, D2, Cl in the charge path and Cl, T2 and R2 in the discharge path. The input voltage is converted into a discharge cur- rent by T1 and T2. These transistors keep the voltage across R2 equal to the input voltage at all times. Since R2 is I k, the current through it (the dis- charge current!) in milliamps is equal to the input voltage in volts. The dis- charge time of Cl is therefore a linear function of the input voltage. The time required to charge Cl is always the same. The charge current is supplied from the low impedance out- put of N 1 . If for a moment we assume that pin 2 of Nl is held high, the oscillator circuit can be more easily understood. The secret that allows Nl to operate as an oscillator is the fact that its switching levels are not the same (hysteresis). If Cl is completely discharged pin 1 is low, making pin 3 high. In this state Cl will be charged rapidly. When the voltage on Cl reaches the upper switching threshold of Nl, pin 3 goes low. D2 prevents Cl from discharging into the output of Nl, and since the input (pin 1 ) is a very high impedance, the only discharge path for Cl is through Continued on page No. 12-41 12-33 fantasia on a MIDI theme After God had created man, He istic of an analogue synthesizer: both are rested. After man had created the interface norms. This would, however, be a (analogue) synthesizer, he also rested, limited one. since MIDI offers many more by D Meyer But only for a little while, because within a very short time digital techniques had shown a different way, enabling sounds, rhythms, and melodies to be stored in memories. And man realized that he, unlike God, had not created perfection. The new techniques made the addition of, for instance, digital-to-analogue converters, digitizers, and other peripheral units necessary, so that within a relatively short time the originally simple synthesizer had grown into a complex array of equipment through man's attemps at idealizing it. The magic word was preset. And soon the win- dows of music shops became tangible evidence of a Japanese invasion of relatively inexpensive digital synthesizers. And everybody sensed that music had taken a new lover: the microprocessor. To make that union a happy one. man hit upon a brilliant idea that would make it possible for communications to be effected between synthesizers, or between a syn- thesizer and all kinds of modulators, rhythm boxes, mixer units, and many more. And he called it MIDI. Now, a little later, it seems that MIDI is likely to ensure the couple's continued happiness and keep them in perfect harmony. And so, the device that began with charging and discharging capacitors, oscillators, filte'rs, waveform generators, and keyboards that generate discrete voltages (1 V/octave), is now augmented by all sorts of memory and digital systems that generate and syn- thesize the most complex sounds. Soon, when digital loudspeakers have become available, there will no longer be any need for digital-to-analogue converters and digitizers, and all that will pass from one unit to another will be Is and Os. But since that stage has not quite been reached, some sort of converter is still required, and its oper- ation must be digital. It might be said that as long as a sound is not heard by the listener, it only exists in a purely digital form, ie., a complex series of digits. A careful comparison may be drawn be- tween MIDI and the 1 V/octave character- facilities than this single characteristic. It is, moreover, not so much MIDI itself that offers these facilities, but rather the fact that the music characters are in the form of numbers. And it is well-known that the fast processing systems currently available are more efficient in dealing with numbers, however complex, than with electrical (i.e., analogue) quantities. All the more reason to stress that MIDI is a standard of communi- cation between microprocessor systems that have been designed specifically for music applications. But MIDI is also an interface that does not contain even a hint of a feature specific to music — see Fig.l — in the same way that a Centronics interface, often used with printers, has no feature specific to printing. The MIDI interface has no intelligence from a logic point of view: it is nothing but a col- lection of protocols for data communi- cations and music characters. Therefore, combining MIDI with new electrophonic instruments becomes beneficial only through the richness and diversity of the associated logic programs. Whether one or a hundred MIDIs are added to a monophonic synthesizer, it will not change into a polyphonic one. It is, therefore, the microprocessors attached at either end of the MIDI chain that give the synthesizer a semblance of being intelli- MIDI and real time Ever since the 19S0s, there has been an obsession among those engaged on elec- trophonic music to eradicate the time delays between electronic actions or instructions and their musical result. Such delays, which may vary from seconds to months, are, of course, detrimental. With digital and modern data communication methods, it is possible to reduce such delays to virtually nothing. In music, time is measured in an implacable and rigorous manner, and in a microprocessor with a clock frequency of 12-34 several MHz such rigidity is, of course, easily attained. It is possible to transmit and process data and musical characters, prior to producing clear sound signals, so that they appear simultaneous to the listener. This implies that the MIDI interface, via which the data and characters required for the production of a sound are transmitted, does not act as a brake on the general pro- cessing. At present, the agreed serial trans- mission rate is 31.250 kilo-baud (=31 250 bits/second), which is quite convenient, but already too low for certain sophisticated applications. It is, however, much higher than is tolerated by the majority of RS232 interfaces with which the MIDI may be com- pared. Additionally, MIDI is typified b» perfect decoupling of the different voltages brought about by the interconnecting of a number of different units. Transmission takes place via an opto- isolator, which ensures the absence of earth loops that are often so troublesome in an array of sound reproducing equipment. The circuit of Fig.2 offers the facility of using a MIDI in conjunction with an existing RS232 interface, provided the associated micro- processor system is capable of handling the 31.250 kilo-baud rate. However, this circuit does not provide an external clock to the ACIA that effects the parallel-to-serial con- version. This ACIA may, for instance, be the 6551 on the 6502-based CPU card featured in the November 1983 issue of Elektor Elec- tronics. There, the maximum baud rate with a crystal frequency of 1843.2 kHz is 19 kilo- baud. An external clock may raise this to 125 kilo-baud. For the present purposes, a crystal frequency of 31.250 x 16=500 kHz is sufficient. Note that pin 7 of the 6551 should be left unconnected. Once the 31 250 baud rate is established on the existing RS232, all that remains to be done is to ensure that the RS232 levels are passed onto the 5 mA cur- rent loop used by MIDI. Desirable features of the average modem synthesizer • At least 32 presets to shape the timbre or the registers. • Touch-sensitive keyboard. • Single-key modulation with aftertouch. Note that this feature is rarely provided by single key, but is normally common to all keys, although this is not always specified in the manual. • Variable portamento. • Pitch bend or modulation wheel. • 49. . .88 keys. • Breath control. • Transposition. • Variety of oscillators (voltage-controlled, digital-controlled) and filters (voltage- controlled). • Envelope generators. • A number of LFOs (low-frequency oscillators) (it is beyond understanding why even the DX7 contains only one of these circuits). • Programmable operators and algorithms (it is a major deficiency that the algorithms imposed by the manufac- 1 2 Fig. 1. Schematic diagram of a typical MIDI inter- face (that of the Yamaha DX7I. Note that pin 2 of the INput socket is not connected to earth. 12-35 -II I turers do not allow the feedback of several operators superimposed on one another). • Phase, amplitude, and pulse-width modulation. • Split keyboard. • And many more. . . This list shows that if all the listed facilities and some others beside are controlled by MIDI, this interface will become very busy, indeed, and yet, there is still no information on the MIDI interface until the synthesizer in question is being programmed with the aid of an external system. This can also be seen from Table 1. It must be understood that all these data are not permanently updated but only when it is necessary to modify one or more of them. But even then, only the data relevant to the characteristic to be altered is transferred. Nothing at all happens about the rest. Moreover, the modification may affect only one of the channels of communication, as it is possible, thanks to MIDI, to serve several individual instruments over the same line, but to address each one separately with data that the others can also understand. An even more restrictive means of communi- cation allows each of the instruments to be addressed separately with specific data that the others cannot understand. These chan- nels allow a microcomputer to control, via the same MIDI interface, not only small and simple synthesizers — polyphonic or monophonic — and smaller units, such as rhythm or other special effects boxes, but also synthesizers that are more complex as far as the number of voices is concerned. It should, perhaps, be stressed again that it is the logic characteristics, and not the chan- nels, that provide the real substance of this Moreover, these channels (at the present time there are sixteen of them available) can be used in three different modes, called OMNI, POLY, and MONO. Briefly, all data available in OMNI are transmitted over all MIDI channels: it is then as if there were only one common channel. In POLY, all data are transmitted over one predetermined channel in such a way that only receivers associated with the channel can receive the data. Finally, in MONO o ily specific data are transmitted over certain channels. In other words, in the POLY mode it is possible to imagine a trumpet, a saxophone, and a trom- bone playing in unison with the piano and the players reading the score over the shoulders of the pianist. The chords of the . left hand are in the score, but the brass is not playing them. In the MONO mode, on the contrary, each instrument has its own part to play and the scores of the other instruments do not matter. It is, in our opinion, fruitless to go further into the theoretical details of the MIDI con- figuration, for only practical use will really show up its effectiveness. In the meantime, we believe that the principal virtues of MIDI have been discussed from a practical point of view, which does not in the least pre- judge the richness (or poorness) of MIDI on the strictly musical plane. . . Realization Its realization, simplicity, and economy make the MIDI truly staggering. The com- patibility it provides between products of different origin makes all the difference: Communication is made possible where it was not before, and more flexible — and thus more efficient — where it was difficult before. And, of course, it also increases the turn-over of electrophonic businesses. The more costly units, such as a dynamic touch keyboard, may be operated at the same time as several other units. Similarly, the microcomputer controlling the whole system contains mass storages such as fioppys, from which the MIDI system may profit. Shared joy is double joy! The fact that all information consists of logic bits makes it possible for it to be accessed more easily, more often, and at less expense than would be the case with buying new material. Finally, there is another advantage the digital system has over the analogue and which is very much appreciated by mu- sicians, and that is its precision. This is as good for pitch (no more problems with tuning, drift, and so on) as it is for rhythm and synchronization. H Did you know. . .? A new museum, devoted to film and tele- vision, will take its place alongside drama, sculpture, fine art, and music on London's South Bank complex. Called the Museum of the Moving Image, it will open in 1987, with displays devoted to the cinema, television, and video, plus futuristic images in fibre optics and lasers. The £7 million museum will be built on a 3000 m* plot next to the National Film Theatre. It will comprise twenty sections, plus a special exhibition, which will change every six to twelve months. It will cater for up to a million visitors a year and will be open up to twelve hours a day, seven days a The museum will be no ordinary collection of static exhibits to be viewed silently: the visitor will be invited to experience and par- ticipate. There will be an operational TV control room overlooking a miniature studio which visitors can operate by themselves, as well as areas dealing with news and documentary. Museum goers will be able to watch an animation film being made, and there will be a place where children can spend an afternoon creating their own animated movies. There will be talks and lectures from actors, directors, and techni- cians, and there will be a library of related video material. H Zinc-carbon Zinc carbon batteries are the least expens- ive, have the smallest capacity, and have a greatly varying capacity depending on the current. Their nominal e.m.f. of 1.5 V degrades linearly in use; when this has reached about 0.9 V, the battery should be thrown away. Alkaline manganese Alkaline manganese batteries are more expensive than zinc carbon types, but have a much better performance, particularly at higher currents and at lower temperatures. Mercuric oxide Mercury batteries offer a virtually constant voltage over their effective life. Moreover, they have a high energy density and good storage characteristics. Their output current is, however, limited at low temperatures. batteries: a danger to the environment? Mercury Many primary batteries contain some mer- cury, albeit usually in very small quantities. Mercury is, of course, used in other elec- trical equipment as well: switches, mercury vapour lamps, and in the cathode of certain rectifiers. Compounds of mercury are highly poisonous, although several are used in medicine. An amalgam with cadmium, also used in dental fillings, is the alloy found in primary batteries. It is interesting to note that mercury — chemical symbol Hg — is the only metal that is liquid at room tempera- tures, whence its use in barometers and thermometers. Mercury does not become inert in its natural state, is not affected by dilute acids, and only dissolves in hot oxidizing acids. The mer- cury contained in discarded batteries that each week land on refuse disposal dumps by the million remains stable and may, therefore, become an environmental hazard in years to come. Zinc Another element contained in primary bat- teries that may cause problems is zinc, but nowhere near as serious as mercury. This is because zinc bums in air, and combines with halogens and sulphur. Alternatives? Before the question as to alternatives to zinc and mercury can be answered, the characteristics of the various primary bat- teries should be considered. The table gives an overview of some of the properties of a number of primary batteries. Lithium Their very small size and extremely high energy density makes lithium manganese dioxide batteries ideally suitable for use in miniaturized electronic equipment. They are able to perform effectively over a tem- perature range of — 20 °C. . . +50 °C, and have an exceptionally low rate of self dis- charge: 85 per cent of their capacity remains available after six years of storage at + 20 °C. Lithium chromoxide batteries also offer a high energy density, low self dis- charge. and a wide temperature operating Over the past twenty years or so, the pro- duction and use of batteries have grown enormously. Whereas in the sixties their main use was in portable radios and torches, nowadays they are found in quartz watches, pocket calculators, hearing aids, portable computers, camera Bash units, and many Batteries consist either of primary, ie., non- rechargeable, or of rechargeable cells. Primary batteries may be sub-divided into zinc-carbon, alkaline manganese, mercuric oxide, silver oxide, lithium manganese diox- ide, lithium chromoxide, and zinc air types. The best known — and oldest — recharge- able battery is the lead-acid type, although nickel-cadmium batteries are catching up fast. As this article is concerned with the effect of batteries on the environment, it deals with primary batteries only, as these are disposed of by the million every day. 2-39 range, but, in addition, maintain an e.m.f. of over 3.5 V with normal loads (lithium manganese dioxide >2.7 V), and have a design (storage) life of 10 years. Lithium bat- teries also have disadvantages in that they are more expensive than alkaline types and, because of their different output voltage of 3.0 V nominal, cannot simply be used as a substitute for 1.5 V batteries in existing equipment. atmospheric oxygen in the presence of a catalyst is used to produce the elec- trochemical potential. The use of oxygen from the air means that the cell can be filled with nearly twice as much anode material, ie., zinc, as, for instance, corresponding silver or mercuric oxide cells. This gives zinc air cells a very high energy density and nearly double the life of silver and mercury cells. Silver oxide The nominal e.m.f. of silver oxide cells is 1.5 V. Storage life is about 2 years at +20 °C: at the end of one year, about 90 per cent of nominal capacity remains available. Bat- teries of this type also perform well at low temperatures. Zinc air In zinc air cells, the reaction of zinc with From these considerations, it would appear that two types of primary battery might pose an environmental hazard: the mercury — self-evidently — and the alkaline manga- nese. The latter appears not to be at first sight, and perhaps rightly as long as a single cell is considered. However, this type of bat- tery is produced in growing numbers as a composite unit with a volume of from 10 to 100 times that of a single cell; in that case, the amount of mercury becomes a matter of concern. It is clear from the foregoing that alterna- tives for both types, depending on appli- cation, are the silver oxide, the lithium, or the zinc air cell. In some instances, alkaline manganese units may be replaced by zinc carbon types. Finally If and when fewer mercury cells will be used, it is likely that more silver oxide cells will be produced to satisfy battery demands. Would it not be feasible for some enterprising concern to collect these cells when used and recycle the silver in them? overload protection for electric drills This circuit prevents the burning out of the motor of an electric drill through an over- load. The mains voltage is transformed into a pulsating direct voltage by a bridge recti- fier. The motor of the electric drill forms the load of silicon-controlled rectifier (SCR) Th,. Resistors /?,. . and diode D, keep the SCR on as long as the current, / M , through the motor does not exceed a given value. The motor current is monitored by resistor X t : its maximum value depends on the setting of P,. If an overload causes / M to exceed the maximum value, transistor T, switches on and triggers SCR Th 2 . The gate circuit of Th, is then short circuited, and the motor is switched off. Capacitor C 2 ensures that Th 2 is kept on until the circuit is reset by switch S,. Resistor R 2 and capacitor C, form a low- pass filter which ensures that Th 2 cannot be triggered in error. Preset P, allows the circuit to be used with electric drills rated between 50 W and 1 kW. 12-40 ele millivolt Continued from page No. 12-33 T2 and R2. This discharge current (and. therefore, the discharge time) is deter- mined by the input voltage, as de- scribed earlier. Once Cl has discharged to the lower switching threshold of Nl, Cl is re- charged rapidly and the cycle repeats. The higher the input voltage the faster Cl will be discharged, and the higher the output frequency will be. This fre- quency is gated into the counters by the internal time base (shown in the dashed box in the circuit diagram). The time base really has three functions: gating the incoming frequency, blanking (turning off) the display during the count cycle, and resetting the counters to zero just before the start of the The duration of the positive portion of the waveform being produced by the time base must be ‘spot on’, otherwise the unit will not be accurate. Therefore PI is provided for full scale calibration of the unit. The timebase output has a very low mark to space ratio so that the count cycle is very short compared to the readout time. During this 'enable' time the output of gate N2 is high. T3 is switched off, blanking the display (this is perceptible as a short blink). The reset pulse (the positive edge of the enable pulse) is passed via C3 to the counter section. Resistors R5, R6 set the voltage level at pin 9 of gate N3 and pin 1 3 of gate N4 between the switching thresholds of these Schmitt triggers. A positive pulse will now switch the gate output to ‘O' and a negative pulse will switch it to T. The reset pulse will turn on T4 briefly, so that the outputs of N3 and N4 w.ill both switch to T. T5 and T6 are turned off and the ‘100s‘ display (D7) is blanked. C4 is discharged. Positive pulses .. supplied by the ‘carry’ output (pin 5) of the decade counter IC2 at each transition from 9 to 0, and resulting step is differen- tiated by C5 and RIO. The first time this occurs, N3 switches to ‘O’. This turns on T5, so that a ‘1’ is displayed (segments b and c of the ‘hundreds’ display). C4 is still dis- charged, so the output of N4 is held at ‘I’. However, C4 is now charged through R8 and T5. The next zero-to-one tran- sition at the output IC2 (overflow) gives a second positive pulse. This time, the output of N4 does switch to ‘O’, causing the display to read ‘H’(for ‘HELP’). At the same time IC1 and 10^ pins number S (display enable) change to low, in- hibiting the readout by D5 atd D6. As a result the D7 character ‘H’ is the total display. Calibration Apply a known voltage between 1 and 2 Volts to the input and adjust PI for the correct readout. Since the DVM only reads to 1.99 V, an external volt- age divider must be added if higher voltages are to be measured. N inverter This inverter circuit can be used to power electric razors, stroboscopes and flash tuoes, and small fluorescent lamps from a 1 2 volt car battery. In contrast to the usual feed- back oscillator type of inverter, the oscil- lator of this inverter is separate from the output stage, which allows easy adjustment of the oscillator frequency to suit different applications. The oscillator circuit consists of a 555 timer connected as an astable multivibrator. The inclusion of D1 ensures that the duty-cycle of the squarewave output is maintained at about 50%. The output of the 555 drives the base of T1 which switches current through one half of the primary of the transformer. T2 is driven from the collector of T1 and thus switches current through the other half of the transformer winding on opposite half cycles of the drive waveform. Zener diodes D4 and D5 protect T1 and T2 from any high-voltage spikes generated by the transformer. The voltage applied to the transformer primary is stepped up and the required high output voltage appears across the secondary winding. Depending on the application the secondary voltage may or may not be rectified. Components The transformer is a standard mains trans- former with two identical secondary windings or a single, centre-tapped second- ary. This transformer is, of course, driven in reverse, i.e. the secondary becomes the primary and the output is obtained from the primary (which is now the secondary). It must be borne in mind that, since the inverter produces a squarewave output, the RMS secondary voltage and peak secondary voltage are identical. This affects the choice of transformer for different applications. The required secondary voltage of the mains where 1 2 V is the inverter supply voltage U ra is the normal mains primary volt- age of the transformer. Up is the desired peak secondary volt- age. An electric razor requires 240 V* RMS = 240 V* peak, so if a transformer with a 240 V primary is used the secondary windings should each be 1 2 V or a single 1 2-0-12 wind- ing. For vibrator type (non-rotary) razors the oscillator frequency should be 50-60 Hz, so the value of Cl should be 330 n and PI should be adjusted accordingly. Rotary razors are less critical of mains frequency. When operated from the normal mains supply, fluorescent lamps receive a peak supply voltage of around 340 V, which enables them to strike reliably. The trans- former secondary voltage should be calcu- lated with this in mind, which means that secondary voltages of eight or nine volts will be suitable. Fluorescent lamps can be operated with improved efficiency at frequencies greater than 50 Hz, and the transformer will also be more efficient. Choosing a value of 56 n for Cl the oscillator frequency may be set to around 250 Hz. At frequencies much higher than this iron losses make the trans- former less efficient. The current rating of the transformer depends upon the load. For electric razors and small fluorescent tubes up to 8 W, 500 mA secondaries will be adequate. Higher output powers may be obtained by choosing a suitable transformer, replacing T1 and T2 by higher power types and reducing the value of R3 and R4 (minimum 1 20 fi). To power strobes and flash tubes the output , must be rectified and used to charge a reservoir capacitor, which should be of a type rated for high discharge currents. The bridge rectifier should be rated to suit the peak output voltage. M * U.K. only. Overseas readers substitute the etebto* mdia december 1985 12-41 play ball with Elektor! Here is a circuit that will appeal to the inveterate gamblers among our readers. It is an electronic version of the pin-ball machine, but is not so versatile as the professional machines found in amusement arcades, bars, and other public places. But then, it is also much less expensive and does not gobble up all your money. The idea for the circuit is based on a study conducted in Las Vegas, which suggests that any money won is of less importance to most players than the fact of getting a reward for his or her skill in playing the machine. With reference to Fig.2, when the player presses S, . bistable N 5 -N 6 toggles and this switches on oscillators N,...N 4 . The resulting clock signal at pin 1 of IC, causes this circuit to commence counting: inputs A. B, C. and D of IC 2 become logic low, and this starts the ball rolling! Accompanied by flashing lights and appropriate sounds, the imaginary ball rushes about, hitting pins galore, and finally disappears. The length of time it is in play can be extended by skilful use of switches S2 and S3. The pins are represented by variously coloured LEDs, although D3„ and D 3I are on all the time. As soon as D 20 or D 23 lights, S 2 or S3 should be pressed, as your skill indicates. If you have judged correctly, the ball travels back upwards and keeps the game alive. Possible positions of the pins (LEDs) are suggested in Fig.l. Circuit description Clock oscillator N, — see Fig.2 — enables speeding up or slowing down of the ball: its frequency is controlled by N 2 . Oscillators N 3 and N. affect the to and fro travelling of the ball: they drive two NAND gates, N 7 and N 8 , which in turn provide logic levels to one of the inputs of XNOR gates N 9 and N l0 . These latter gates invert the logic levels at inputs A and B of IC 2 at irregular intervals that depend on the setting of the relevant oscillator. If only the least significant bit LSB is inverted, the ball jumps one LED higher or lower, again depending on the rel- evant logic level. If input B is inverted, the ball travels two LEDs up or down. Oscillator N 2 controls the clock signal: it operates in an erratic manner. Its prime task is to switch T, on and off and thereby short- circuit C 2 or not in the process. The effect of this is that the clock is speeded up for a short time, which causes the ball to travel faster, just as in a professional machine. The tendency of the ball is, of course, to move downwards (prior to disappearing), so that at some time D 25 will light. The Q 15 output of 1C 2 will then become active: transistor T 2 switches off, which resets bistable N 5 -N 6 ; the clock oscillator then stops, and the game is over. All this can. however, be prevented by rerouteing the ball upwards with S 2 or S 3 . Note that these switches should only be pressed at the precise moment that D 20 or D 23 lights. If this is done skilfully, IC, is reset, and the clock oscil- lator continues to operate; if not, the ball is suspended and the last lit LED remains on. The 0 9 and Q 12 outputs are not connected direct to D 20 and D 23 respectively, but via drivers T 6 and T 7 . The logic level present at the bases of these transistors is applied to NAND gates N 13 . . .N 16 . When either S 2 or S 3 is pressed judiciously. N 13 is provided with a 1 at its pin 9. Together with the 1 at its pin 8 (from 0 9 or 0 12 ), this results in a 0 at its output. This is inverted to a 1 by N, 4 , which is applied as a reset pulse to IC, via D 4 . ICi resumes counting and D u lights again. The logic 1 at the cathodes of D 9 and D 10 is also applied to N 15 , which then generates a 0 at its output. If Sg or S 3 has been pressed timely, one of the inputs of N l6 is 1. This means that T 3 remains on. Pin 1 of N 6 is then high, and the bistable remains set. If Sg or Sg was not pressed timely, which means that the logic 1 from Q 9 or Q 12 has already disappeared at the time of pressing, both inputs of N 16 are high. The base of T 3 then becomes 0, and C 8 discharges slowly. After a short delay, the transistor switches off, and this pulls pin 1 of N 6 low. The bistable is then reset and the clock oscil- lator stops. The game is over! The ball is put back into play with S,: IC, then receives a reset pulse, and the oscillators start as before. To make the game even more realistic, a small loudspeaker has been added which provides the typical sound effects of a pin- ball machine. Every time N 7 or N e provides a logic 1, the leading edge is differentiated by one of four RC networks and applied to pin 1 of N i9 via D 5 . . .D 8 . The signal triggers the chain N ;9 . . . N, 7 . The frequency of oscil- lator N 17 may be set as required by preset P 5 . The rectangular signal is amplified by T 8 . The value of J? !4 'has been chosen to give a suitable volume for most purposes; if more power is wanted, the value of the resistor 12-43 S' S'* may be reduced, but not below about 50 ohms: its rating should then be increased to W. ensure that the trailing edge of the pulse is also made audible, the output signal of N r N a is inverted by XNOR gates N„ and N, 2 and then differentiated. The remaining presets may be set to individual taste, experience, and reaction time by trial and error. The pin-ball will operate whatever their settings, because they all have a series resistance. To use the pin-ball machine competitively, a counter and associated display is required — see Fig.4. The versatile counter circuit described in our April 1985 issue, P. 4-44, is ideal for this purpose. Note that if more digits than shown in Fig.4 are required, these are simply added by connecting the power lines and the Co line of the previous section to the Ci of the added section. although a possible set-up is shown in Fig.3. Whatever enclosure is chosen, it must, of course, be large enough to house the PCB and it is advantageous to have the display section at an angle. The best position for Si is in the right-hand bottom comer of the front panel, while S 2 and S 3 are best fitted in the left- and right-hand side panels respect- ively near the front of the unit. Construction In the mechanical design of the can give your imagination fr< 5 The circuit is constructed on the printed- circuit boards shown in Fig.5. The indicator board serves also as decorative front panel (the decorations are not shown here, but are provided on PCB 85090-2 available through our readers' services). If you want to provide the unit with its own power supply, this should be taken into account when planning the enclosure. The unit requires 9 ... 15 V at about 100 mA. The power supply may, of course, be combined with that for the disday section as shown on page 47 of our April 1985 issue. Where the display of Fig.4 is used, connect its clock input to pin 4 of N 5 . M Did you know. . .? Phoenix, the British Army’s new suveillance system, contains a fully-equiped pilotless aircraft for real-time targetting and battle- field surveillance. Apart from the small air- craft with its advanced avionics and infra- red imaging equipment, the system com- prises an air-to-ground data link, a mobile communications ground station, and logis- tics vehicles for launch and recovery. The air vehicle will have low radar, infra-red, and acoustic "signatures" to make it hard to detect. Its modular construction and small size make it easy for soldiers to assemble, launch, and recover. ** 1 2-46 elektor in Computers aid survival of rare plants by Martin Redfern There are about a quarter of a million different plant species in the world, but botanists estimate that, if man continues to exploit the natural environment at the present rate, one- sixth could be extinct before the year 2050. Many could disappear without ever being properly identified and without their potential for man being realized. The International Union for the Con- servation of Nature (IUCNI and the World Wildlife Fund (WWF) have responded to the emergency with a Plants Campaign, aiming to raise US$4 million to protect the habitats and save individual plants. One of the problems is knowing exactly which plants are in danger and where they still survive. The IUCN Conservation Monitoring Centre (CMC) can provide some answers. It is based at Cambridge and publishes the Red Data Books on threatened plants and animals. The botanical section of the centre is at Kew Gardens in London, where plant enthusiasts of past centuries often brought their discoveries. The CMC now operates an electronic database which scientists at Cam- bridge and Kew can update and expand each time new information is available. Apart from governments and organizations that are members of the IUCN, more than 2000 scientists worldwide are contributing infor- mation, together with researchers on 300 IUCN or WWF field projects. Data Bank Data on more than 14000 threatened plant species have been entered in the computer store, both as key facts and as pages of descriptive text. In addition to a plant by plant analysis, there is information on the status of plants in different countries. Researchers can ask questions of the database, such as: "Which plant species are under threat in Poland?" or "Are there any species of wild potato threatened with extinction?". There is also an inventory of protected areas such as national parks. Plants that are particularly rare, or even extinct in the wild, sometimes survive in the 130 leading botanic gardens around the world. At the push of a button, it is possible to see how many gardens hold a particular plant, whether the specimens came from cer- tified wild stock, and whether gardens can provide seeds or duplicate plants. Botanists can tell at a glance which plants are held only at one or two gardens and which are widespread, and so priorities can be adjusted accordingly. The database actually discovered one plant thought to be extinct. Sophora toromiro, once a native of Easter Island in the Pacific, had been assumed extinct for 50 years until the computer traced specimens in the Gothenburg Botanic Garden, Sweden. The activities of the Threatened Plants Unit at Kew are not just in the interests of science and the love of nature. Many endangered species may hold great economic potential for man, even if the potential has not yet been recognized. A good example is the Yicib (or Yeheb) a bush that used to be common in Ethiopia and Somalia. But famine, drought, and over-grazing have taken their toll. Tasty Nut The Yicib bears a tasty nut that is rich in protein. It was once a traditional food, grew wild, and was ideal for passing nomads. But the expanding population has taken all the nuts, leav- ing none to seed new plants. Even the bushes themselves have been used for firewood. Now, Yicib is on the lUCN's list of the 12 most endangered plants. Another on the list - the African violet - has become the symbol of the Plants Campaign. In cultivation, it is an abundant houseplant, with world trade valued at US$30 millon per year. But in the wild it is restricted to a couple of mountain forests in Tan- zania. Eighteen out of 20 species are unique to Tanzania and the com- monest houseplant, Saintpaulia ionan- tha, was known only on the Usambara mountains, already badly degraded by In 1983, Jon Lovett, a British botanist working for the WWF, found the species growing in the Uzungwa mountains. But a sawmill is under con- struction there with a view to logging 40 000 ha of the forest. The Plants Campaign has promised £60 000 to help turn part of Uzungwa into a nature reserve. Through the CMC database, it has been possible to draw the world's attention not only to the plight of the African violet, but to that of other threatened species in the same area. Other houseplants, for example the African primrose, Streptocarpus, and the busy lizzie, Impatiens, have been bred from only a few wild species. There are many more that are hardly known and which could offer tremen- dous potential to horticulturalists. Others may yield new medicines or 2-47 other products. Of 40 known wild mation on more plant species and will source, information on such things as species of coffee, ten are unique to become part of a worldwide network climate, soils, water resources, pol- Tanzania. Only three species have of environmental data. Already, it is lution, animals, and plants for any part been used so far for cultivation. Coffee part of the Global Environmental of the world. With such knowledge of is one of Tanzania's main exports and Monitoring Service run by the United the full environmental implications, it the wild species could represent a Nations Environment Programme should be possible to plan projects valuable new resource. (UNEP). with greater care and concern for the UNEP is now setting up an even bigger natural world, system called GRID, which will com- (LPS) Worldwide Network bine all the data into one huge com In the future, the database at the CMC puter network. Decision makers will is to expand to include more infor- then be able fo find, from a single Moves towards a cashless society by David Lascelles First came cash, then cheques, and small number of banks covering the minal at the check-out point. Each then credit cards. The next step along whole country with their many bran shopper is identified by punching in a the way the world pays for its goods ches. It is much more suitable than the secret personal identification number, could consist of electronic blips and sprawling United States of America or and the shop assistant keys in the flashes. West Germany, with its multitude of amount of the transaction. The As befits the computer age, bankers small banks. message goes down a dedicated tele- and retailers in many countries are Not that the British have reached their phone line to the shopper's bank studying push-button payment that decision easily. The banks dithered for where a computer checks that there is could largely do away with paper years over whether EFTPOS was enough money in the account to pay transactions and revolutionize the worth the massive investment the bill. meaning of money. But. like all revolu which will be several hundred million If there is, it debits the account and tions, their deliberations are fraught pounds - and whether people even sends a message to the shop's bank to with uncertainty, not least because the wanted it. Nor could they agree with credit its account there. If there are not shopper may prefer to stick with good retailers how to share out the cost of sufficient funds, the message comes old-fashioned notes and coins. installing the special terminals and back to the check-out counter that the The new method goes by the odd hooking them up to bank computers customer cannot pay, and the trans- name of EFTPOS, which stands for via telephone lines. action is cancelled. All this can be Electronic Funds Transfer at Point Of On the other hand, banks were faced done in less than 20 seconds which is Sale. The idea is that the shopper pays with a steadily rising tide of paper- about the same as a cash transaction for goods at a check-out through a ter- work, with the proliferation of cheques takes and is much less than writing out minal which instantly transfers money and credit card slips, and they knew it a cheque or credit card voucher, from his or her bank account into that would engulf them in the end unless For the banks, EFTPOS offers poten- of the shop. No coins or paper slips are they acted. So they have agreed to a tially huge gains in efficiency: the involved: only electronic messages pilot project in two year's time with big whole process is rapid and automatic, whizzing between banks and ter- retail chains in Britain like Burton and and cuts out vast amounts of paper- minals. Debenhams. This will pave the way to work and human labour. Ironically, It is not as futuristic as it seems, the fully fledged system they now however, the banks' own projections Several countries in Europe, the Far want, once acceptable charge rates for show that cheques and credit card East and North America already have transactions are set and technical slips will continue to grow even after small EFTPOS systems, most of them details agreed with retailers. EFTPOS is launched because of the operated by chains of shops or petrol The share-out of cost between banks burgeoning level of retail transactions, stations. But none has caught on in a and shops will be crucial to the sue- But bankers say that without EFTPOS, big way and some did quite the cess of the British experiment. To the growth of paper would be even opposite, conspicuously failing to avoid an unseemly squabble. Deloitte faster. And perhaps by the 21st cen- appeal to shoppers. Haskins & Sells, a leading account- tury, the tide will start turning against ancy firm, has been engaged to inves- paper slip transactions, tigate the benefits of EFTPOS and its Experimental Territory report will become the basis for cost However, EFTPOS received a big allocaton. There is still a danger, how- Cheaper And Quicker boost with the decision by leading ever, that the storeowners, large and For the shops, EFTPOS offers speed English and Scottish clearing banks small — who have deeper reservations and certainty of payment and, and retailers in January 1985 to set up about EFTPOS than do banks - will therefore, fewer bad debts. Because what is likely to be the world's first baulk when they learn what their con- the money is transferred immediately, nationwide system, with 250 000 ter- tributions are to be. it will improve shops' cash flow. EFT- minals usable by 20 million people by POS could also reduce the need to the end of the decade. The United keep tills full of coins, and it will main- Kingdom is, in many ways, an ideal How It Works tain a detailed record of transactions, proving ground: geographically a small EFTPOS comes in several forms, including issuing receipts, which will country with a large urban population. Usually shoppers have plastic cards help with accounting. All told, this a sophisticated banking system, and a which they insert into a special ter- should offset a good part if not- all of 12-48 elektor India december 1 985 the cost of the system. Bob Woodman of Burton, who heads the retail consortium's EFTPOS policy committee, said: "EFTPOS will be a cheaper and quicker way of making payment in shops." Less clear, however, is what EFTPOS does for the shopper. The banks say it speeds up the check-out and saves the customer having to carry money around or sign cheques and credit card slips. On the other hand, EFTPOS means that a customer's bank account is debited right away whereas there is a period of credit if payment is by cheque or credit card. However, some EFTPOS systems debit a credit card account rather than a bank account, and even those that are plugged into a conventional bank account can have inbuilt delays before the amount is actually debited. Educating The Public However, no matter whether the banks or the stores end up paying the lion's share of the cost, it will all eventually have to be passed on to the consumer. So unless EFTPOS produces genuine gains in efficiency, it could push up prices — and sharpen consumer hostility. The banks hope people will take to the new equipment as quickly and easily as they did to cash dispensing machines, which took about three years. But they are making no assump- tions. A big campaign to educate shoppers to the new technology is being developed, and video cassettes are already being circulated showing happy people using Britain's only sizeable existing EFTPOS system, that operated by the Clydesdale Bank at a small number of supermarkets and petrol stations in Scotland. There is also the worry of security. Although EFTPOS operators can never guarantee 100 per cent safety, planners of the British system say the secret codes and equipment they use will pre- vent fraudsters tapping other people's accounts or altering EFTPOS messages. Generally, however, EFTPOS could help reduce robbery by cutting the use of cash and fraud by using an electronic system that should be very hard to crack. Highly Sensitive Project Banks expect the biggest users to be petrol stations, department stores, and supermarkets, but there is no reason why EFTPOS terminals should not be installed in other busy places like railway and airline terminals, travel agencies, and restaurants. Special equipment has even been designed for installation in small corner shops, although these would probably not be hooked up to central computers all the time. EFTPOS will operate on British Telecom's IBT) telephone network with equipment and software made by IBM UK, subsidiary of the American computing giant. BT and IBM UK have been co-operating for some time on information systems in Britain, and made an EFTPOS proposal that was accepted. However, the award of business in this huge project is highly sensitive because of the large sums of money and jobs involved, and the British Government is keen to see other suppliers come forward when the system gets under way. Britain's largest indigenous computer company, international Computers Ltd IICL), will probably play a role, as will companies in the United States of America such as NCR, a leader in high technology banking hardware. Several European countries are also interested, particularly France where EFTPOS systems have been in operation for a number of years but have never been pulled together nationally like the British scheme. Helpful To Everyone An important consideration for the designers of the system is the danger of being accused of setting up a cartel. Britain's Office of Fair Trading, which watches out for signs of excessive col- laboration between competitors, is keeping a close eye on EFTPOS and is being consulted by the banks as they go along. David Robinson, general manager of Williams & Glyn's Bank, who chairs the clearing banks' EFTPOS policy committee, said: "We are all aware of the complex work that lies ahead but are determined that in developing the system and network, the right balance will be struck between cooperation and competition so that benefits will be shared among all the participants." While the banks will be operating the system together, they intend to market their EFTPOS products individually. Competition should produce variations in service. Some banks may promise customers longer delays before pur- chases are debited, in effect giving them credit. Others may charge less for EFTPOS accounts than their com- petitors, or build extra services into them such as credit cards and over- drafts. It will be a good ten years, however, before anyone can say that EFTPOS is the system of the future, and the i British experiment will doubtless be ] closely watched all round the world. 1 (LPS) 2-49 high-resolution colour graphics card - 3 The third instalment of an article describing a 512x512 or 512 x 256 pixel, black & white or colour, graphics card. by P Lavigne & D Meyer Circuit description Address decoding The eight most significant address lines, A e . . .A, 6 in Fig. 15, are applied to the P- inputs of 8-bit comparator IC , The O-inputs of this IC are polarized by resistors . ./? 8 and switches S|...S,. When the most significant address byte on the bus is the same as the binary word programmed by the user with S,. . .Sg, output P = Q of IC, goes low, which enables IC 2 to read address lines A-. . . A 7 . Since address line A 7 is tied to enable input G 2A of IC 2 , this circuit is active only for addresses between XX, Y and XX 7Y . where XX is defined by A 8 . . -A, 5 of IC ; and Y by IC 3 . Note that the presence of 2 on enable input G ; of IC 2 ensures that the address decoding is synchronized with the signals of the 6502. Of the eight outputs of IC 2 , only two are used, each of which defines a block of six- teen addresses. One activates the GDP's E(nable) input so that this is decoded between XX M and XX SF ; the other enables IC 3 , which deals with the sixteen bytes between XX,, and XX 6r . The first address used in thi s blo ck is XX 64 . Presenting the R/W (read/write) line to input A, of IC 3 (low for write; high for read operations) allows some increase in efficiency. Two different decoding signals are obtained at the same address, depending on whether the oper- ation in progress is a read or a write. This means that when XX, 4 is written, it is input to IC 12 , but when XX, 4 is read, IC )3 is accessed. At address XX, 5 (XX is user-defined), is write-only register IC 8 . Read address XX, 8 is not used. Address XX 66 is the location of write-only register IC„. Once again, this register's read address is not used. When the GDP (IC5) is active, associated data buffer IC 4 is also active. The data transfer direction is from data bus to GDP for writing and from GDP to data bus for reading. This is defined by the R/W line from the 6502, which is applied to pin 1 of IC ; . The other registers on the graphics card communicate direct with the data bus. The reason for this will be considered later. The address decoding is summarized in Table 3. has absolutely no effect on the operation of the devices. GDP and control signals The GDP. with its eleven addressable registers that are accessible from the micro- processor bus, is the heart of the circuit. Pins DAD,. . .DAD 6 output the addresses of bytes that have to be accessed for reading (display), writing, or refreshing. As it hap- pens, the first word to appear on these pins is A,. . . A 6 , ie„ the address line for the RAM (RAS). The second word output is A 8 . . . A 14 of the RAM. Initially, consider the circuit as if IC 8 and IC 9 do not exist, and the DAD outputs are connected direct to inputs A,. .A, of the memory. The timing for the entire card is controlled by clock CK. This clock is derived from the dot-clock generated by the quartz oscillator based on N, 6 . . .N 18 . The frequency of this oscillator is 12 MHz in the EF9367. and 14 MHz in the EF9366, which is manifested by a smaller image in the former than in the latter. The high clock — HCK — signal times out- put shift register IC15 direct. Apart from this, it is also fed to IC 29 , a counter that con- tinually repeats all possible binary con- figurations from 7 to 0 on its 0,. . ,Q 2 pins. This provides a cycle of eight clock pulses (= eight pixels = o ne b y te), which estab- lishes signals CK, RAS, CAS, STR, and A 7 via PROM IC3Q. The three least significant address lines of this PROM are controlled by IC 29 , so that the addresses from 7 tot 0 are scrolled continually. The most significant address lines are con- trolled either by PS, and PS,, or by PS, and MSL3X. In the sequential scanning mode, PS, and PS, allow 4x8 different addresses in the PROM to be accessed. The signals in these four blocks are identical, with the exception of A 7 . In the interlaced mode, MSL3X replaces PS,, and PS, takes care of the switching between the two high- resolution pages in the memory. The signals output by IC 30 are synchron- ized to the HCK signal via six bistables con- tained in IC, 6 . A complete listing of the contents of the PROM is given in Table 4. The ti ming diag ram of Fig. 16 shows the HCK, EES, CAS, STR, and LlJ signals - which are independent of the page selected — and the A 7 signal, which is determined by the condition of lines PS, and PS, or MSL3X. When CK is high, the line addresses are enabled by RAS; when it is low, the column addresses are enabled by the CAS signal. Whereas CES is applied direct to the memories, RAS is not, because, as has been stated before, there must be some means of 12-50 Table 4 The signals showi in Fig. 16 are generated the eight address blocks of PROM IC30 Whatever the block concerned (anc this depends on which page has been selected ti fill the screen), reading takes place from the most significant to the least significant address, e g., 87 06 05 84 03 02 01 00 07 06 . 01 00 07, etc. discriminating collective from individual accessing. Theref ore, th e RAS pulse is com- bined with the ALL and MSLj . . . MSL ? signals. ALL is, of course, low to indicate col- lective memory access. The MSLj . . . MSL 2 lines specify which of the eight bits in a byte is accessed. A second PROM, IC> 0 , is used for the decoding of the RAS. ALL, and MSL signals. Its contents are listed in Table 5, which shows that outputs RASj. . .RAS, are activated together only if both RAS and ALL are active, ie., low, simultaneously. When RAS alone is low, the output that is activated is determined by MSL,. . .MSL2. Applying the RAS signal to a dynamic memory IC not only enables addr esses A„ . . . A,, but also, in combination with CAS, addresses A 8 . . . A, 5 . The result is that selec- ting RAS for one eighth of every cycle selects one bit from every byte. A SM applied to a me mory IC that has not already received a RAS has no effect. The leading edge of the STR signal activates buffer 1C 7 , which toggles the levels of the MSL, ALL, BLK, DIN, and DW lines at the start of each cycle. These logic levels are only active for a much shorter time than the period of the CK signal: hence the need for an intermediate register. The function of N 5 , IC 6 , IC e , and 1C 9 will be discussed later. PROM 82S123 IIC30) £ £ s m O s 1 2 $ 3 s 1]I DATA HEX R4 R.3 R? R1 R0 §X ■ < Bo’ D6 D5 D4 03 02 ojoe 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 00 01 02 03 04 06 06 07 y 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 j ; a ' ||| ill 1 , 0 0 0 0 1 0 0 ! 0 0 0 0 0 0 0 16 9* 10 1C 1C 1C 17 0 3 0 0 0 ] 0 0 ; ! 0 0 0 0 1 0 0 0 08 09 0A 0B 0C 00 0E 0F 0 0 0 0 0 « 0 3 0 0 0 0 0 0 0 0 0 0 0 0 ! 0 1 0 0 0 0 11 0 0 0 1 I 0 V 0 1 0 0 i 0 0 0 0 | 0 • ° 16 00 10 10 1C 3C 3C 37 i 0 0 0 0 0 0 0 rr 0 0 0 0 0 0 ? 0 0 0 ! 10 11 12 13 15 16 17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 SIS 36 20 30 38 3C 1C ] 0 j 0 0 0 0 0 1 0 1 0 0 1 18 1 A IB ID IE IF 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ; 0 1 11 0 0 0 0 0 1 0 0 0 I 1 0 0 0 0 1 0 0 : 0 0 ; 0 0 ! 0 Hil 36 20 30 30 3C 3C 3C 37 The memory and its peripherals The memory on the mother board consists of eight Type 4164 (64Kxl bit) ICs. These have a common CM line, but separate RAS inputs. Address lines A„ . . . A 6 are con- trolled by the GDP. but A, is driven by IC 30 via IC3,. Every address appearing on multiplexed address lines (A 0 . . . A 6 ) + A, refers to a com- plete byte. If only one of the eight RAS lines is active, a single bit is selected. Each of the Q 0 . . ,Q, lines of the memory may have one of two outputs. In collective accessing, these lines are all active when shift register IC 15 is being loaded. The LD signal is applied to the SH/L (shift/load) input of IQ l5 via N,, so that the shift register gets loaded at the end of every cycle of IC 30 , but, because BLKX is fed to the same input, the register cannot be loaded outside the display window. If the BLKX signal were removed, the GDP would manipulate the memory at the edges of the screen, but this is not within the purposes of this article. The 0 H output of IC, 5 supplies the video signal obtained from the contents of the memory. If two successive bits in a byte are high, the video signal remains high until the end of the second bit. This serves to reduce flicker on the screen, and is also the reason for routeing the signal'via gate N 6 . This gate is triggered by the clock and compensates for any excess energy — see Fig. 17. The drawback of this arrangement is the need for a much larger bandwidth than that of a 12-52 elektor 1 typical monitor. In selective accessing, all the RAM outputs are inactive (forced high by r esistors R n . . .f? 34 ), except when the RAS line is active. The output of NAND gate N, 9 is low when the addressed bit is high (because the other seven bits are also high), and high when the addressed bit is low. The logic signal on the II line is, therefore, the con- verse of the addressed bit. The bit is loaded into register IC, 3 and is read by the host microprocessor via the data bus, D 0 , at address XX M . Note that the bit is loaded into register IC 13 only when t he puls e enabling the LD signal arrives and MFREX is active, indicating that the selected bit originates from the coordinates fed to the GDP by the microprocessor. The II signal is also fed to the colour selec- tion circuit containing the RMW logic already described. Writing to the screen memory The screen memory can be written to only when the DW line is activated by the GDP. The write signal is appl ied to the WRITE pin of the RAMs when the WRIS (write I select) line is activated to select memory plane 1 in the colou r mod e. In the monochrome ver- sion, the WRI S lin e is active continuously. Whether the DW signed is applied to the RAM depends on the logic level output by AND gate N 9 , which is part of the RMW cir- cuit. Initially, assume that this output is low and that WR is, consequently, activated by the DW signal. The data to be written into the memory appears on the DINX line. It can be fed to RAM input D ln via N 3 and N 4 only in combi- nation with the level on the DIS (data I sel- ect) line and the output of AND gate N| 0 . When the latter gate (RMW logic) outputs a low level, and DINX is high, the GDP tends to quench the pixel in question. The DIS line, whatever its state, can then change nothing: the only external method of preventing the pixel being dimmed is taking the WRIS line in IC I2 high, which disables the writing to plane 1. When DINX is low, the GDP tends to light the relevant pixel. Should the colour combi- nation demand that the pixel be dark (eg., when the pen and paper colours are differ- ent), t he DIS line must be taken high while WRIS is active. This is of importance for the colour facility only. It is essential to include this from the start, however, because it can- not be added at a later stage — see Table 6. In the read-modify-write circuits — Fig.12 and 13 — gates N, and N 4 , connected to register IC 12 , are used to select the colours. The other gates, N 2 , N 3 , N 9 , and N 10 are associated with the RMWs line provided by register IC M . When this line is low, it is as if the four gates do not exist; when it is high, however, the circuit is in the RMW mode. Here again, when the GDP tends to quench a pixel (DINX = high), it cannot be stopped, except when- the colour select logic disables the writing to one or more memory planes. The combination of RMWS and L5 signals disables writing in the RMW mode, except at the precise moment when the bit 16 HCKjinjuuinnnjinnnnjuui Fig. 16 This diagram shows that the timing of the memory activating logic combinations at A7. Note that these combi- Table 5 PROM IC, 0 is not read cyclically like IC30. since its role - when RAS is low - is to RASg. . . RAS7 when ALL them one after another when ALL has ceased to logic combination with lines MSL. 12-53 to be changed is output by N 19 — see Fig.16. The possible combinations of RMW logic in the monochrome mode are shown in Table 2. Further possibilities will be dis- cussed later. Table 6 Some examples of logic combinations of the colour select lines when the read-modify-write mode is disabled. Scroll The scroll logic consists of OR gate N s , register IC 6 , and high-speed 4-bit adders IC 8 and IC 9 . The manner in which the GDP relates the addresses on the screen to the memory addresses has already been dealt with. Since the GDP has not a scroll function (which is of particular interest in alphanumeric applications), the vertical addresses provided by the GDP must be modified. This is effected by the read- modify-write circuits. There are two suitable types of scroll: one changes the screen contents with respect to the memory contents, and is quite complex; the other, which is much simpler, makes use of the visual effect of an endless drum revolving in front of the viewer. The latter type is used in the present card. The vertical addresses (y-axis) are present during part of the addressing cycle only, since they are multiplexed with the horizon- tal addresses (x-axis). The latter are not modified, because this would result in horizontal scroll or a combination of vertical and horizontal scrolls. Instead of demultiplexing the addresses, a basic characteristic of binary logic is used. Adders IC 8 and IC 9 receive the GDP address bits on their A-inputs, while their 15- inputs accept a 7-bit binary word that is added to the A-word. The sum, which appears at the Q outputs, is either the unchanged horizontal address or the modified vertical address needed for the scroll. The 7-bit word, provided by IC 6 , is determined by the software, and is based on the position of the cursor relative to the bot- tom of the screen. The horizontal and vertical addresses, which follow each other at pins DAD, . . . DAD 6 , arp demultiplexed by N 5 . This gate combines two signals: BLKX, which is active (hig h) outside the display window; and RAS, which enables the horizontal addresses^ The resulting MUX signal controls the OE (output enable) cir- cuit of IC 6 ; the outputs of this chip are high impedance when the device is inactive. The MUX signal is also applied to the Cl (carry in) pin of IC 8 : when it is high, the word fed to be B-inputs of the adders is 111 1111 (because of resistors /? 9 . . .7?| 5 ). In this state, anything input to the A-inputs of the adders appears unchanged at their O-outputs. The MUX line is. therefore, kept high when no addition is required, as for positions outside the display window, and for horizontal addresses within the window that must not be modified. The BLKX signal disables the adders out- side the window, while inside the window, for vertical addresses only, they are enabled by RAS. The procedure described may seem somewhat contradictory, but this is only apparent, because the purpose of a cor- rectly specified RAS signal is the enabling of horizontal lines. The horizontal addresses are seen by the RAMs in the state they have on the trailing edge of RAS: this is one of the characteristics of dynamic RAMs. The logic levels output by the adders are, therefore, exactly as input to A when RAS goes low. The horizontal addresses are unchanged, but the vertical addresses appearing on the A-inputs will be added to the word fed to the B-inputs. When 1C 8 and IC 9 are returned to the by-pass mode (by RAS reverting to 1 at the end of the addressing cycle), the modified vertical addresses are enabled by the CAS signal. A final note about register 1C 6 : when this contains FF HEX (=1111 1111), both vertical and horizontal addresses remain unchanged. M Pan 4 will appear in our January 1986 issue. Ludus was conceived as a teaching aid in reading and writing classes, enabling the student to quickly learn, for example, the distinction between upper and lower case letters. The unit compares keyed-in given answers (yes/no or right/wrong) with previously programmed correct ones. It is therefore eminently suitable for use wherever a simple distinction between yes or no is to be made for a small outlay. programmable learn-and-play unit H. Rehbock The unit provides eight cycles of twelve questions each, that is a total of ninety-six questions. When all twelve questions in a cycle have been answered correctly, a light begins to flash. It has been found that this type of reward has a strong motivating effect on the user. The relative simplicity of the circuit limits the questions to two-choice ones which of course increases the chances of a 'good guess’. The answer to each question posed is given by pressing either of the two keys, Lfleft) or R(ight). This answer is compared with that previously stored in a register by means of the same two keys. If the two are the same, a point counter advances one step. Resetting is effected by pressing the reset switch S5 or the input keys simultaneously. operation When input keys S2 and S3 are pressed, debounce flip-flops N1/N2 and N3/N4 are actuated; when the switches are released, the flip-flops return to their quiescent state. The output state of gates N2 and N4 is compared in gates NS and N6 with out- puts 0 and 0 of an 8-bit serial shift register, IC2. If the outputs of N2 and 0 are both logic 1 (which means that switch S2 — left — was pressed correctly), the output of N5 goes logic low. Gate N8 then imparts a 1 to input B of monostable MMV2 which in turn triggers point- counter ICS. The same happens when the output of N4 and 0 are both 1. Task counter IC4 is triggered by monostable MMV1 and counts one up every time an input switch is pressed. The monostable is triggered at input B by the leading edge of the pulse coming from N1 or N3 via gate N7. The outputs of N1 or N3 are 0 when S2 or S3 respectively are pressed. Apart from the trigger pulses for the point and task counters, the two monostables provide the clock pulse for IC2 and the control signal for the result in- dicator lamp Lai. The time determining elements are C8/R14/D1 (MMV1) and C9/R15/D2 (MMV2). The trailing edge of the 20 ms pulses at output Q triggers the ap- propriate counter. At the same time, switch S4b connects a 0 to input A of MMV2 which enables the monostable to react to a leading edge (from N8) at input & When MMV1 changes state, the trailing edge at 0 causes a positive clock pulse to be given to shift register IC2 via N16 (provided the second input of this NOR gate is 0). Simultaneously, input A of MMV2 goes high which prevents the point counter and lamp being triggered. Monostable MMV2 is also rendered in- operative when switch S4b is open. The two 4-bit binary counters type 74LS93, apart from indicators, also function as con- trol elements. Outputs C and D of task counter IC4 provide a logic 1 to the inputs of NAND gate N10 after the twelfth input (binary 1100). The output of N10 is then 0 which keeps both monostables inactive until pressing reset switch S5 enables the acceptance of a fresh program. A further, important, function of counter IC4, together with NAND gate N9 and NOR gate N15, is to prevent once per pro- gram, in counter position 0110 (sixth input), the provision of a clock pulse to the shift register. In this situation both inputs of N15 are 0, one input of NOR gate N16 is therefore 1, and the clock pulse for IC2 is blocked. We now see how an 8-bit shift register may function as a 8 x 12-bit memory: in- puts A and B are connected to output 0 via switch S4a and IC2 then functions as a ring register. Because the clock pulse is absent at the 6th input, the contents of the register are run through after the 9th in- put: the last and first three steps in a pro- gram are therefore identical. The memory is thus shifted by three bits every program and the original line sequence does not recur until after eight through-runs. The output of NAND gate Nil is normally logic 1 and only becomes 0 when the user has reached the 12th input, that is, when the point counter indicates 1100. The output of NAND gate N12 is then 1 which actuates oscillator N29 and this causes lamp Lai to flash at a frequency of about 2 Hz via inverter N17 and driver stage Tl. lb keep the switch-on current of the lamp low, it is pre-heated by a quiescent cur- 2-56 2 rent provided by Rll. This current is kept at about 25 per cent of the operating cur- rent by emitter resistor R13. The value of R13 is thus determined by the operating current of the lamp used: about 27 Q for 30 mA, 10 Q for 80 mA, and 4Q7 for 160 mA. Oscillator N28 is normally blocked by the logic 0 at the output of NOR gate N13: its output is therefore high and this causes NOR gate N14 to keep input A of MMV1 low. When both input switches are pressed, the output of N13 goes high, and the oscillator causes the task counter to count to the next '12', and IC2 to shift to the next program line, via MMV1. To pre- vent unjustified points being awarded, the reset inputs of IC5 are connected to the output of N13, and in the above situation are logic 1. The normal reset pulse is derived from diode D3. teach you the binary coding if you are not already familiar with that* The best are miniature LEDs (may be obtainable in an array) which can be connected directly to the +5 V line via a biasing resistor. Make sure that the LEDs all have the same brightness before soldering them in. As regards the power supply, no more need be said than that the maximum cur- rent consumption is about 250 mA. Voltage regular IC1 must be fitted onto a heat sink. The circuit is best built into a case which is suitable for use on a desk or table top. programming It is not particularly difficult to devise your own program. The circuit should first be reset and switch S4 closed so that LED D4 lights. You then choose an arbitrary se- quence of eight L/R instructions for shift register IC2, for instance, R-L-L-L-L-R-L-L, and extend this arbitrarily with four Xs. Your input may then look like that shown in figure 2. Instruction 6 (R) is not written in the first program line, because the clock pulse for IC2 is then suppressed. To avoid this in- formation getting lost, instruction 6 must be repeated. The first three inputs (R-L-L) are omitted from the ring register but are added at the back (second register con- tents). This procession is repeated until after eight runs the original input line recurs. In this method of programming shifts oc- cur and the program lines and the register contents are therefore not identical. As the program runs as a cycle, it may be read from any arbitrary position by selecting a line of the register contents and accepting the following line as the first program line. In this way it is possible to obtain eight program blocks from one (as, for instance, block A in Table 1), and by interchanging L and R, even sixteen. To start with, this will certainly be sufficient. After a pro- gram has been written, S4 should be set to its original position, and the circuit reset by means of S5. M construction We have chosen coloured LEDs for the binary indication, which are inexpensive and look good. As an aside, they quickly •None the less, especially for younger users, figure 3 shows an alternative way of indicating the number of the question and the points scored. 1 2-58 eleklof i Digi-Course II selex-7 In the six chapters of our Digi-Course covered so far. we have learnt about basic gates. These basic gates had fixed relations between inputs and outputs, defined by their individual truth tables. From this issue, we start the second phase of our Digi- Course. In this part we shall learn about sequential logic circuits. The term sequential means that the relations of the outputs are not only dependant on the present input conditions but depend also on the previous conditions of inputs and outputs. The word "sequential" is derived from the Latin word sequentia which means succession A simple example of sequential operation of a device is the table lamp with a push button switch. When you push the switch the lamp lights. Push the same switch again and the lamp goes off. The action is same but the results are totally different. This example illustrates the meaning of sequential logic. If we consider the push button switch as the input and lamp as the output, we can see that for an input condition-"push the switch" the output condition — "lamp lights" is either true or false depending on the previous condition of the output. FLIP FLOP The first sequential logic circuit that we shall see is the Flip flop.This consists of two NAND gates connected together as shown in figure 1 . The truth table for this circuit is not as simple as the truth tables we have so far seen for the static logic gates. Table 1. S R I Q Q 0 0 1 1 0 11 0 10 0 1 1 1 | 0/1 I/O This truth table will have input-output relations which are time dependant. The truth table for the circuit of figure 1 is given in table 1. It has three inpu* combination which have definite output combinations each. The last input combination, where both the inputs are ”1" has no definite output combination, it depends on the previous conditions on outputs Q and Q. the previous condition of Q and 5 is retained when R & S_are both "1 ", The flipflop inputs are marked S and R Decause they are called Set and Reset inputs to the circuit, the dashes above S and R indicate that the active value of these inputs is "O". The dash is the_sign of logical reversal. The outputs are called Q and Q but the letter Q here has no special significance. The flip flops are practically used as logic switches. The circuit of figure 1 is in a simplified form given for understanding the basic principle behind the operation of flip flops. In the case of practical flip flops the two outputs Q and Q are always opposite of each other. The term "Set" means setting the output Q to "1" and "Reset” means resetting it to "O" The flip flops are frequently used in practical applications alongwith mechanical switches and relays. 12-59 y9— 5 OUTPUT OUTPUT nriui_ Mechanical Contact However, this circuit has a different truth table compared to the truth table of the circuit shown in figure 1. This is natural as we have used NOR gates in place of NANO gates. The input condition when both R & S are "O" retains the previous conditions of Q and Q Flipflop output The flip flop switches smoothly just by giving a pulse to the set or reset input; and has no transient noise like a mechanical contact. This is a very critical requirement in case of sensitive measuring circuits which can pick up all the transient noise of a mechanical contact and cause malfunctioning of the system. Similar to the flip flop of figure 1 constructed using two NAND gates, we can also connect two NOR gates to obtain a flip flop. The S and R inputs have no dash shown above them here, because the set or reset is done by a "1 " and not by a "O” as in the previous circuit. These two circuits can be connected on the Digilex Board and the input output relations can be studied by giving suitable input values of "O” and "1”. Remember that an open, unconnected input on any gate will behave as if it was on "1". As the inputs of this type of flip flop are for Reset/Set function, this flip flop is called the RS-Flipflop. Four of such RS flip flops are available in a single 1C, 74279. Two of these flipflops have one of their NAND gate with 3 inputs. The schematic diagram for this 1C is shown in figure 5 below. SN 74 279 selex Audible Continuity Tester selex the current flowing through R3 charges the capacitor Cl . As the charge on Cl builds up to such a level that 'the base voltage of T2 exceeds 0.7V, T2 starts conducting due to this forward bias on its base. As a result of this, the collector voltage of T2 suddenly drops, and due to capacitor C2 the base voltage of T1 suddenly drops. This low voltage on base of T1 brings it out of conduction and it is now cut off. This time resistor R2 charges capacitor C2 till the voltage on base of T1 again rises to about 0.7V, and T1 again starts conducting. This cycle continues so long as the power supply is connected. If we see the emitter current of T2 during this process, we find that it repeatedly becomes low and high at a frequency determined by the values of Cl, C2, R3, R2. In our circuit, the frequency of oscillation is around 1000 cycles per second (1 KHz). The oscillating emitter current of T2 causes T3 to pass an oscillating current through its collector circuit which contains the speaker. This oscillating current through the speaker produces a sound of 1 KHz frequency which we hear as continuous tone. If the supply is given to the circuit momentarily, it produces a beep sound. As we have already seen, the circuit of figure 2 can test conductive paths having resistance only upto 1000. For testing conductive paths with high resistances, the circuit needs to be modified as shown in figure 4. This modified circuit will work with contact resistances even upto 1 MO. If you want a tester with both facilities, use a three position switch as shown in figure 5. The entire circuit can be assembled on a small standard PCB as per the layout shown in figure 6. Avoid using bare copper wires for jumpers as it may create short circuits unintentionaly. Start soldering the passive components first, then solder the transistor. The PCB and speaker can be fixed in a small case as shown in the photograph. Speaker can be directly pasted with adhesive. (Be careful not to drop any adhesive on the cone of the speaker.) Soldering wires to banana sockets requires patience, because the area being larger, it takes quite some time to get hot enough for a good soldered joint. Figure 4. The moditied circuit of tester for high resistance Figure 5. 3-Position switch connections for the High-Low Ohms combined version. Component List R1. R4 = 2.2 Kft R2. R3 = 56 K ft R5 = 10 ft Cl. C2 = 10 nF T1 . T2. T3 - BC 547 LS = 8ft Loudspeaker (0.2W) Other parts 1 Battery holder for 2 penlight cells. 2 Penlight cells. 1 Suitable Case. 2 Banana Sockets. 2 Banana plugs. 2 Crocodile Clips. 1 3-position double pole switch. 1 Standard PCB and suitable length of insulated flexible wire. selex Potentiometers In the last issue we have studied the properties of Resistor combinations - parallel and series. We have seen how two resistors connected in series and the combination across a voltage source behaves. The two resistors R1 and R2 divide the voltage of the source between themselves in such a way that current through both of them remains same, as they are connected in series. The voltage across each of them is simply the product of the current and the resistance value in Ohms. 1 Figure 1 shows a practical circuit which has a series combination of two resistors R1 and R2 across a voltage source. The resistance values are 1000 and 2200 (approximately this gives a ratio of R1/(R1+R2) as 1/3. This means that the voltage U1 across the resistor R1 will be about 1/3 of the total battery voltage, the exact value being 1.4V. Assuming that this is the voltage we wanted to derive for some practical application from the 4.5V battery, it means that the remaining 2/3 voltage which appears across R2 will be wasted as heat dissipated by heating up the resistance R2. The voltage dividers using resistances like this are the simplest means of obtaining the required voltages from a larger source. However, as we have seen it is the most inefficient method of doing so. Fortunately, as the total i energy consumption of electronic circuits is generally very small, the inefficieny of the voltage dividers can be over looked in favour of the simplicity they present. For AC voltages, transformers are the most efficient voltage dividers, and series combinations of resistances are seldom used. Potentiometers The voltage divider in the most widely used form is a "Potentiometer", generally refered to as "Pot" for simplicity. A potentiometer is not just a voltage divider similar to the one seen before, but it is a variable voltage divider. The R1/R2 ratio is variable, keeping (R1+R2) constant. The specified value of a potentiometer is this sum (R1 + R2); which in fact is a single resistance and not a combination of R1 & R2. The resitance is either made of a track of resistive material or may consist 2 Figure 1 The current in the circuit flows through both R1 & R2. 'developing voltage drop proportional to their individual values. The resistors thus divide the battery The slider contact of the potentiometer varies the individual values of R1 and R2 such that R1 + R2 remains constant equal to the potentiometer's total 1 resistance value. The ratio R1/(R1+R2) thus becomes variable. 12-63 12-64 selex Polarity Tester Figure 2. Complete circuit diagram of the Polarity Tester. As you might have correctly guessed, a polarity tester can be constructed by connecting a diode and a lamp in series. When a voltage is applied as shown in figure 1, The lamp will glow. If the polarity is reversed the lamp will not glow (Figure 1.) Even though this circuit is very simple to construct and use, unfortunately it has a very important disadvantage. It functions only if the voltage is sufficient to light up the bulb. With a small voltage, less than the value required by the blub, the circuit will not function with whatever polarity you connect. This flaw can be overcome, with a little help of electronics. Figure 2 shows a modified circuit of our simple polarity tester. Figure 1 . The lamp glows if the polarity of the test voltage is The basic circuit of the polarity tester consits of the lamp L a 1 the transistor T 2 and resistor R3. The transistor T2 works as an electronic switch for turning the lamp ON and OFF. Transistor T1 controls the operation of T2. How this happens is illustrated in Figure 3. It can be seen that when the Base - Emitter junction of transistor T1 is forward biased, it supplies base current to transistor T2. At the same time the collector current of T1 flows through R2. The combination of base and collector current cofTl leaves the emitter of T1 and is again distrubuted between the base of T2 and diodes D2 and D3. As the diodes are forward biased, they conduct. This gives rise to a voltage of (0.7 + 0.7) = 1 .4 Volts at the base of T2. The base emitter junction of T2. is also forward baised and produces a voltage of 0.7V. Effectively the voltage across R3 is 0.7V, independent of other voltages in the circuit. From our knowledge of Ohm's Law we can derive the emitter current of T2 to be 70mA I = V/R = 0.7V/1 0 fl • =70mA At this point you will recollect that we have already seer a similar circuit - the constant current source. The base current being small compared to the emitter current, we can consider the emitter current to be almost equal to the collector current. Thus the current passing through thelamp Lai is always 70mA, provided that T1 provides the base current for T2. This is possible when the voltage connected at the test terminals is of the correct polarity. In case a reverse voltage is applied. T1 does not supply the base current for T2 and the lamp does not glow. 2-65 selex With a reverse voltage at the test terminals, diode D1 conducts and provides protection to the remaining circuit from damage due to reverse polarity. 5 in case the test voltage is too low to drive a lamp, as we have seen before, the switch SI can be thrown to position 2 so that the 4.5V battery supplies the current for the lamp. The current drawn from the test voltage in this case will be just about 5 microamperes. (1 microampere is one milionth of an ampere) To reduce the current drain from the 4.5V battery, and LED can be used in plae of the lamp. This requires the resistor R3 to be changed to 4711 instead of 1011 Construction : The circuit is very simple to construct, and if it is constructed properly as per the layout shown in figure 5 it should work at the first attempt. The usual precautions must be taken and the terminals of diodes and transistors must be correctly identified to avoid wrong connections. A seperate switch for the battery is not provided, as the idle current is not more than 1 micro ampere. The polarity tester can be used for voltage from 3V to 45V, and the lower lever of range can be brought down to 1 ,5V by using an LED in place of lamp and using a 9V battery in place of the 4.5V battery. Component List : R1 = 22 kll R2 = 4.7 Kll R3 = 1011 D1 . D2. D3 = IN 41448 T1 , T2 = BC 547B/BC1 47B Lai = 3.7V/70mA Bulb SI = SPDT switch Other parts : 1 Standard PCB 1 4.5V (or 9V) Battery 1 Bulb socket 1 suitable cabinet Changes for LED version : 1 LED instead of Lamp Lai R3 = 4711 instead of 101! 12-66 EEC have developed a new Insulation tester which is intrinsically sate. It can be used even in hazardous areas, where danger of igniting combustible gases is always present. The insulation tester can measure low resistances upto 100 ohms and voltages upto 25 V AC/DC. The instrument has been developed at CMRS, Dhanbad and has been approved by CCE, Nagpur. DGMS, Dhanbad, and DG FASLI at Bombay. The instruments is housed in an unbreakable fibre glass enclosure and powered by rechargeable Ni— Cad cells. It is ideally suited for use in Chemical. Petrochemical, Oil & Gas Industries, Mines, Ammunition dumps MICROPROCESSOR TRAINER Hi-Rel Electronics manufacture adjust- able frequency AC Inverter Drives for various industrial applications. The range is upto 150 HP in 2 different models. Model HR-XL is upto 50 HP and Model HR-PF is above 50 HP. The drives are capable of controlling the speed of any squirred cage motor from 1 0% to 200% of the rated speed, without affecting motor characteristics. The drives can handle all types of loads including constant torque, constant HP and variable torque. It is claimed that the system can work with minimum maintenance, even in a hostile atmosphere. PEP now offers a microprocessor Trainer EC-85B based on the single power supply 8-bit microprocessor Intel 8085A. The circuit of EC-85B has been laid out on a single PCB and functional blocks of the system like microprocessor, ROM, RAM and ad- dres/date bus are marked on the component side for understanding of the system architecture. Six-digit hexadecimal display has been provided, alongwith a 28-key keyboard and 4K byte monitor. For further information, write to HI-REL Electronics Pvt. Ltd. Shanti Chamber. Opp. Dmesh Hall. Navrangpura, Ahmedabad-380 009. For further information, write to: Professional Electronic Products. Post Box No. 316, Delhi Road Meerut 250 002. For further information, write to Electrical Equipment Corporatic 13/63, Punjabi Bagh, New Delhi-110 026. TRANSISTORS DIGITAL MILLI OHM METER Solid State Electronics manufacture a complete range of Bipolar Power Transistors and Power Darling tons covering 30 to 1000 Volts in 1 to 100 Amps ratings. The series covered are BDW, BDX, BUR, BDY, BUT, BUV, BUW, BUX These are in addition to the previous JEDEC series. The power darling tons are fully integrated devices with biasing resistors, speed-up diodds with biasing resistors, speed-up diodes and fly back diode for fast switching. The Isb, tON, ts and tOFF are checked on cent per cent basis. The devices are offered in TO-39, TO-66 and TO-3 packages. For further information, writ Economy Electronics, 15, Sweet Home, 2nd Floor, 442, Pitamber Lane, Mahim, Bombay-400 016. For further information write to: Solid State Electronics Pvt. Ltd. 9/123, Marol Co-Operative Ind. Estate, J.B. Nagar, Bombay-400 059. For further information, write tc Brisk Electro Sales Pvt. Ltd. 390-A, Lamington Chambers, Lamington Road, Bombay-400 2-67 switchboard FOR SALE - Casio PB300 Personal Computer with printer 2 Speak & Spell American Teaching Aid For English. R.K. Kondaveeti. C-4/A. 36C, Janakpuri, New Oelhi 110 058. WANTED - Elektor UK Oecember 82 or photo copies ol crescendo published in that issue. Ajith Kumar. B.N.H.S. Research Centre. 331. Rajendra Nagar. Bharat pur. Rajasthan 321 001. WANTED — 1C 7137 & Eveready 950 size Nicad Rechargeable cells 10 Nos. Mukul Pani. Pant Niwas. Ranidhara Road. Almore 263 602 (U.P. Hills) WANTED - Full Details technical-PCB Etching and Drafting, For Preparing PC Patterns lor individual Requirements R. Murali Krishnan. No. 6. Choodiamman Pettai Street. Jones road. Saidapet. INTERESTED - In-Linear AC and DC Power Sensing Component or Circuit. Vijay Kumar Narsinghani. C/o. Sah Industrial Research. Institute. SA 15/171, Gautarn. Budh Raj Marg. Sarnath, INTERESTED — In-Contacting young short wave listeners also amateurs. Ganesh Prasad Sahu. Near Railway Gate. Premnagar Satna. M.P. Pin 485 001. WANTED - Full Details about satelhte TV signals Receiving Equipment dish antenna and Microwave to WHF converter. Padma Kumar U. Nedumpurathu. Thodi- yoor North. Thazhava. Kerala 690 523. WANTED - ZX-81 -want To Purchase have the manual (Or Copy) Separately if possible R. Cheema. 1162. 15-R Chandigarh. 160 015 (Phone 25214). WANTED 57105. MM53114. LM 1889. 4.433618 MHZ Crystal. Ne 5532 Exchange Possible also all one each. Except 5532-Needed. C. Sanjay. 11. Balaram Street. Adyar. Madras 600 020 WANTED 3 Digit Digital Event/Pulse counter with reset & blind temp controller 30 90C & 50-300'C H.S. Shab. 402. Indira Apts. Carmichael Road. Bombay 400 026 Tel. 4924037 WANTED — Circuit of Electroic Ingition used in two wheelers Mr Praveen Kumar Jain. 1 Parao Delhi Gate. Meerut City. Pine 250 002. INTERESTED — In purchasing an old Oefeclive personal computer SK. Agrawal, 31. Raja Manindra Road. Calcutta 37. WANTED — CRT Controller Chip EF 9364. Software for Character Generator ROM of Elektor Oct. 83 VOU Card. FNA -30 Display. Mr Sandeep Nirankari 3282-A. Sector-24-0, Chandigarh 160 023. WANTED — MCM 68 10-RAM (Motorola) N.S. Sehra, 201. Space Age Apartments. Green Acres. Din Ouarry road. Deonar. Bombay 400 088. WANTED - Information Regarding Amateur Radio Magazines. Manoj. TC- 13/1187. Kunnukuzhy. Trivandrum 37. Tel. No 70012. WANTEO - Circuit Diagram For Philips Two in one 90AR467/70R P V Singh. Mecon. Ispat Bhawan. Bhilai 490 001 WANTED - Ic Based ckts for 27MHZ am receiver and ICs S042P Tea 440. R.S. Muchakani. Kalabadevi Bldg , Srinagar. Dharwar 580 003. SEEKING EXTENDED — Capability tor El-7001 Sharp memownter 2 availability of B8C micro spectrum or a commodore S N Saigal. C/o. General Industrial Co.. F/54. Okhla Industrial Area. Phase- 1. New Delhi 1 10 020. FOR SALE — 20W Stereoamp factory assembled frequency response 3042-18 5KHZ stable operation. Samuel Jayaraj. J . 240 Pillanna Garded. 57 Thomas Town. Bangalore 560 084 FOR SALE — Programmable Music Generator Cum Violin. R.V. Dhekale. 453. Ganapati Ali. Wai (Satara) 412 803. WANTED — Circuit of Electronic Ignition system used in two wheelers (Magneto Type). Mr. Anant Kumar Tudu. 6. ASB/2 Line. P.0. Mosaboni Mines 832 104- Bihar. FOR SALE -Commodore-64 wilh peri- pherals and software send S.A.E. for Details. Mr P Raja Sekhar. Flat No. 304. •Mehta Towers" 1-1-293/4, Ashok Nagar. Hyderabad 500 020 (A.P.) WANTED — Circuit Diagram for Hitachi TRK 808W cassette tape recorder with Radio. Mr Naresh Panjabi. 602 Blue Moon Apts. Cornerot. 14th and 30th Road Bandra, Bombay 400 050. Dehradun. U P. 248 001 INTERESTED — In-Exchange of software for ZX Spectrum. Sumit Kumar. 20 Montieth road, Egmore, Madras 600 FREE advertising lor our readers! deluding address and/or telephone number). One advertisement per reader per month. To enforce this rule, a switchboard voucher Ads will be placed in the ore they are received. We reserve the right to refusr BLOCK CAPITALS PLEASE - ONE CHARACTER TO EACH BOX ElEkTOR ElECTRONiCS pvT hd. R.N.No. 39881/83 MH/BYW-228 LIC No 91 OUT OF THE WORLD We are moving ahead. Our Cosmic energy, skill and long-standing experience of thirty years, have worked in creating a unique technological galaxy of electronic goods. One such rising star is our television model 3324. a colour Pnnwf & Publisher -CR chandarana. 2 Kouman. 1 4th a Road Khsr. Bombay-400 052 » Pnntedat Trupb OHsel 1 03. Vaaan udyog Bhavan. 00 Tula P|pe Road, lomr Parel. Bombay. 400 01 3