mwsmmi up-to-date electronics for lab and leisure • 0 2xl40watt mosfet power amp the power supply : to 30 volts, 3 amps a phototropic whatsit i ioniser : 7 kV off 12 volts contents <*) elektor december 1982 — 12-03 selektor 12-15 triopede 12-17 An idea from J. Cornelissen A cybernetic model belonging to the 'Triopodus Electrus Diclopus' family or TED for short. In normal terms it means: a light seeking, three wheeled, electromechanical hybrid. For your domestic pet it will be a close encoun- ter of the third kind. precision power supply 12-22 A precision power supply capable of providing up to 3 amps at 35 V, incorporating both current limiting and short circuit protection. Meters are included to enable current and voltage output levels to be monitored. Ideal for testing high power equipment! the XL system 12-29 An introduction to the new Elektor XL audio system. A brief outline of I the different parts which will ultimately form a complete high quality audio system. crescendo 12-30 The start to our XL system. A high quality symmetric/complementary MOSFET stereo power amplifier easily delivering MOW into 8 J2. It is short circuit proof, includes speaker protection, power meters, temperature and power-up indication. An amplifier which will leave any serious minded audiophile speechless! home telephone help 12-40 W. Wijnen The beginning of our technical answers section. Further modifications to the telephone system published in the September issue which improves the volume level of the speaker in the handset. soft switching An extra refinement to the HighCom sound reduction system published in the March 1981 issue. It provides softswitching by changing the response time of the circuit to increase the decay time, thus eliminating inter- ference. culmulative index volume 8 1982 in-car ioniser A theory which is rapidly gaining credence relates to the quantity of nega- tive ions in the air. This circuit is one way of increasing the concentration of these ions in the surrounding air, refreshing the environment. In other words fresh air on wheels! floppy-disk interface for the Junior - part II G. de Cuyper The second and last article describing the modifications that must be made to the hardware and software of the Junior, to be able to run Ohio Scien- tific software available on diskette. No tracks need to be cut and no mech- anical changes need to be made! a dozen and one sounds A single chip containing all the ingredients of a BBC effects laboratory, producing a large number of refreshing sounds. 12-41 12-42 12-45 12 48 12-59 The next issue may be more of a surprise than just being a new issue starting a brand hew year. As far as projects are concerned the January issue will continue with the XL system by introducing the accessories for the power amplifier, a video/audio modulator, and . . . it will be a great start to 1983! stop-signal override for model railways 12-62 Model railway enthusiasts know what happens when the stop signal dis- connects the supply voltage from a section of track. This little circuit gets over the problem and allows shunting in spite of the stop signal. switchboard 12-63 market 12-64 advertisers index 12-82 12-08 — elektor december 1982 advertisement aJLa. aJa ^Ta aJa vjv vjv vjw yjw vjy vjw yjv ^p ^p rp SPECIAL OFFER to Elektor readers 4|JL| a|a ^JU a|a a| a aWa aJa aIa a|a a|a aTa a|a aTa a|a a|a a|a ^ a|a a|a aJa aJa a|a ^p ^p ^p ^p ^p ^p ^p ^p ^p VJW ^p vjv ^p vjv ^p ^p vjw ^p ^p ^p ^p ^p ^p ^p VJV 3i DIGIT LED PANEL METER for only £11.95 + 75p P&P (inc.VAT) The DP 355 is a 3% digit LED panel meter with a full scale input of ± 1 .999 V D.C., and a basic accuracy of 0.5%. It has a very high input impedance (> 1 G£2) and good linearity, typically 0.1%. The whole instrument operates from a single 5 volt rail. Polarity and overload indication are automatically displayed. There is provision on the pcb to fit either attenuator resistors to alter the input voltage for full scale or shunt resistors for current measurement. The decimal points are user programmable to correspond to the full scale voltage/current. Full specifications and applications for use are included. Don’t miss this excellent offer closing date-31st Dec. 1982 Only a limited number of meters are available, so order now while stocks last. Complete the coupon below and mail it, with your cheque or postal order, to Elektor Publishers Ltd., lOLongport, Canterbury, Kent CT1 1PE. Please allow 28 days for delivery. Please send me panel meter/s @£ 11 .95 + 75p p&p each. I enclose cheque/P. 0. for (payable to Elektor Publishers Ltd.) NAME: ADDRESS: El 282 advertisement elektor december 1982 — 12-09 HOME LIGHTING KITS Theaa kns COM**" naw i»**y component* lull iniliuclions & «r« dmgncO 10 replace a standard wan twitch and control up 10 300 w o' lighting TDR300K Remote Control £14.30 MK6 TrlTnTmjtler for above £ 4.20 TD300K Touchdimmer £ 7.00 / | , ?: W *V i* o Extension kit for 2 way - _ switching for TD300K £2.00 v\ HOME CONTROL CENTRE This New Memote Control Kit enables you to control up to 16 different appliances any- where in the house from the comfort of your armchair. The transmitter injects coded pulses into the mains wiring which are received by receiver modules connected to the same mains supply and used to switch on the appliance addressed. Receivers are addressed by means of a 16-way keyboard, followed by an on or off command. Since pushing buttons can become rather boring, the transmitter also includes a computer interface so you can programme your favour- ite micro to switch lights, heating, electric sUSSI MW RADIO KIT Based on ZN4UIC. eludes PCB. wound aerial and crystal earpiece and all components to make a sensitive miniature radio lr. 5.5*27x2 cms. Reau.res PP3 9V b »« erv toBAL FOB BEGINNERS ite micro to switch lights, heating, electric blanket, make your coffee in the morning, etc., without rewiring your house. JUST THINK OF THE POSSIBILITIES The KIT includes all PCBs and components for one transmitter and two receivers, plus a drilled box for the transmitter. Order as XK112. £42.00 Additional Recievers XK111 £10.00 ELECTRONIC LOCK KIT XK101 This KIT contains a purpose designed lock 1C. 10-way keyboard, PCBs and all components to construct a Digital Lock, requiring a 4-key sequence to open and providing over 5000 different combinations. The open sequence may be easily changed by means of a pre- wired plug. Size: 7x6x3 cms. Supply: 5V to 15 V d.c. at 40uA. Ouput: 750mA max. Hundreds of uses for doors and garages, car anti-theft device, electronic equipment, etc. Will drive most relays direct. Full instructions supplied. ONLY £10.50 Electric lock mechanism for use with latch locks and above kit . DVM/ULTRA SENSITIVE THERMOMETER KIT Thu new design >s based on . Ih# ICL7126 (a lower power version of the ICL7106 chipl I , _ i | and i3'^ digit liquid crystal JtSHSJ display This kit will form the 1 basis of a digital multimeter (only a few additional resistors and switches re required — details supplied), or a sensitive ' digital thermometer l-50*C to + 150*0 reading to 0.1*C. The basic kit has a * sensitivity of 20OmV for a full scale reading, automatic polarity indication and an ultra low power requirement — giving a 2 year typical battery life from a standard 9V PP3 when used 8 hours a day, 7 days a week Price £15.50 DISCO LIGHTING KITS CM. 1000K This value-f or- money kit rtT - ] features a bi directional sequence, speed of sequence and frequency of direction change, being variable by means of potentiometers and incorporates a master dimming control. £14.60 DLZ100K A lower cost version of the above, featuring undireclional channel sequence with speed variable by means of a pre set pot. Outputs switched only at mams zero crossing points to reduce radm interfere" --e to a minimum. Optional opto input DLA1 Only £8.00 Allowing audio ("beat') —light response DL3000K This 3 channel sound to light kit features zero voltage switching, automatic level control & built m m>c. No connections to speaker or amp required. No knobs to adjust - simply connect to mams supply & lamps. Only £11.95 (IKw.ChanneO 1 "OPEN-SESAME" The XK103 is a general purpose infra red trans- mitter receiver with one momentary (normally open) relay contact and two latched transistor Output Designed primarily for controlling motorised garage doors and two auxiliary out m the home for switching lights. TV, closing curtains, etc Ideal for aged or disabled persons The Kit comprises a mains powered receiver, a four button transminer, complete with pre- drilled box. requiring a 9V battery and one opto isolated solid state switch kit for inter- facing the receiver to mains appliances As with all our kits, full instructions are supplied. ALL PRICES EXCLUDE VAT THE MULTI-PURPOSE TIMER HAS ARRIVED Now you can 'un your central heating, lighting, hi-fi system and lots more with just one programmable timer At your selection it ts designed to control four mains outputs independently, switching on and off at pre set times over a 7 day cycle, e g to control your central heating (including different switching times for weekendsl. just connect it to your system programme and set it and forget it — the clock will do the rest. Ik FEATURES INCLUDE VVl' * 0.5* LED 12 hour display. \ * Day of week, am pm and output status indicators * 4 zero voltage switched mains outputs. -■ \ - * 5060Hz mains operation. \ O'*-© * Battery backup saves stored programmes and continues ^ b time keeping during power failures (Battery not suppliedl a \\' r ~ ^ * Display blanking during power failure to conserve battery power ^ * 18 programme time sets _ «-^\0 'V. £ * Powerful "Everyday* function enabling output 7 O* k_S' to switch every day but use only one time set / y * Useful "sleep" function-turns on output for one hour I £ * Direct switch control enabling output to be turned on A a immediately or after a specified time interval 1 * 20 function keypad for programme entry 1 1 \ * Programme verification at the touch of a button. f 1 (Kit includes all components, PCB, assembly and programming instructions). ORDER AS CT5000 Safe ONLY £23.75 REMOTE CONTROL KITS MK6 SIMPLE INFRA RED TRANSMITTER Pulsed infra red source complete with hand held plastic box Requires a 9V battery £4.20 MK7 INFRA RED RECEIVER Single channel, range approx 20ft Mams powered with a tnac output to switch loads up to 500W at 240V ac £9 00 (RC500K -Special Price for MK6 and MK7 together £12.50 MK8 CODED INFRA RED TRANSMITTER Based on the SL490. the kit includes all components to make a coded transmitter and only requires a 9V (PP3) battery and keyboard 8x2x1 3cms £5.90 MK10 16-WAY KEYBOARD For use with MK8 and MK18 to generate 16 different codes for decoding by the ML928 or ML926 receiver (MK 12) kit £5.40 MK11 10-Channel + 3 Analogue o/p IR Receiver Based on ML922 decoder 1C Functions include on standby output, toggle, control of volume, tone and lamp brightness. Includes its own mams supply £12 00 MK12 16- CHANNEL IR RECEIVER For use with MK8 kit with 16 on off outputs, which with further interface circuitry, such as relays or tnecs. will switch up to 16 items of equipment on or off remotely Latched or momentary out puts - please specify when ordering Includes its own mams supply. £11.95 MK13 11-WAY KEYBOARD For use with MK8, MK18 and MK1 1 kits £4 35 MK16 Mains Powered IR Transmitter Mains powered for continuous operation - single channel, for applications such as burglar alarms, automatic door openers, etc. Range approx 6 ft £2.50 MK17 12V d c IR RECEIVER For use with MK6 or MK16. Relay output with DP 3 Amp change over contacts, may be used as latched, momentary or "break beam' receiver Operates from 6 13V d.c. . £9.50 MK18 HIGH POWER IR TRANSMITTER 4 0 Similar to MK8 but with range of approx 60ft £6 20 J\ Ancillary Kits MK2 Solid State Relay — fj Opto- isolated with zero voltage switching No. tnac supplied £2 60 MK15 DUAL LATCHED SOLID STATE RELAY Comprises 2 x solid stale relays and latch for use with momentary version of the MK 12. 2 output triacs required (not supplied! £4 50 . 24 HOUR CLOCK/APPLIANCE TIMER KIT Switches any appliance up to IkW on and off at present times once per day Kit contains AY-5-1230 1C. 0.5* LED display, mams supply, display drivers, switches. LEDs, triacs. PCBs and full instructions CT1000K Basic Kit CT1000K with white box (56 131 x 71mml (Ready Built) e SHORT FORM CATALOGUE - send SAE (6"x9"). We also stock Vero, Books, Resistors, Capacitors, Semi-Conductors etc. FAST SERVICE TOP QUALITY LOW LOW PRICES ELEKTOR PCBs NOW AVAILABLE Add 55p postage & packing + 15% VAT to total. Overseas Customers; Add €2.50 (Europe). €6.00 (elsewhere) lor p&p. Send S.A.E. for further STOCK DETAILS Goods by return subject to availability. APT Kl 9am to 5pm (Mon to Fri) vJl Clil 10am to 4pm (Sat) ELECTRONICS^ 11 Boston Road London W7 3SJ | TEL: 01-567 8910 ORDERS 01-579 9794 ENQUIRIES 01 -579 2842 TECHNICAL after 3i 3 CHANNEL MICRO- PHONE OPERATED SOUND TO LIGHT. 300 W PER CHANNEL. 240V A C. £11.58 mcV.A.T. 600 MHz PRESCALER f 5 V DC £ 28.42 inc V.A.T. TRANSISTOR AND DIODE TESTER £ 14.73 inc V.A.T. UK 666 0-30 V. 2.5 A REGULATED SUPPLY WITH DIGITAL DISPLAY £ 76.82 inc V.A.T SOUNO OPERATED SWITCH. 9-1 2 V D.C. COMPLETE WITH MICROPHONE £8.42 inc V.A.T. MINIATURE F M. TRANSMITTER NOT LICENCEABLE IN U K. £ 9.46 inc V.A.T advertisement elektor december 1982 — 12-11 ELECTRONIC KITS 7 HUGHENDEN ROAD, HASTINGS, SUSSEX. TN34 3TG Telephone: HASTINGS (0424) 436004 Post & Packing 50p for KS kits. 75p for UK kits, £1 .00 for Cabinets. Send 20p S.A.E for catalogue of our extensive range of kits & Cabinets. Trade. Educational & Export enquiries welcome. Mini FM Transmitter 88 108 MHz 9 Vd.c. (not licenceable in UK I £ 7.37 inc V.A.T. UK 232 # ANTENNA AMPLI FIE R FOR AM/FM. A M GAIN 25 dB FM 15 dB £5.27 inc V A T THREE WAY STEREO MIXER. PHONO. AUX. MICROPHONE £24.21 me V A T ... KS 262 T ' £ 10 CHANNEL CHASE LIGHT, 300W PER CHANNEL. 240V A.C. £ 14.73 me V .A .T STEREO STUDIO MIXER. INPUTS 2 PHONO. 1 TAPE. 1 AUX. 2 MICRO PHONE £69.45 me V A T 6 CHAN HIGH POWER VU DISPLAY. 300W PER CHAN. 240 V A.C. £ 15.79 inc V.A.T. PORTABLE SIGNAL TEACER, COMPLETE WITH R.F. PROBE £24.21 inc V.A.T. CENTRALIZED ANTI THEFT ALARM £ 33.67 me V A T. L E D. VU DISPLAY, INPUT 1-100W 5 12V DC £ 9.48 inc V A T KS 200 services to readers EPS print service Many Elektor circuits are accompanied by printed circuit designs. Some of these designs, but not all, are also available as ready-etched and pre-drilled boards, which can be ordered from any of our offices. A complete list of the available boards is published under the heading 'EPS print service' in every issue. Delivery i time is approximately three weeks. It should be noted however that only boards which have at some time been published in the EPS list are available; the fact that a design for a board is published 1 in a particular article does not necessarily imply that it can be supplied by Elektor. Technical queries j Please enclose a stamped, self-addressed envelope; readers outside UK please enclose an IRC instead of stamps. Letters should be addressed to the department concerned — TQE (Technical Queries). Although we feel that this is an essential service to readers, we regret that certain restrictions are necessary: 1. Questions that are not related to articles published in Elektor cannot be answered. 2. Questions concerning the connec- tion of Elektor designs to other units (e.g. existing equipment) can- not normally be answered. An answer can only be based on a com- parison of our design specifications with those of the other equipment. 3. Questions about suppliers for com- ponents are usually answered on the basis of advertisements, and readers can usually check these themselves. 4. As far as possible, answers will be on standard reply forms. We trust that our readers will understand the reasons for these restrictions. On the one hand we feel that all technical queries should be answered as quickly and completely as possible; on the other hand this must not lead to overloading of our technical staff as this could lead to blown fuses and reduced quality in future issues. Elektor, the first in the field to provide a printed circuit board.... and now, the first to bring you a fascia panel The Fascia Panel for the Elektor Artist Guitar Preamp joins the Elektor Printed Circuit Board Service. The Panel is manufactured from a high quality self-adhesive plastic which is scratch resistant and washable. Finished in three colours on a silver background it provides the elusive finishing touch. Price £1.70, available from Elektor Canterbury using the pre-paid order card in this issue. Elektor sets the pace selektor elektor december 1982 — 12-15 new sense of direction solid-state gyros . . . Conventional gyros have to be accu- rately machined and are very expensive. Moreover, they are delicate mechanisms vulnerable to shock or vibration. Far more robust and cheap to produce is a new device based on a cylinder's flex- ural mode of vibration and the effect of Coriolis force. Imagine a fleet of fork-lift trucks moving about a warehouse or factory, following precisely the routes they are instructed to take and automatically correcting any error in direction, but all without drivers. This is but one application foreseen for a novel device with a diameter of only 17 mm and 25 mm long. The device can be called variously a solid-state gyro or a rate gyro or a rate sensor. It was invented and is being developed by GEC Marconi Electronics at its research laboratories at Great Baddow in eastern England. The scientist concerned who demonstrated the device recently to members of the Royal Society in London is Dr R. M. Langdon, who has used the three different names and has researched the subject for a couple of years. The word gyro could perhaps be a little misleading because of our preconcep- tions. We know that it is a wheel or disc or cylinder rotating at a very high speed and often mounted in a complete set Basic shape of the solid-state gyro. of gimbals. Simple gyros have been sold as toys, but the complete mathematical behaviour of such a device is far from simple (it involves vector analysis) and need not be considered here. Gyros have had and still have many successful uses, for example as compasses for navigation, in stabilizers for ships, in monorail vehicles, and so on, apart from military applications. These are the gyros we are familiar with, but the new Marconi device is quite different. For one thing it has no heavy, fast- rotating wheel or cylinder driven by a motor. It has no bearings in which anything moves. Yet it can do what many conventional gyros can do. That is why it is often called a gyro. It measures the rate of rotation of any vehicle in which it is mounted, so it is of the type known as a rate gyro. Virtually indestructible The reason for the search for newer devices to do most of the things that mechanical gyros do it that the latter are very expensive: their machining has to be extremely accurate so that there is negligible friction and no imbalance. So the gyros that are used for such tasks as navigation are very expensive — the smallest may cost as much as £ 1000. Furthermore, they are vulnerable to severe shock or vi- bration. A much cheaper device to do the same job yet be so robust that it is virtually indestructible is obviously a desirable product with a big potential market. Several scientists and engineers have used their skills to invent such new devices, relying on periodic vibrations instead of mechanical rotation. The earliest was a tuning-fork mounted vertically, its stem being of such dimensions that any torsional vibration would be of the same frequency as that of the tines. When the whole fork was rotated, the torsional vibration set up was measured and was proportional to the rate at which the fork was rotating. It was in effect a rate gyro. Other devices have utilized vibrating wires or rods, but little is known about them. As for the tuning-fork device, despite intensive work in the USA and at the Royal Aircraft Establishment in Britain, it has not come into pro- duction, presumably because of the complexity needed for a satisfactory instrument and the cost of careful machining. Coriolis force At this stage of the argument it is necessary to introduce the notion of the Coriolis force, named after the French engineer and mathematician who postulated it in work published in 1835. He was concerned with the limitations of the Newtonian equations of motion in a rotating frame of reference. So he added another force in such a system, and when he did so the Newtonian equations could be used. (Physicists and mathematicians spent about a century searching for an 'absolute' frame of reference only to find that they were chasing a will- o'-the-wisp, so a radical change of approach was needed, culminating in Einstein's Relativity Theory.) Early use of the Coriolis-force Idea was in calculations of ballistic trajectories, be- cause the Earth is a rotating body. Oceanologists and meteorologists have invoked the Coriolis force to account for ocean and air motions. More recently scientists have used it in connection with rotating systems such as gyros. It is the Coriolis force that comes into the explanation of how the Marconi device behaves. Basically, the device is a cylinder of metal, so asymmetric difficulty is avoided. It looks somewhat like a wine-glass, and everyone knows that is such as glass is struck near the rim it emits a ringing sound, just as a bell does. The fundamental frequency of the tone it gives off depends on the dimen- sions and the elastic modulus of the material. If we could observe it from above, the vibrations would be seen as shown, grossly exaggerated, in the dia- grams a, b and c. So the first problem is how to produce and maintain the vibration. This is done by means of piezoelectric crystals. Such a crystal will produce a vibratory motion in response to an applied oscillatory Flexural vibration of the cylindrical cross-section, shown undistorted (al and subjected to forces as indicated in (b) and (c). 12-16 — elektor december 1982 selektor 'Superimposing the distortions caused by flexural vibration, produced by applying an oscillatory voltage to transducers at Ai and A 2 , shows that nodes appear at points marked C. Other transducers at Bi and B 2 serve as sensors to pick up the movement and convert it back into an alternating electrical signal. electrical signal; conversely, it produces an oscillating current when made to vibrate physically. Moreover, it can be made very thin and small. In the combined diagram, the three parts of the previous diagram are superimpo- sed, with piezoelectric crystals mounted as shown at Aj and A 2 and at right angles at Bi and B 2 . The oscillatory circuit is connected to Ai and A 2 while Bi and B 2 are used as sensors to pick up the mechanical movement and convert it into electrical alternations. There are also piezoelectric crystals quipositioned between the As and Bs, and it can be seen that when the cylinder is stationary there is a node at each position marked C. That is to say, there is no movement of the cylinder wall at those places, so there is no electrical output from the C crystals. The symmetrical way the crystals are arranged helps maintain the overall symmetry. In practice there are various electronic circuits connected to several of the crystals, but we need not go into the detail. Some of the circuits maintain the frequency, others effect corrections, and so on. Distorting force Now let us suppose that the cylinder rotates about its central longitudinal axis. Immediately, the Coriolis force comes into play. The effect is to add a tangential force to the force along the diameter of the cylinder, that is, the force direction for the vibration. As a result there is distortion of the move- ment and the points marked C are no longer nodes; instead, there is some movement outwards and inwards. This is sensed by the transducers at C and can be measured electronically. The amplitude of the electrical signal so produced is directly proportional to the rate of rotation. When the signal is rectified, and thereby changed into a direct-current signal, its strength can be shown on a meter which, when suitably calibrated, gives a direct reading of the rotation rate. This sounds simple enough, but a great deal has been left out of the explanation in order to keep it readily understand- able. For example, it is not easy to find the exact point where there is a node, so two transducers may be needed, with subsequent electronic treatment. To overcome the difficulty, a better method of measurement has been used. A signal that is 90 degrees out of phase with the input signal at A] is added to the output from C. Fairly simple mathematics then shows that the signal to be used is shifted in phase and that the shift is directly proportinal to the rate of rotation. With this technique there is greater sensitivity. Such a device must be inside a cylin- drical cover for protection. The first, made for investigation, had an overall length of 85 mm and a diameter of 50 mm. It was made of brass and had Sensor circuit arrangement for variable-phase output. Summing amplifier Miniature version of the rate sensor, measuring 25 mm x 17 mm. a fundamental vibration frequency of about 3 kFIz. Many experiments were done with it to discover the next practi- cal steps necessary in the research and development. For example, would tem- perature affect the performance? What adhesive would be best both for adhesion and electrical conductivity to the metal? How much would the distance between active cylinder and the surrounding cover affect the accuracy? Many such questions affecting the accuracy, stab- ility, sensitivity, ease of manufacture, and so on were raised for continued investigation. And what about minia- turization? A much smaller one was made and it worked satisfactorily. It was that version which was shown recently in London. So the correctness of the basic as- sumptions has been proved, and the company has been careful about world- wide patents. However, there is still a lot to be done and it will be a few years before there will be a product suitable for marketing. The device is indestructible and un- affected by shock or vibration, and it would be inexpensive to produce. For the most delicate applications it would not rival the conventional gyro, but it would still be sensitive and accurate enough for many uses. For example, once a rate signal is obtained it can be integrated to find the total angle of turn. Through modern electrics and electronics, the signal could be used for directing a vehicle automatically. Trac- tor machines could be made to steer according to a preprogrammed plan, and fork-lift trucks could be guided about a works complex quite auto- matically. This is the outlook for say, five years time. Spectrum, no. 180/1982 (833 S) triopede elektor decamber 1982 — 12-17 triopede Tripodus Electrus Diclopus (TED) The built-in electronic instinct is pro- grammed to seek the brightest point in the room. A deviation routine in the logic circuit ensures that it cleverly avoids the many obstacles on its path towards the light. For specialists it is clearly a simple cybernetic model. But to other, un- biased observers it is a spectacular electronic monster. For domestic pets it is a strange encounter of the third kind. For electronic hobbyists it is a lot of fun. When a technical creation becomes self-controlled it usually appears mon- strous to us humans. Inanimate objects are supposed to be 'dead' and indepen- dent action is associated with living beings. Exceptions to this rule are spooky. A locomotive which becomes self-controlled is a ghost locomotive and J. Cornelissen People will be astonished to see this three-wheeled creature moving along the floor, buzzing and chirping. Not only does it make strange but it appears determined to reach some destination. First it travels forwards, bumps into an obstacle, makes an elegant reverse U-turn then continues in a determined fashion towards some invisible destination. After careful observation of its behaviour, the only conclusion that can be drawn is that the creature is a light- addict! Hence the name Triopodus Electrus Diclopus, or TED for short. In normal terms it means: a light seeking, three wheeled, electro- mechanical hybrid. when inanimate objects move without being instructed to do so it is eery. This is what we are dealing with here. It is small, a bit weird but a lot of fun. Action and reaction As in the case of a living being, TED re- quires an internal locomotor system. It is driven by two electric motors that draw their energy from a rechargeable battery. The locomotor system is com- pleted by three wheels. Another charac- teristic of living beings is the ability to react in a specific manner to external stimulus. This results in a behavioural pattern. Our TED is equipped with simple sensory organs: two light-sen- sitive cells consisting of light-dependent resistors provide the electronic creature with directional, light-seeking behaviour. Simple contacts allow it to detect obstacles. 12-18 — elektor decamber 1982 triopede The two eyes (hence diclopus) are fitted at the front at an adequate distance from each other; they contain LDRs. If the light impinges laterally, then one of the two LDRs is subjected to less light than the other. The result is that the motor's electronic control circuitry executes a change of course which lasts until the light impinges from the front and both LDRs are subjected to the same amount of light. If there is only one light source in a room with dark walls, our mon- ster will always travel towards this light source. The direction it faced when it was placed on the floor is immaterial (figure 1). There is, however, one excep- tion: when travelling away from the light source, no direct light impinges on the LDRs. Both are dark and TED will travel aimlessly into the darkness. The situation is somewhat more com- plex in daylight conditions. On an infinitely large area subjected to sun- shine, our stubborn electronic creature would travel east in the morning, south at noon and west in the evening, as long as the battery still delivered energy. However, these conditions do not prevail in a room; several windows provide several light sources and the creature must select one of them. The direction of travel depends on the initial direction when it was switched on. There are often strange results which cannot be foreseen in the develop- ment stage: when faced with a 'difficult decision' the entire system of the crea- ture can start oscillating; this is made noticeable by the creature swinging to and fro on the spot as though it can- not make up its mind. This phenomenon is occasionally encountered in more intelligent beings! The problem was largely solved, however, by an appro- priate modification to the circuitry. Since most windows do not extend down to the floor, the cybernetic creature often changes direction when approaching a window: when it enters the shade area it may suddenly detect a light object that was previously unde- tected on account of the bright light from the window. Instead of con- tinuing its journey towards the wall, the creature changes direction and heads for a light waste-paper basket. During this turn, it can also happen that the light from another window appears in the view of TED and becomes the new destination. The creature is even capable of avoiding obstacles. An object placed in its path, such as a carboard box or waste-paper basket or even a human, is avoided if — and this is important — the object is in front of the light source and appears darker than the background to the LDR eyes (figure 2). If the obstacle has a light colour and is illumi- nated from the side towards which the creature is travelling, it is immediate- ly attacked. Even the light-coloured trousers of Elektor staff members were not spared. On the other hand, the legs of chairs and tables are not detected. The crea- ture can 'see' about as much as a human who tries to walk towards a light win- dow with his eyes closed. Two sensor contacts are interconnected by means of a bracket which forms a kind of 'front bumper'; this informs the creature when it has bumped into an undetected obstacle. When the bumper touches an object the creature's 'vision' is briefly disabled and it executes an 'instinctive action'. It reverses for a short distance (to obtain sufficient space for a change in direction) then makes a turn to the left or right, depending on the location of the obstacle. Luckily, our electronic creature is not capable of intelligent thought. Its creators wisely refrained from providing it with a computer brain and merely made use of simple and easily constructed, hard-wired logic circuitry. The circuit or simple nervous system From the electronic viewpoint TED is a fairly simple creature: it has simple sen- sors, simple logic circuitry and simple drives. Let us consider the sensors first. The most important sensory organs are the two LDRs which form a potential divider (R1, R2). When subjected to the same amount of light the two LDRs have the same resistance; half the oper- ating voltage is then applied to the mid- point. Two triggers without hysteresis ensure that any deviation from this midpoint results in a binary signal for 'right/left'. The two LEDs D3/D4 indicate the initial state and are in- dispensable for alignment, as described later. The following memory stage IC2 prevents direct feedback from the drive (direction control) to the direction in- dicators (LDRs). If the LDRs were to control the drive motors directly, hunting or oscillating could easily occur. In order to avoid such nervous over- reactions the creature was given a 'tranquilliser' in the form of IC2. This ensures that the results of its sensory perception are passed on with a slight delay, whenever clock generator A3 emits a pulse. TED blinks, so to speak, at the rhythm of the clock gener- ator. LED D5 also flashes with each clock pulse, to make the blinking noticeable. Direction control and drive are inter- related: this is an integrated function of the type used by track-laying ve- hicles. Each of the two drive wheels has its own motor and each motor has its own control system. When the directions of rotation are in opposition the TED turns on its own axis. When the directions of rotation are the same it 1 Figure 1 . If there is only one light source in a room with dark walls, TED will always travel towards this light source. The direction it faced when placed on the floor is immaterial. triopede elektor december 1982 — 12-19 travels forwards or backwards, depend- ing on the polarity of the motor supply. A bridge circuit of four transistors for each motor caters for polarity inversion. Driving of the bridge output stages is simply performed by inverting CMOS buffers (N17...N20). Since the motors are controlled digitally and not proportionally, the power loss is very low. As much of the battery energy as possible is made available for the drive system. So that TED will not 'give up the ghost' in front of every obstacle, the circuitry contains a deviation logic circuit. A front bumper is mounted on two sen- sors that require only slight pressure for actuation. When an obstacle is bumped into, these sensors deliver a logic '0' to the inputs of the flip-flop consisting of the two NAND gates N1 and N2, and to the inputs of gate N3. N3 represents an OR-f unction for the sensor signals: as soon as a sensor is actuated by a collision, N3 delivers a pulse to the two monostables N13/N14 and N15/N16. The duration of the toggled state of N15/N16 is about twice as long as that of N 1 3/N 1 4. The shorter time governs the duration of reverse travel and the longer time governs the total duration of the avoiding action. The whole operation consists of reverse travel of about one length of the crea- ture, followed by a 90° turn. In the event of a reaction to light (no collision) pin 9 of N9 and pin 5 of N 10 are at logic '1'. The information from the latch is allowed to pass. At the following OR-gates N7 and N8, pin 13 (N7) and pin 1 (N8) are at logic 'O'. This means that only the 'visual' infor- mation is allowed to pass here also (logic signals at the outputs of N9 and N10). When an obstacle is bumped into, both monostable outputs (pin 4 of N14 and pin 10 of N16) are first at logic '1'. This causes pin 3 of N4 to go to logic 'O'; N9 and N 10 are inhibited. The outputs of N7 and N8 are now at logic '1' until the monostable consisting of N13 and N14 reverts to logic '0' at its output: TED travels backwards! Where do the two 'ones' come from? The outputs of N5 and N6 are at logic '1' because the interconnected inputs (pins 9 and 5) receive a logic 'T from the monostable. Nil and N12 allow these logic 'ones' to pass, because their interconnected inputs receive a logic '1' from the second monostable. If monostable 1 toggles back to zero, the state of the flip-flop must be trans- ferred to the motors: it determines whether a turn to the left or right is made. If the output of the first mono- stable goes to zero, N5 and N6 deliver the flip-flop information via Nil, N12, N7 and N8 to the motors. N1 1 and N12 are only inhibited when the second ('longer') monostable reverts to zero somewhat later. N9 and N10 are en- abled again, thus restoring the original state. 2 82179 2 Figure 2. Since most windows do not extend down to the floor, the cybernetic creature often changes direction when travelling towards a window: when it penetrates the shade area, light objects are suddenly detected which were not noticed in the bright light of the window. When placed in its path, an object such as a cardboard box, waste-paper basket or even a human is avoided if — and this is important — it is between the light source and the creature and is therefore considered as background when viewed by the LDR 'eyes'. Figure 3. The circuit of the cybernetic creature, our TED evaluates the light impinging on R1 and R2; T1 . . . T8 provide the motor control circuitry; SI and S2 evaluate 'contact with the enemy' and the sound generator with loudspeaker gives the creature its speech. 12-20 - elektor december 1982 triopede 4 Figure 4. Track pattern and component layout of the printed circuit board for the Triopede. Ensure that all ten wire links are inserted and that the cooling surfaces of all nine power transistors are facing R9/R8. Construction The electronic circuitry should present no problem and can simply be as- sembled on the printed circuit board shown in figure 4. Care must be taken to insert all ten wire links and to ensure that the cooling surfaces of all nine power transistors are facing R9/R8. When all components have been fitted to the board and inspected, we can turn to the mechanical construction of TED. Figure 5 shows the solution as tested in the Elektor laboratory. Of course, many versions are possible. A base supports the printed circuit board and the re- chargeable battery. The battery voltage is 6 V. Although the CMOS ICs and LEDs consume relatively little power, the motors draw approximately 250 to 300 mA under load (rolling resistance, friction of the gearbox and drive belt). Normal dry batteries would be in- adequate; for this reason a rechargeable battery with a minimum capacity of 0.6 Ah was chosen. It powers both the motors and the electronic control cir- cuitry. Any interference generated by the motors has no effect on the con- trol circuitry. The two LDRs which form the 'eyes' are mounted externally on the left and right. The loudspeaker and battery switch S3 must also be connected and mounted somewhere on the base. Switches SI and S2 are assigned par- ticular locations as are the LDRs. They are mounted on a bracket which serves as the 'bumper'. If the creature bumps into an obstacle, one of the two sensors closes and activates the electronic cir- cuitry. LEDs D3, D4 and D5 are first mounted on the printed circuit board. Later, when TED has assumed its final form, they can be mounted elsewhere to improve the appearance. The base is propelled by two motors. We used two 6 V/350 mA motors. The worm drive was fitted with a gear and pulley that supports the drive belt at the motor end; the belt is a rubber band of the type used for sealing preserving jars. Situated at the wheel end is a gearbox which steps down the rotary motion transferred by the belt. Mounted at the input of the gearbox is a pulley of the same size as that at the motor. Thus the motor speed is transferred to the gearbox at a ratio of 1:1. The drive wheel, which rotates approximately 4 . . . 16-times more slowly than the belt pulley, is positioned at the output of the gearbox. With this design the creature reaches a speed of approximately 10 cm per second, which corresponds to 0.36 km an hour. The LDRs are darkened at the rear using an indelible marking pen or black adhesive tape. No shielding against lateral light is required. Alignment PI. P2 : These are used to establish the 'blind zone' in the visual range of TED. A change of direction of a light source within this blind zone must not cause any change in course. If the range of tolerance becomes too small or even zero, 'motor hunting' can occur, as already mentioned. If it is too great, obstacles can no longer be detected in time. A room is chosen with a light source in the form of a single spot if possible (small window or light bulb). Place TED on the floor at a distance of a few metres from the light source and facing it with both 'eyes'. Then rotate it about 10 degrees to the left. More light now impinges on LDR1: the volt- age at point A becomes greater than Parts list Resistors: R1,R2 = LDR R3,R4,R5 = 470 il R6.R15= 22k R7 = 1 k R8,R9,R17,R18 = 10 k R10 = 2k2 R11.R12 = 47 k R13 = 470 k R 14 = 1 M R 16 = 33 k P 1 ,P2,P3 = 1 0 k trimmer Capacitors: C1,C2,C3 = 4p7/16 V C4 = 470 p/16 V C5,C6 = 47 n C7 = 100 n Semiconductors: D1,D2= 1N4148 D3,D4,D5= LED T1,T3,T5,T7 = BD239C T2.T4,T6,T8,T9 = BD 240C IC1 = LM 324 IC2 = 4042 IC3 = 4011 IC4 = 4071 IC5 = 4081 IC6 = 4001 IC7 = 4049 IC8 = 4093 Miscellaneous: LS = loudspeaker 8 ill 0.2 W S1,S2 = sensors (pushbuttons) S3 = on/off switch 6 V rechargeable battery, 0.6 Ah min. U/2. Adjust PI so that LED 1 just begins to light up. Now rotate TED the same number of degrees to the right. Adjust P2 so that LED 2 just begins to light up. During forward travel (with the light source dead ahead), neither of the two LEDs must light up. LED 1 and LED 2 must never light up simultaneously. This can, however, result from an in- correct adjustment of PI and P2. Fine adjustment of the clock frequency can only be made while the creature is travelling. Its speed is very important. The clock frequency can only be ad- justed to suit the mechanical construc- tion of the propulsion system after a few trial runs. It is also interesting to observe the change in behaviour of the creature at different clock frequencies (approximately 2-10 Hz). When it is switched on TED acts as though it had just bumped into an obstacle. Wait until the reverse travel and subsequent rotation have ended. For a functional check it is advisable to disconnect the link between the motors and the drive wheels. Connecting the motors The motors must be connected with the correct polarity so that TED will not unintentionally travel backwards or in a circle. The procedure is as follows: 1. Illuminate the LDRs in such a way that D3 and D4 do not light up. The two motors should now rotate for for- ward travel if they were properly con- nected. 2. Rotate TED so that the light source is at the front right (as viewed by TED). The right drive wheel should now rotate backwards and the left wheel forwards. 3. Rotate TED so that the light source is at the front left. The motors should now change their direction of rotation. The left wheel should run backwards and the right wheel forwards. 4. If, after bumping against an obstacle and reverse travel, the creature turns towards the obstacle instead of away from it, SI and S2 should be inter- changed. Finally . . . ... all that is left is the final appearance of TED and this is a matter for individ- ual appreciation. However, we would advise against a terrifying exterior; the 'cybernetic look' is probably quite sufficient. Have fun. M 12-22 — elektor december 1982 precision power supply Good control with high power If any circuit is to be accurately and safely tested a good power supply must be used. It is not sufficient for it to be just a stabilised supply, it must also in- clude some form of protection against faults arising in the circuit under test. This usually takes the form of current limiting and output short circuit protec- tion. In order for it to fulfil its function cor- rectly, a power supply should have the following facilities. • The ability to deliver fairly high cur- rent levels at voltages of 24 V or more. • It must be completely stable at all output conditions. • The output must have some form of short circuit protection. • Current limiting control up to the maximum current output. • An output voltage control that is fully variable from 0 to maximum. • Accurate indication of both current and voltage output levels. • Sense inputs to allow compensation for voltage drops when long supply cables are necessary. Although the last two points are not strictly necessary, their inclusion makes the power supply more versatile and easier to use. The precision power supply here fol- lows the standards set by commercial equipment and includes all of the above features. It has a variable output voltage range of 0 to 35 V and continuously variable current limiting up to 3 amps. The performance is on a par with fairly expensive commercial power supplies but approaches the stabilisation prob- lems with a rather novel circuit design. The principles The vast majority of power supplies use either 'series' or 'pass' regulation. This means that the stabilising power transis- tors are connected (effectively) in series or in parallel to the load. In common with most designs the circuit here utilises series pass regulation. The originality in the circuit design is the method used for stabilisation. The block diagram in figure la illustrates the principle of a conventional series regulator. The active element of the cir- precision power supply Any item of test equipment is useful but only one is absolutely necessary and that is some form of power supply. These normally provide a voltage output of up to 25 or 30 volts at about 1 amp which is fine for most purposes. However, this current level can be rather limiting when testing computers, audio amplifiers and other high power equipment. It is essential too that some form of protection such as current limiting is included in the circuit design. The precision power supply here is capable of providing up to 3 amps at 35 V and incorporates both current limiting and short circuit protection. Meters are included to enable current and voltage output levels to be monitored. precision power supply elektor december 1982 — 12-23 cuit is opamp A and its output is the source of the load current, that is, in series with the load R|_. The non-invert- ing input of the opamp is held at a refer- ence voltage, U re f. The inverting input of the opamp is at a voltage level that is a proportion of the input voltage — de- rived by potentiometer P. Under these conditions the output of the opamp will become stable at the point where the voltage difference between the two in- puts is zero. That is, the opamp will maintain a condition where the reference voltage and that at the wiper of poten- tiometer P are equal. It will be obvious that the output voltage will therefore be dependant on the position of P. With the potentiometer in mid position the output will be double the reference volt- age. The disadvantages of this system are that the stability factor is dependant on the setting of potentiometer P, the output can never be lower than the refer- ence voltage and the operation of P will not be linear. Two of these points may not be so significant in some cases but an output minimum that is restricted to the reference voltage will be embarrassing to say the least! The block diagram of figure 1b provides another solution. In this case, the opamp is used as a unity gain amplifier and P becomes a voltage divider connected across the reference voltage. The output of the opamp will now be proportional to the voltage level at the wiper of P. In this configuration the output range will be between 0 and the reference voltage. This sounds better but it is still far from ideal. The opamp will now require a negative voltage supply rail, an added dis- advan- tage. The refer- ence volt- age must be at least as high as the maximum required output, not an ideal situ- ation! Finally, the stability factor is still a question of poten- tiometer P. Figure 1c goes a long way towards re- moving the problems by replacing the reference voltage, as far as the opamp is concerned, with a reference current. The output voltage is now determined by the current passing through P. The advantage is that the circuit is no longer dependant on the reference volt- age level. We now arrive at figure Id which, in principle, is very similar to 1c. The ref- erence current in this case is derived from the output voltage via a series resistor R. The idea is not entirely new but the method used here is a little unorthodox. As previously mentioned, a current source is achieved by placing a resistor in series with a reference voltage derived from the output. Flowever, for this to happen in practice, the value of poten- tiometer P has to be much lower than R. The opamp still tries to balance out the difference between the voltage levels at its inputs but now the output voltage will be equal to the level on its non- inverting input. The series resistor is effectively placed between the two inputs of the opamp. However, due to the high impedance of the inputs, theoretically at least, no current can enter the opamp. In effect then, the current derived from the refer- ence source follows the path shown as a dotted line in the block diagram. Since Ui = U 2 (the opamp ensures this) the current level remains constant, totally independant of P and the load. The U re f current level is equal to — — . The opamp will balance out the voltage across P and, in doing so, the reference current is compensated for any change in load. The result of all this is that the circuit conforms to what we are looking for, a constant reference current (even at 0 V) using a reference voltage source and a resistor. The precision power supply The major difference between the block diagram of the precision power supply in figure 2 and that of figure Id is the fact that two opamps and a series pass power transistor are included. The cur- rent source (U re f and R) and the poten- tiometer PI are very similar. The second opamp A2 is responsible for output current limitating. The voltage across the emitter resistor R s of transis- tor T is proportional to the output load current. A proportion of the reference voltage is derived by the setting of P2 and this is compared to the voltage across R s by opamp A2. When the volt- age across R s becomes higher than that set by P2, the opamp reduces the base drive current to T until the difference is reduced to zero. The LED at the out- put of A2 functions as a current limiter. The circuit diagram So much for the theory, now for its practical application. The circuit of the power supply, shown in figure 3, has two independant power supplies (if that makes sensei). The power for the out- put stage is provided by transformer Tr2 which, of necessity, will be rather a hefty beast. Transformer Trl provides power for the reference source and the opamps. The reference source is derived with the aid of the inevitable 723 (the worlds longest living chip?). The components 82178 Id Figure 1. The drawings here, in conjunction with the text, illustrate the advantages of why the use of a constant current reference source is preferable to a reference voltage. 12-24 — elektor december 1982 precision power supply around this 1C were chosen to provide a reference voltage of 7.15 V. This appears at the junction of R1/R5, R15/R16 and R9. For ease of understanding it should be noted that R4/R5 represents R and IC2 corresponds to A1 in the theoretical diagram of figure 2. The reference voltage eventually arrives at the non-inverting input of IC2 (pin 3) while the inverting input is connected to the zero rail via R8. Diodes D2 and D3 are included to protect the inputs of the opamp against surge voltages. The output of IC2 controls the power out- put stage, consisting of transistors T3, T4 and T5, by providing the base drive current for transistor T2. A word about transistors T3 . . . T5. These are connected in parallel and their outputs are combined via emitter re- sistors to provide the power supply out- put via R21. This resistor is the practical counterpart of R s in figure 2. The use of three 2N 3055's in this configuration provide an economical power stage that can handle up to 3 amps comfortably. The voltage across R21 is compared in IC3 with a voltage level determined by the setting of P2. This latter is derived from the reference source via R15/R16. The output of IC3, like that of IC2, is fed (via D5) to the base of T2. When the output current is higher than that set by P2, the output current is reduced by IC3 until the two levels are matched. Transistor T1 and its surrounding com- ponents cause the LED D7 to light when current limitation is in effect. Two meters are included to allow direct monitoring of both voltage and current levels at the output. Each meter is pro- vided with a series potentiometer, P3 and P4, to allow for fine calibration. These can be replaced with fixed re- sistors if desired once their values have been found. Capacitor C3 in the reference voltage circuit (IC1) serves two functions. It reduces any noise produced by the inter- nal zener of the 723 and it also provides a 'slow start' for the reference voltage supply. This means that when the power supply is first switched on, the opamps are giving time to 'settle down' before being asked to do any work, a sort of early coffee break! If this slow start was not designed in it could possibly allow the maximum voltage level to appear at the output, albeit very briefly, but still potentially damaging. The diodes D1 to D8 in various parts of the circuit are included to guard against the possibility of accidental connection of an external voltage to the output ter- minals of the power supply when it is switched off. For instance, this could quite easily occur when working with a circuit that has a built in battery back-up. Components R7 and C6 increase the reaction time of the circuit when changing output voltage levels while capacitors C7 and C8 eliminate the possi- bility of oscillation in the opamps. For stable operation of the circuit a minimum Figure 2. The basic block diagram of the precision power supply. Opamp A1 provides the voltage regulation while A2 takes care of the current limiting. 4 + U S + U 1 1 Powersupply 1 1 RL -u 1 1 S' -Us 82178 4 Figure 4. The two sense inputs are used in the manner illustrated here to enable the circuit to compensate for voltage drops caused by the use of long cables. B80C 50 00/3300 Figure 3. The circuit diagram of the precision power supply. Resistors R4/R5 correspond to R in figure 2, IC2 to A1, IC3 to A2 and R21 to R s . Of the two transformers, Trl provides the supply for the reference current source while Tr2 supplies the power for the output stage. 12-26 — elektor decamber 1982 precision power supply 82178 precision power supply elektor december 1982 — 12-27 Parts list Resistors: R1,R3,R6,R8,R12,R13,R14 = 4k7 R2 = 22 to R4.R16 = see text R5 = 10 k R7.R10 - 1 k R9 = 2k2 R 1 1 =470 ill 1 W R15 = 15k R17 = 10U/1 W R18.R19.R20 = 0,22 a/3 W R22 = 4k7/1 W R23.R24 = 47 a R25= 5k6 R26 - 270 k PI = 50 k potentiometer P2 = 1 k potentiometer P3 = 2k5 preset P4 = 250 k preset Capacitors: C1,C2 = 100 p/25 V C3= 100 p/10 V C4 = 100 p C5 = 10 p/25 V C6 = 1 n C7 = 100 p C8 - 56 p C9 = 47 p/63 V CIO = 4700 p/63 V C11 = 820 n C12 = 100 n Semiconductors: B1 = bridge rectifier B40C1000 B2 = bridge rectifier B80C5000/3300 D 1 ,D8 = 1N4001 02 . . . D5 = 1N4148 D6 = 3V3 400 mW zener D7 = LED red T1 = 8C559C T2 = BD 241 T3,T4,T5 = 2N3055 IC1 = 723 IC2,IC3 = 741 Miscellaneous: SI = double pole mains switch Ml ,M2 = 100 pA meter Trl = 2x12 V/400 mA mains transformer Tr2 = 33 V/4 A mains transformer F = 1 A fuse output load resistance is necessary. This is taken care of by R22. It will be noted that there appear to be more output terminals than the usual power supply needs. The two extra out- puts, +U S and — U s , are in fact inputs. These so-called 'sense' inputs are used to allow for voltage drop compensation when working with long connecting cables between the power supply and its load. Figure 4 illustrates how the inputs are used. Two extra wires are connected as shown between the load and the sense inputs. The result of this is that the supply voltage level is now effectively measured at the load and not at the out- put terminals of the power supply. This enables the circuit to compensate for any voltage drop resulting from the resistance in the main supply cables. It should be noted that if the total resist- ance of the two main supply cables is 1 at the current level of 1 A the voltage drop will be 1 V. In normal use. 6 Figure 7. The design of the front panel that is available from Elektor. It is manufactured from scratch resistant polycarbonate material and is self-adhesive. The illustration is at a reduced scale, the actual size is 11 cm by 30 cm. Figure 6. The practical wiring diagram for the power supply. Obviously care must be taken with connections, especially with respect to the transformers and power transistors. Errors in this area will not become visible until the smoke clearsl shorting links can be placed between +U and +U S , and — U and — U s . Construction The maximum output current of the circuits as shown here is 3 A at 35 V but In principle different current ratings are possible. It must be remembered that any change in this direction must be accompanied by a change in the ratings of both C9 and CIO. The limiting factor is the maximum collector/emitter volt- age capability of transistors T2 . . . T5. This is 60 V for the 2N 3055. The other deciding factor will of course be the current rating of the transformer for the power output stage. The maximum out- put of the power supply is a factor — — of the current supplied by the trans- V2 former which explains why a 4 A trans- former is required to achieve an output of 3 A. The three power transistors in parallel are used because each 2N 3055 cannot dissipate more than 50 W. The consider- ation is that when the output voltage is at 0 V the maximum dissipation required is the maximum level of the rectified voltage multiplied by the maximum current. For an output of 1 A at 35 V only one 2N 3055 would be sufficient. One more power transistor can be added without any modification to the circuit providing that the correct value for the emitter resistor is calculated. A 2°C/W heatsink is needed for each power tran- sistor or a 1 °C/W for each pair. Capacitor Cl 2 is mounted directly onto the out- put terminals as shown in figure 6. Do not mount the resistors R4 and R16 initially as their value will depend on the maximum output voltage and cur- rent. For this reason it will not be poss- ible to mount the printed circuit board into the case until test and calibration is completed. Set PI to maximum, switch on and connect a multimeter to the out- put of the circuit. By trial and error find the actual value of R4 which gives the maximum required output voltage. This can be done by connecting different resistors in parallel to R5. When the correct value has been found it can be soldered in place on the board. Repeat the exercise with P2 and R16 (in parallel with R15) until the maximum current level is found. The remaining calibration is that of the meters by adjustment of P3 and P4. It is possible to build the power supply using only one meter. In this case a 2 pole 2 way switch connected to points x, y and z is required to switch between volts and amps. H the Elektor XL range elektor december 1 982 — 1 2-29 die Elektor XL range an audio system that sounds good — and looks good! Over the years, Elektor has built up a reputation for producing reliable and well-designed circuits for home con- struction. In the audio field, we have produced preamps that vary from the cheap-and-n . . . not-too-bad to the top quality all-singing-all-dancing variety; power amps from the milliwatt range up to 'quality-power'; noise reduction systems, indicators, distortion measuring gear, a stylus pressure gauge even. Now we've got some more in the pipe-line; the XL range. In this case, we feel that a general intro- duction is called for. What is the XL range? What is special about it? In a nut- shell, it comprises a series of very high performance audio components that can be built at sensible prices. The quality should come as a pleasant surprise, even to the serious audiophile, and the final appearance needn't be a disgrace to the living room. In other words, it is not intended to be competitive in the low- budget category; not it is expressly intended to be a competitor in the £1000+ range. High quality, plenty of features if you want them, reliability and a reasonable price: these were the main design goals. Our new front panels add the finishing touch. How it all started Way back in 78, one of our many hi-fi enthusiasts drafted an eight-page memo on what we ought to do in this field. It suffered the fate of most memos: interested study, brief discussion, filing cabinet . . . However, the idea kept on nagging in the back of our minds. And when an idea is irritating the grey matter in some two dozen fertile, tech- nically oriented brains, something must happen eventually. Then, early this year, your editor was wondering what large project would be worth tackling in the winter months, as a series of articles. From past experience, a truly consecu- tive series seemed unwise: no matter how carefully it is planned, the designers always get second thoughts halfway through — and want to redesign some forthcoming part. It then becomes a battle royal to keep the articles appear- ing at monthly intervals. (For those readers who have wondered why nearly all magazines tend to leave gaps in a series 'for reasons of space': now you know the true reason!) The two things came together. A series of articles describing a complete audio system! You can publish the various components in any order — and if you miss a month, nobody's any the wiser. Now (second thoughts again!) we de- cided to play it straight: we'll tell you what we intend to publish, but we don't make any promises as to when! You can take your pick from one of the stock phrases: in the near future, soon, over the next few months, starting now . . . Over the past six months, we’ve been having a lot of fun: no-wild-ideas-barred brainstorming! What features would you like? What features seem even vaguely feasible? To give you some idea of the things we've been thinking of, we can run through the various items in the system. No promises, mind! These are 'brainstorm' ideas, and not all of them are economically and/or technically feasible. Power amplifier: Crescendo 100W? 250 W? Bandwidth from DC to passing-bat frequencies, or restricted to suit the human ear? Transistors, opamps or MOSFETs? End of discussion: the final circuit is described elsewhere in this issue. As specified in that article, it uses MOSFETs to deliver plenty of undistorted power — enough to coax pleasing sounds at a more than adequate level from even the most inefficient of 8 £2 speakers During our think-tank sessions, we dreamed up some features. DC protec- tion for the speakers seemed a 'must' — so we'll describe that next month (promise!). A power-up countdown with visual indication is fun, and avoids nasty thumps — so it is included. Then there's a choice: where do you put the amplifiers? In the rack, on the floor, or in the speaker cabinets? The design caters for all options — but if you mount it in a rack, you need something to put on the front panel. The mains switch, of course; power output indi- cators, possibly? Hey . . . why not add a thermometer on the heatsinks! When the system is switched off, it will double as room temperature indicator. Preamp: Prelude What should a preamp do? It boosts the input signal, selects the one you want, provides tone, volume and balance con- trols, feeds signals to a tape deck and headphones, and may do a lot more things as well: noise reduction, remote control, level indication . . . Wow! A de- sign that includes everything might be rather expensive, and one that does not include everything might disillusion some audiophile. So what do you do? Make it 'modular'! That way, all the various bits and pieces can be added or omitted according to personal taste. The design that we're finalising at present includes: • both dynamic and moving-coil pre- amps that can be tailored to suit virtually and high-quality cartridge • tone controls that include all mod cons • for the true audiophile: tone control cancel option • separate headphone amplifier • input sensitiviry adjustment • dual input channel switching: one for listening and the other for re- cording on tape • remote control option, as outlined below Remote control: Baton Your editor would have liked a system that included motor-drive of the control potentiometers. No quality loss, and just think of the 'effect': hit a button on the remote control unit, and the corresponding knob on the panel starts to rotate! However, this would involve a difficult and expensive mechanical system, so we settled on an electronic (infra-red) version. For top-quality- philes: the unit can be omitted or, if it is included, cancelled. Loudspeakers: When did we last publish a loudspeaker design? Way back! Why haven't we published one since? Because loud- speaker design is a specialised art — it's much easier to design a bad one than to create something really good! — and because manufacturers have a provided several good designs for the home con- structor. Even so, we felt that the XL range ought to include one or more designs for this all-important component. At present, we have several interesting and promising possibilities in the pipe-line. But we haven't seen or heard any of them yet, so we are very reluctant to make any firm promises. Suffice it to say: if and when we publish a design, rest assured that it comes from a repu- table specialist in this field! FM tuner We haven’t even thought of a 'musical' name yet! What we have in mind should be revolutionary; in fact, 'in-house' it has been dubbed the FM-2000. High performance, easy to operate, micro- processor-controlled . . . Wait and see! XL Why did we call this the Elektor XL range? What does XL mean? An honest answer: we don't know, but it sounds nice! We picked those letters six months ago, because they have a nice high- quality 'sound', and we've been racking our brains ever since for a good expla- nation. 'Excellent'? 'Extra Luxury'? 'Exclusive'? Forget it! It's simply a series of audio gear that we'll be describing from now until . . . then. 12-30 — elektor december 1982 crescendo idea to try to avoid the ailment and reduce the dose of medicine. Let us return to the analogy of the passport photograph. Looking at the block diagram of figure 1 or the circuit diagram of figure 2, if we draw an imaginary line along the middle we can see that each component has its coun- terpart. Components which are only present once are assigned to the com- mon input or output (including feed- back). There is one exception: we only need one trimmer potentiometer (PI) to set the quiescent current. The input circuitry in figure 1 begins with a highpass filter consisting of R1, Cl and C2. This filter is necessary for two reasons: firstly, to prevent very low frequencies from reaching the bass loadspeaker, and secondly to block any DC that may be present at the input. Any DC component would appear at the output and unbalance the quiescent current setting of the output stage. The lowpass filter con- sisting of R2 and C3 was designed to pass signals with frequencies of up to approximately 160 kHz. The double differential amplifier con- sisting of T1 . . . T4 is represented with two operational amplifiers in figure 1. The output signals of the two differential amplifiers (voltage over R11 and R13) result from the comparison between the input signals and the output signal attenuated by R4, R5 and R6. In other words, the feedback first module v in the audio XL range Power MOSFETs are available as P-channel and N-channel types. This makes it possible to design a class B amplifier in which both types operate with 'time sharing' — to borrow a term from the computer field. Since there are limits to the power rating of such a MOSFET, we have used what could be described as double 'job sharing' in this design: an output stage with two N-channel and two P-channel MOSFETs. The result is an output stage capable of great output power and exhibiting extremely low distortion. The principle If one takes a passport photograph and folds it downwards in the middle, over the nose and between the eyes, the two halves will certainly not be identical. The same applies to many amplifier designs. In our case, of course, we are less interested in the visual distortion than in the harmonic and non-harmonic distortion as a result of non-symmetrical amplifier design. These types of distortion can be avoided by proper, symmetrical design of the circuit. The reason is as follows: the distortion of even harmonics produced on each side of the symmetry line finally cancel each other out (as AC voltage into the load impedance). The result is less 'cosmetic circuitry' in the form of feedback or other alternatives, in order to meet the high requirements for quality of the output signal. Since the medicine (feedback) is usually worse than the malady (lack of stability, dynamic distortion (TIM), it is a good crescendo elektor december 1982 — 12-31 Figure 1. 'A line along the centre’: the block diagram of the power amplifier shows the symmetrical arrangement. effect is contained in the voltages over R 1 1 and R13. For AC voltages the amplification is 1 + (R5/R6I/R4 = 32. For DC voltages the amplification is 1. This is catered for by C4 and C5. Without these two capacitors, the offset of the double differential ampli- fier would also be amplified by a factor of 32. This offset consists of the different base-emitter voltages at the given collector currents and of any difference that may be present between the voltage drops over R1 + R2 and over R5/R6 (resulting from the base currents of T1/T3 and T2/T4 that may not be equal). However, this last contribution to the offset voltage can be ignored if R1 and R2 in series are equal in value to R5 and R6 in parallel. This explains the apparently unnecessary parallel connection of 39 K and 150K instead of one 33 K resistor. Ideally, the current flowing from the base of T3/T4 will be as great as that flowing into the base of T1/T2. In this case the voltage drops are zero. Incidentally, this article also explains how the low output offset of +20 mV can be reduced even further. The current source 'supply' (T5, T6) turns the input stages into good differ- ential amplifiers and poor analogue adders: amplification and non-linearity (excluding the feedback effect) for common signals are low. The influence of slow and rapid variations in the operating voltages (100 Hz hum plus sinusoidal halfcycles at the signal frequency) on the desired signal is zero squared. One more comment on the differential amplifiers: R12 and R14 ensure that the collector-emitter volt- ages of T1 and of T2 and those of T3 and T4 are almost identical. In this way, thermal balance between the two differential amplifiers is achieved, re- sulting in a favourable effect on the offset. The differential amplifiers supply the drive voltages (via R11 and R13) for two current sources: the sources of the drive current for the output stage consisting of T11...T14. Each of these current sources consists of a cascade circuit of two transistors: T7 and T8 at the top and T9 and T10 at the bottom. This apparently unnecess- ary duplication of components provides many advantages. Each cascade forms a 'super transistor' with a current amplifi- cation factor of at least 400 and with a straight and almost flat lc (UCE) characteristic extending up to collector voltages of 250 V, and with an equally linear, frequency and voltage-indepen- dent collector-base capacitance of a few tenths of a picofarad, which can be made as small as the p.c.b. layout will permit. What we have is an ideal, text- book current source for frequencies from DC to 0.5 ... 1 MHz. How is this achieved? There are two contributing factors, the first is 'job sharing'. This high current amplification is provided by T7 and T9, operating at low level (a few volts), whilst T8 and T10 handle the high voltage and dissipation. The current amplification of T8 and T1 0 can be ignored, because the currents flowing in their emitters are almost 100% equal to those flowing in their collectors. The second factor is screening. Two parallel, metallic surfaces present a capacitance. Now what if this capacitance is un- desired? The simple solution is to take a third metal surface and place it between the other two. This third surface eliminates the original capaci- tance and creates two new, considerably lower and less harmful capacitances. In figure 2 the base of T7 and (T9) forms one electrode, the collector of T8 (T10) forms the other electrode and the base of T8 (T10) forms the intermediate electrode. Why is this screening so important? The 'eliminated' collector-base capacitance in this amplifier is present anyway (a few picofarads) and it is non-linear. One can imagine it as a varicap diode with a capacitance which is greatly depen- dent on the inverse voltage. In many amplifier designs this non-linearity is cured by connecting a capacitance in parallel which is 50 to 100-times greater. This Miller capacitance, as it is known, also serves as a stabilizing capacitor. The solution is not optimum, however. It is better to avoid capaci- tance altogether, no matter how linear it may be! The collector-base capacitances of T8 and T 1 0, which are special video transis- tors, are very low. This capacitance (2 x approximately 2 pF) is effectively between collector and ground and cannot cause any mischief. However, this cascade solution has its price: the dynamic range of the output stage is restricted as a result of the necessary DC voltage setting. But this does not present a problem because it is better for the top end of the dynamic range to be restricted by the cascade circuit than by the saturation phenomenon in the output stage ifself. This in turn has a favourable effect on recovery after an overload. The DC voltage setting of the cascades and of the differential amplifiers is performed by zener diodes D1 and D2, which are connected to the operating voltage via R17 and R18. The result of this 'joint operation' of the zener diodes with C8 ... Cl 1 , R 1 9 and R20 is that the DC setting of T1...T4 does not change in the slightest (in the event of operating voltage variations). Let us now consider the output stage. It can handle a peak current of 14 A and a dissipation of 320 W at 50°C. Short-duration current limiting is per- formed by D3/D5 and D4/D6. In the event of long-term overcurrents, fuses FI and F2 ensure the necessary shut- down. The output stage is adjusted to a quiescent current of 2-times 100 mA. This current is more than adequate to provide sufficient 'overlap' (simul- taneous conducting) of the two output- stage halves to keep any distortion (that might arise from insufficient or even no quiescent current) at a very low level. With drain currents upwards of about 100 mA, this current decreases as the temperature rises at a constant gate-source voltage. This negative tem- perature performance of the MOSFETs ensures that the output stage cannot overheat. With conventional NPN/PNP output stages, certain measures must be taken to prevent this hazardous situation. No such measures are re- quired with our design. It is merely necessary to set the quiescent current with PI. No additional diodes and transistors are needed. The MOSFET output stage exhibits a considerably flatter output current/in- put voltage characteristic than that of a conventional output stage. This results in advantages and disadvantages. Let us first consider one significant disadvan- tage. The output stage is configured as a complementary source-follower. This means that maximum drive is deter- mined by the operating voltage less the drive voltage. Since the drive voltage for MOSFETs must be higher than normal with the same AC output voltage, the result is a reduced dynamic range at the given operating voltages (a further restriction is created by the voltage drop over the relatively high saturation resistance of a MOSFET). Now let us look at the advantages. By its very nature, the flatter voltage-in/ current-out characteristic provides less opportunity for static and dynamic distortion. Furthermore, a particular quiescent current can be adjusted with PI. There is no jump from 0 to about 1 A when the quiescent-current trimmer is rotated by 10°. The MOSFET output stage T11...T14 can handle high frequencies as well as high output power. The relationship between out- put current (drain current) and input voltage (between gate and source), i.e. the slope, remains unaffected by the frequency right up into the megahertz region. However, this can also result in a tendency to oscillate. The risk of crescendo elektor december 1982 — 12-33 Table 1. Technical data General: Output power Input sensitivity Input impedance Frequency range Attenuation factor Output offset voltage Extras • Fully symmetrical/complementary hi-fi MOSFET output stage • Good dynamic charcteristics because the internal 'inertia' of the input is utilized for full frequency compensation (designed for unconditional stability). • 140 W into 8 fi when both channels are simultaneously driven at a distortion factor which does not exceed 0.01% (—80 dB) within the frequency range 20 Hz ... 20 kHz. Total power output: 280 W. • 1 80 W into 4 f l when both channels are simultaneously driven at a distortion factor which does not exceed 0.01% {— 80 dB) within the frequency range 20 Hz ... 20 kHz. Total power output: 360 W. • 180 W max. per channel into 8 (fully driven) • 250 W max. per channel into 4 H (fully driven) 1 Vrms for 1 30 W into 8 25 kfi 4 Hz ... 1 60 kHz + 0/— 3 dB (with a source impedance of 600 J2). 100 Less than ± 20 mV DC • DC output protective circuit, combined with activation delay • Temperature sensor for heatsink • VU meters 3 optional Figure 3. This is the hefty power supply for the amplifier. oscillation is largely avoided by keeping the wiring as short as possible, de- coupling capacitors (C6, C7, C14 and Cl 5), limiting resistors (R23...R26) and, with somewhat less influence, (R27 , . . R30. The two capacitors Cm and Cp in the block diagram of figure 1 represent the input capacitances of the MOSFETs. Over the operating range of the ampli- fier, the voltage over C[\| and Cp is independent of the frequency and relatively proportional to the output current. The frequency-dependence is based on the drive current delivered by the driver stages T7...T10, which rises with the signal frequency and/or the output current. However, this drive current is none other than the charging/ discharging current of Cm and Cp. The result of having a current that rises with frequency is that the voltage over R15 (R16) also increases with fre- quency. The same applies to the voltage over R 11 (R13). The advantage of this arrangement soon becomes clear. As in all amplifiers with feedback, the so- called open-loop gain (= gain without feedback) must drop with increasing fre- quency, as of a particular point. The drop must be such that the frequency rolloff at which the open-loop gain is one is not quite 1 2 dB per octave, i.e. just before the point of causing a 180° phase shift. This is not an invention of Elektor, but was established by two gentlemen called Bode and Nyquist some time ago. This not-quite-1 2 dB, not-quite-1 80° situation characterizes the stability minimum. If one wishes to remain in the safe region of 6dB per octave and 90° phase shift, in order to allow capacitive loads or capacitive contents of loads (crossover networks), the open- loop gain must roll off with 6 dB per octave from a 'safe' frequency onwards (which is now considerably higher). In all cases this must be up to the frequency at which the amplification has become one, but beyond that point if possible. The maximum open-loop phase shift in this case is 90 in the quiescent state; the system is uncon- ditionally stable. It is almost always necessary to create these conditions by affecting the open-loop gain, i.e. by providing compensation. In most cases this is accomplished by inserting a compensation capacitor. With our circuit, however, this is un- necessary. If the open-loop gain de- creases with 6 dB per octave from a safe frequency onwards, there is a point in the amplifier at which the drive voltage or current increases with 6 dB per octave at a given drive level. We have already found this 'point': the drive current through C(\| and Cp, or the drive voltage over R11 and R13 which amounts to the same thing. To put it another way: the frequency- dependent input characteristic of the MOSFET output stage T11...T14, which is present in any case, is used for crescendo elektor december 1982 — 12-35 Parts list Resistors: R1 = 27 k R2 = 3k9 R3 = 10 n R4 = 1 k R5= 150 k R6 = 39 k R7. . .RIO = 150 R11 . . . R14 = 6k8 R1 5.R16 = 82 n R1 7,R1 8 = 10k/1 W R19,R20 = 2k2 R21.R22 = 5k6 R23 . . . R26 = 220 £7 R27 . . . R30 = 0,22 !J/5W R31 ■ 1 0/1 W carbon film R32 = 10 0/1 W carbon film PI = 250 O preset Capacitors: C1,C2 = 820 n MPF' C3 = 220 p ker. C4.C5 = 220 p/10 V C6.C7 = 220 n MPF' C8 . . . Cl 1 = 100 m/10 V Cl 2,C1 3 = 330 n MPF' Cl 4, Cl 5 = 1 00 m/ 1 00 V C16 = 22 n MPF' *MPF = metallized plastic foil Semiconductors: D1 ,D2 = zener diode 3V9/0.4 W 5% D3.D4 = zener diode 12V/0,4W5% D5.D6 - 1N4148 T1,T2,T6= BC546A T3 . . . T5 = BC 556A T7 - BC 560C T8 = BF 470 T9 - BC 550C T10= BF 469 T11.T12- 2SK135 (Hitachi) T13.T14 = 2SJ50 (Hitachi) Miscellaneous: LI = approx. 2 mH, 2x 10 turns Copper enamel (dia. 1mm) on R31 FI ,F2= 3.5 A fuses, slow-blow, with p.c.b. fuse holders Two heatsinks for T8agd T10 (e.g. SK09/ 37.5 mm; approx. 8.5 C/W) One heatsink for Til . . . T14 (e.g. SK53/ >1 50 mm, black anodised, not predrilled (see figure) Aluminium angle section 40 mm x 40 mm, 1 80 mm long (see figure and figure ) Mounting and isolating hardware for Til ... T14 (see figure I Thermolube Power supply (without printed circuit board): B1 ,B2 = bridge rectifiers 100 V/25 A (see figure ) Cl 7 . . . C20 = 4700 . . . 10000 m/ 80 ... 100 V or Cl 7, Cl 9= 10000 m/80 ... 100 V; C18.C20 can then be ignored. Preferably can electrolytics with screw terminals. F3.F4 = 2 or 2.5 A slow-blow fuses Trl ,Tr2 » mains transformers 2 x 25 V/ 6 A (300 VA) or 2 x 50 V/5A (500 V) toroidal R 32 = 10 fl/1 W carbon film PI » 250 n trimmer potentiometer Figure 4. Track pattern and component overlay for the power amplifier. The four MOSFETs are also situated on the p.c.b. This is not the simplest solution but certainly the most reliable one. purposes of frequency compensation. We can be greatful for the very high, overall collector impedance of the cascade T7...T10 which makes this possible. Why is this advantageous? The 'inertia' of C[\| and Cp is present in any case, whether it is used for compensation or not. The alternative would be complete or additional compensation elsewhere in the amplifier. This would mean that a Miller capacitance would have to be added to the circuit of Figure 2. It would provide additional 'inertia' but would also give rise to TIM problems. Let us assume that such a capacitor were present between the base of T7 (T9) and the collector of T8 (T10). The maximum available charging/ discharging current is approximately 300 /zA (collector current of T1 or T3). Clearly there will be an audible frequency at which a 'clean' sinewave at the input will result in an equally 'clean'triangular waveform at the output . . . Fortunately, that is not the case with our design. Those readers who are interested in the slew rate can rest assured that the ampli- fier does not easily become unstable, in other words it would take a great capaci- tive load at very high frequencies. With 180 W/10 kHz into 4 Si the peak value of driver AC is 0.6 mA, although 14 mA are available. In principle the figure is even higher, but the driver stage then changes to class AB. At higher fre- quencies the AC increases proportionally and reaches a value of 14 mA at a par- ticular frequency. However, the input filter consisting of R2/C3then becomes effective. Let us briefly consider some of the com- ponents which serve to stabilize the cir- cuit. Firstly, R32 and Cl 6 which are rated so that R32 does not go up in a puff of smoke when the amplifier is tested at full output with 100 or 200 kHz. Bacause of the high self -induct- ance, R32 must not be a wire-sound resistor. LI and R31 in parallel partially or fully compensate for the phase shift caused by a capacitive load at the ampli- fier output. R31 serves for damping the resultant LC circuit so that square-wave reproduction is faithful in this case too. That completes the description of the output stage. The amplifier requires a power supply to power it from the mains. The power supply is shown in figure 3 and is rated for stereo operation. One transformer and one bridge rectifier are required for each operating voltage. Two channels are thus provided with a common positive and negative rail. However, there is no risk of crosstalk between the two channels, and certainly not via the power supply. As already mentioned, interference to the operating voltages does not impair operation of the amplifier. It is preferable to use toroidal transformers in the power supply. The constructor can choose between a medium-rating power supply (600 VA) and a high-rating power supply (2-times 500 VA). The choise depends on the amound budgeted and on whether the amplifier is going to be 'pushed' into a load of 4 Si. The DC smoothing also requires a little thought. The minimum is 4700 pF and the maxi- mum is 10000 pF per operating voltage and per channel. The maximum values for Cl 7 . . . C20 are not only governed by the cost, but are also a matter for technical consideration. The 'bigger' the capacitors, the lower is the 100 Hz hum and the greater are the charging current peaks. The technical data in table 1 apply to the medium-rating power supply with minimum smoothing. When the amplifier is switched on, the output voltage will initially contain a DC component. This is because the feedback is not immediately effective. Also, electrolytic capacitors C8 and C9 require a certain time before they are charged up to the zener voltage of D1 (D2). There is one more situation in which the DC voltage can be present at the ampli- fier output: this is in the event of an overload on the output stage, if fuses FI and F2 do not blow or do not blow simultaneously. This voltage and the 'starting voltage' already mentioned can be hazardous to our expensive loud- speakers. For this reason it is a good idea to equip this amplifier with a DC protection and activation-delay circuit, which is also suitable for other ampli- fiers. For reasons of space, we will discuss such a circuit in the next issue of the magazine. Construction When wiring an amplifier to its power supply it is easy to make mistakes that can impair the audio performance. We therefore recommend that the following text and the corresponding illustrations be followed closely. One difficulty has already been removed: figure 4 shows a printed ciruit board layout. The cost of this p.c.b. is negli- gible compared to that of the power supply and the MOSFETs. For this reason we strongly recommend that only this p.c.b. be utilized. Even very slight interaction between the output and inputs of the cascades on the one hand and the 'wiring' of the MOSFETs of the p.c.b. on the other hand are critical. The method applied to position the MOSFETs directly on the p.c.b. is not the simplest way, but has proven to be the most reliable one. Although a few holes have to be drilled in a heatsink for the transistors, this is far preferable to positioning the MOSFETs on a heatsink away from the p.c.b.; the resultant wiring would very probably cause the output stage to oscillate. Our arrange- ment may be unusual, but it is much more reliable. Those familiar with Murphy's Law will know that 'Amplifiers always oscillate, oscillators often fail to oscillate'. 12-36 — elektor december 1982 crescendo We shall tackle the mounting of com- The heatsinks for T8 and T10 (see bolted together and the p.c.b. can be ponents on the p.c.b. first. All com- figure 6b) are positioned vertically, used as a drilling template. Otherwise ponents are inserted except for T1 1 ... The heatsink for T10 must not make the drilling diagram in figure 6c should T14 and R23 . . . R26. R27 . . . R30 contact with the lead of C7 which is be used. Some precision is required here; must have a little clearance from the very close to it. It is advisable to try T1 1 . . . T14 must be installed in a fully p.c.b. when soldered in. This provides this out visually before drilling a isolated fashion. A poor fit can cause the resistors with some cooling. The mounting hole for T10 in its heatsink. short-circuits between the metal parts, enamel must be carefully removed We must not forget the wire link (input Finally, the aluminium bracket is from the ends of coil LI so that a good ground). On the component side, the mounted on the p.c.b. together with the electrical connection is made with R31 pins of T1 . . . T7 and T9 should not be MOSFETs as shown in figure 6c. It is when soldering. As soon as PI has been longer than 4 mm. We now turn to the better to apply too much rather than inserted, it should be rotated fully anti- positioning of T1 1 . . . T14.T23 . . . R26 too little thermolube to each side of the clockwise to avoid any problems later. and the heatsink. First the aluminium mica insulation. When the assembly is If the amplifier is to be trimmed to bracket is cut to the width of the p.c.b. finished, use another ohmmeter to minimum DC output offset (more about and then the holes are drilled. This can ensure that there is no contact between this later), it is better to use transistor be done in two ways. When the two the TO-3 and the aluminium, sockets for T1 . . . T4, and insert the mounting holes have been drilled, the Mechanical assembly is then followed transistors later. aluminium bracket and the p.c.b. can be by electrical installation of transistors elektor december 1982 — 12-37 crescendo Til ...T14. The gate and drain ter- minals are simply connected to the appropriate tracks on the soldering side. A solder lug is required for each source terminal. Til and T12 are the furthest from the right-hand edge of the p.c.b. and T13 and T14 are the closest to it (as in figure 4). Resistors R23 . . . R26 are drawn with dashed lines in the component overlay of figure 4. They are positioned as closely as possible to the gate terminals of the MOSFETs on the soldering side. There are no mounting holes. The leads of the resistors are cut as short as poss- ible (1 cm maximum), then bent and soldered to the copper surfaces, leaving a clearance between the resistors and the p.c.b. The p.c.b. with components fitted is now connected to the large heatsink (see figure 6b). The manner in which the two units are assembled largely depends on the housing. Since the heat dissipated by the MOSFETs is trans- ferred to the heatsink via the aluminium bracket, extremely good contact is required between the two. At least six nuts and bolts (M4x 15, for example) and a generous amount of thermolube between the two metal surfaces will ensure good thermal contact. If a stereo amplifier is desired, the entire procedure must be repeated. The housing and power supply are the largest parts of a power amplifier. Figure 5 shows a suggestion for the complete stereo power amplifier. The additional circuitry mentioned is not yet included. However, the wiring will not be greatly changed by adding these extra circuits. The wiring is fairly critical. Performance of the amplifier can be greatly impaired by accidental metallic contact or induc- tive or capacitive pick-up. A likely source of interference is pick-up from the operating voltage leads at the input. If the amplifier is driven with a strong sinusoidal signal, the two supply currents look like half-wave rectification with many harmonics. The higher the har- monics, the greater the likelihood that they will be picked up at the input. 12-38 — elektor december 1982 crescendo Accidental metallic contact is prevented by only having one ground point (at the electrolytic capacitors). The amplifier is provided with an input ground and an output ground which are interconnected internally via R3 and connected exter- nally to the main ground point. The rating of the power supply depends on various factors: the transformers chosen for Trl and Tr2, the smoothing capacitors, the required power output and, of course, the constructor's budget. Some care should be taken in choosing the smoothing capacitors. It is better to use a 4700 pF can electrolytic with screw terminals than a 10000 pF elec- trolytic from the junk box. Moreover, an advantage of the can electrolytic is that none of the terminals is connected to the metal can. During construction, alignment and any measurement, care must be taken to ensure that the negative operating voltage never makes contact with ground. Figure 5 shows a possible power supply arrangement using can electrolytics. Vehicle-type spade ter- minals can be utilized. The stranded wire must have the greatest possible crosssection. Activation The power supply should be tested before its load is connected. The wiring should be inspected first. An incorrectly wired electrolytic capacitor can actually explode! When the mains switch is turned on, the positive and negative operating voltages should be 70 . . . 75 V. The electrolytics are then dis- charged again. But do not use a screw- driver; the best method is to employ a resistor of a few kilohms, or a relatively low-impendance voltmeter. In this case one can see the voltage gradually decaying. The entire amplifier can now be tested. If the stereo version has been built, the instructions in the following text must be followed twice. First, position all ground leads as shown in the figures and as described. The positive and negative leads from the power supply are then connected to the output stage. Use two 10 £2, %W resistors temporarily instead of fuses FI and F2. Quiescent-current trimmer PI is rotated fully anti-clockwise. The loudspeaker terminals remain discon- nected. Now switch on the power supply. Should there be a short-circuit somewhere in the output stage, the 10 £2 resistor will go up in smoke. In this case, switch off the amplifier, find the fault and correct it and insert two new resistors. When everything is in order, connect a multimeter over one of the resistors (3 V or 6 V DC range, see also figure 2, FI and F2). The voltage over this resistor should be zero. Otherwise PI is not at its extreme anti-clockwise setting. The voltage should rise as PI is slowly rotated. At the setpoint voltage of 2 V, the current through this resitor is 200 mA, i.e. 100 mA per MOSFET. When this adjustment has been made, switch off the amplifier and insert the two fuses FI and F2. Switch on again and measure the voltage between the amplifier output and ground. It should not exceed 20 mV (positive or negative). The amplifier is now ready for oper- ation. The test points marked in figure 2 can now be checked. One final voltage at the amplifier output can be reduced further. This is done as follows: let us assume that the values for T5 and T6 are established. Two dif- ferent types can be utilized for T1 and T3. There is thus a choise of four possibilities for T1 . . . T4. One of the possibilities results in the lowest DC output voltage. However, if the DC output voltage is only 20 mV, there is probably little point in trying to reduce it further, although we suspect that perfectionists may not agree with us. Something not mentioned up till now concerns the case for the power ampli- fier. The illustrations show a typical 19 inch rack type case, which is ideal. Some of you may be tempted to use steel plates for the sides top and bottom as these are robust and will therefore protect your amplifier extremely well. Steel reacts with the field of the trans- formers to produce quite a loud hum, so we recommend the use of aluminium for these items. If the amplifier is for P.A. applications we strongly recommend using a larger case for extra cooling and the shielding (by means of a small aluminium plate) of the inputs from the power supply. One good way to help matters is to place the inputs as far away as possible from the power supply. M Literature: Toroidal Transformers Elektor April 1982 Equivalent circuits MOSFETs as source-followers When MOSFETs are utilized in output stages, the P and N-layers and junctions are of less interest than the performance of a combination of resistors, capacitors and voltage or current sources. When considered as a whole, these provide an accurate equivalent circuit of the P and N-layers and junctions. Figure a shows the equivalent circuit of a MOSFET configured as a source- follower. Resistance rg represents the internal series resistance of the gate, which is externally 'extended' by resistors R23...R26 in figure 2. Capacitance C represents the input capacitance of the source-follower. The drive voltage u is applied over this capacitor; when multiplied by the slope S, it determines the AC output current through load R(_. In the fre- quency range of interest, S (frequently denoted Yf s ) is independent of the frquency. If the source-follower is voltage driven, the formula of figure a applies. The gate current contribution i to the output current can be ignored. Figure b represents an equivalent circuit for figure a. In this case, however, the input impedance is demonstrated more clearly. The formula applies purely to current driving with a current i. In contrast to the voltage drive assumed in figure a, resistance rg no longer has any influence. However, the maximum level to which the device can be driven is limited as a result of the voltage drop over rg and by the base resistor. The ideal current drive of figure b is not found in practice. For this reason, figure c shows an equivalent circuit which is more realistic. The joint, design-related effect of the two cascades T7 . . . T10 of figure 2 is contained in the current source i s and resistance R. R represents the joint col lector resistance of T8 and T10. The formula exhibits a certain frequency-dependence which leads to the automatic, inherent fre- quency compensation for unconditional stability as mentioned in the text. The necessary rolloff of 6 dB per octave is found in the relationship between the output voltage and the sum of the cascade input voltages. A given open-loop gain is associated with a particular cutoff frequency W3dB- The higher the one value, the lower is the other. The fact that the frequency compensation of the amplifier can take place entirely in the combi- nation of driver stage and output stage (i.e. no other capacitors are required, with all their disadvantages), results from the use of the cascade circuitry. The circuit provides a high resistance R and ensures that no series capacitance is connected between output and inputs. Incidentally, if true current drive were utilized as assumed in figure b, the total effect of the driver stage and output stage would be an integrating one. We would then have a cutoff frequency crescendo elektor december 1982 — 12-39 a b c <*>3dB • — ID C 1 »SR L 1 «SRi RC Uj (TJ) uj 1T9> ' S " R15 " R16 821 BO - c of 0 Hz. Of course, this would not be satisfactory. Suppression of distortion and interference, effective as a result of the feedback, must be frequency- independent over the largest possible frequency range. So far we have only dealt with one source-follower (fugures a...c). In reality, however, there are four such devices (Til . . . T14 in figure 2.) Hence the equivalent circuit of figure d. At low drive, all four MOSFETs conduct; at high-level drive, either Til and T12 or T13 and T14 conduct. In figured the time constant 1/W3dB is established by the product of R and parallel connection (sum) of the four capacitors. It should be noted that if two MOSFETs turn off and the other two conduct, there are two small capacitances (i.e. the 'conducting' capacitances for which S 4 0) and two large capacitances pertaining to the MOSFETs which are turned off. Capacitance Cm is not exactly equal to Cp. This means that the output stage is not fully symmetrical. However, there is no point in inserting the differ- ential capacitance somewhere between the gate and the source, particularly since resistance r^ cannot be removed from the MOSFtf chip. Capacitances Cn and Cp exhibit a strange response. At low values of gate-source voltage Uqs. Cn and Cp greatly depend on this voltage. If their response curves were drawn as a function of Uqs. the result would be a deep notch in the flat line: this notch would be in the region of Uqs = 0, whilst the characteristic would be almost flat at high positive and negative values of UqS- The depth of the notch can also be assigned a value. For Cn it is approximately 270 pF and for cp approximately 160 pF 'in depth'. As a result of the quiescent current setting, these two characteristics shift in such a way that the notch becomes more of a 'ditch' which is almost horizontal at the bottom. This means that as of a particular drive level, the driver stage must not only supply or draw the 'normal' current i = C x dU/dt but also the current U x dC/dt. This performance can be observed using an oscilloscope connected over R 15 or R 16 (the oscilloscope chassis must be floating), with a high drive level and by adjusting the quiescent-current trimmer PI. The practical, negative results for linearity at high-level drive and high signal frequencies can be ignored. However, we should point out that the problem can be solved by using three MOSFETs for each half of the output stage. The changes in Uq$ (i.e. the six AC voltages u over C in figure a) are then only effective at the bottom of the wide 'ditch'. M 82180 -d 12-40 — elektor december 1982 used (see figure la) may differ depending on the type of tele- phone in use. If your system is already operating, you will already be aware of the wiring connections referred to. However, if in doubt, the easiest method of checking is to undo the handset microphone and speaker covers and make a note of the colour coding used. This can then be tracked back down into the main body. Figure la may be of some assistance here. In order to prevent the possibility of feedback the connections to the speaker should be exchanged (reversed phase) as shown in figure 1b. The power supply for the mini line amp is derived from the microphone line as a dc voltage exists at this point when the hand- set is lifted. The emitter of T1 is connected to the negative line while resistor R2 is connected to the positive. Care must be taken to ensure that connections here are correct. Readers have advised us of prob- lems encountered with the use of old surplus type telephones. Basi- cally the pulses generated by the dialling mechanism (in certain cases) are incorrect and the exten- sion dialled fails to respond (no ring out). In the normal course of events, when the hand -set is lifted and the dial is turned the speech line is shorted to earth by means of an internal switch within the dialling mechanism, so in effect you cannot hear what is going on. By disconnecting the appropriate wire from the mechanism (modern telephones it is blue) should solve the problem. You will then hear through the speaker the dialling pulses and eventually the exten- sion dialled ringing. home Speak up, you're through! This short article is aimed at those readers who have found a use for the home telephone published in the September issue of Elektor. Some readers have expressed a desire to improve the volume level of the speaker in the handset. There is one, very quick and easy answer to this problem. This simply consists of replacing R6 on the printed circuit board of each extension with a wire link. This will serve to increase the output level but, in this case, simple is not quite the best! Unfortunately, the quality of the speech signal can suffer badly. Our good friends at British Telecom are able to use line amplifiers to retain speech quality (we have this on good authority!). This answer is not available to us unfortunately, since the cost for the complete system would probably treble. We do have an answer though! The modification we have arrived at entails the addition of a single transistor amplifier stage ( a sort of mini line amplifier). One of these is needed in each extension. The circuit is very small, physi- cally, it can be assembled on a piece of Vero board measuring only 2x2 cm. Figure 1b illustrates how the mini line amp can be fitted into the handset of the extension. Obvi- ously the tag numbering system 82193-1 10p/35V| 82193-2 Soft switching elektor december 1982 — 12-41 Soft switching improvement to the High Com based on information from the Telefunken Television and Radio Company, Audio Development In case the title of this article could be misinterpreted, the purpose of this circuit is not to replace the noisy toggle switches that are typically used in home-made equipment. This modification providessoft switching by changing the response time of the HighCom circuit to increase the decay time, thus eliminating interference. With this modification the HighCom circuit becomes even better than it was and Telefunken will be offering the im- proved version as an integral part of the circuitry in future. The HighCom circuit was further refined in the Telefunken development laboratory. The result is an improvement in the form of the auxili- ary circuit in figure 1. The track pattern and component layout are shown in figure 2, but we shall discuss this later. These are the improvements: 1. The distortion factor at frequencies of about 20 Hz is reduced by two thirds. The compander distortion thus becomes noticeably less than band dis- 1 Figure 1. The additional soft switching circuit eliminates interference caused by drop-outs. Additionally, the distortion factor for low-frequency signals and the response time at high frequencies are improved by a slight modification to the HighCom module. 2 Parts list Figure 2. Track pattern and component over- lay for the additional board. It is placed on the HighCom 1C in the module in piggy -back fashion. When choosing components, please observe the specified tolerances. T2 and T3 should have the same specifications if possible. Resistors: R1 9 = 1 k 2% R21 = 220 k R22 = 220 k 2% R23.R24 = 68 k 2% Capacitors: C24 = 470 n 5% C25 = 220 n 5% Semiconductors: T1 = BC 557 T2,T3 = BC 557B fortiori in the audible range. To quote actual figures: the distortion content of the first harmonic of a sinusoidal 20 Hz signal is reduced by about 6% and that of the second harmonic by 3.5%. The total distortion factor with the new version is only 3.31%, compared to 10.38% with the old version. The im- provement was achieved by increasing the decay time which is effective during the holding time. 2. Interference caused by very short pulses, such as electrostatic discharges when playing records, is now eliminated by modifying the response time of the circuit. The network for this function now contains two time constants in- stead of one. This greatly improves the compander's capability and has been demonstrated by the better response time on a sinusoidal 1 0 kHz burst. 3. Short drop-outs result in undesirable, audible fluctuations in level. The HighCom circuit sometimes responded to these drop-outs with uncontrolled regulating effects. For example, the loudspeakers sometimes exhibited inde- pendent audio levels. On the other hand, the type of noise effects in signal pauses that are familiar with other noise suppression systems could not be heard. Telefunken achieves this performance by controlling the decay with a long time constant during the holding time, followed by a short time constant. How- ever, this automatic changeover between the two time constants was so abrupt that it sometimes resulted in the effects mentioned. Our soft switching circuit eliminates these problems, even with critical signals. There is no doubt that this is a considerable improvement to the HighCom circuit. The circuit in figure 1 can be con- structed on a small vero board or a printed circuit board can be made as shown in figure 2. Terminals 1, 4, 5, 6 and 23 correspond to the pins of the HighCom 1C, U401. Solder pins should be inserted at those points. The High- Com module is modified as follows: R6 and C21 are discarded. The value of C7 is changed to 47 n/5% (if this has not already been done). If the module already contains capacitor C7 with this value, the 470 n and 1 k connected in series between pin 1 and pin 6 on the copper track side should be removed. Wire links of about 2 cm in length are now carefully soldered to pins 1, 4, 5, 6 and 23 of the 1C. The additional board is then placed on the 1C in the HighCom module and the wire links are connected to the solder pins on the printed circuit board. This completes the modification to the HighCom module and the module can be replaced on the printed circuit board. The result is probably one of the best noise suppression system available. M Literature: Elektor HighCom, Elektor, March 1981 12-42 — elektor december 1982 comulative index 1982 <*iimulaliv(‘ || |4iO index tiKxZ Audio A dozen and one sounds 12-59 Automatic switch for output amplifiers 7-39 Class AB amplifier 7-50 Crescendo 1 2-30 CX and DNR 241 DNR printed circuit board 3-28 High com monitor extension 3-14 High quality tape playback pre-amplifier 746 Low octave switch 742 Miniature amplifier 7-28 Mixing console 7-31 Sound effects generator 7-78 Safety switch for stereo equipment 1-39 Softswitching 1241 Stereo power amplifier 7-28 Super low noise pre-amplifier 7-72 Switched filter 1C MF 10 (applikator) 942 Test tone generator 5-14 The Elektor Artist 5-32 (missing link) 9-53 The simplest PDM amplifier 749 Universal VCF 7-65 Voltage controlled filter 7-84 100 W amplifier 4-18 Car and bicycle Car alarm 647 Car lock defroster 7-68 In car ioniser 1245 Lead acid battery charger 3-32 Optical speed indicator 140 Rear light monitor 7-86 Domestic Automatic outdoor light 7-25 AC motor control 7-54 Double alarm 7-36 Electronic chimes 1-14 Electronic dog whistle 6-30 Electronic starter for fluorescent lights 6-22 Electronic thermometer 7-61 Economical battery tester 7-36 Fluid level detector 7-62 Fluorescent light dimmer 642 Gas detector 9-16 Home telephone system 946 (missing link) 1045 Home Telephone system modifications 1240 Induction loop paging system 1-32 Inductive sensor 9-25 Infra-red remote control receiver 7-74 Infra-red remote control transmitter 743 Keyless lock 7-66 Kitchen timer 1 1-58 LCD thermometer 10-20 Light sensitive switch 7-14 Low cost temperature indicator 7-53 Open door reminder 1-38 Single channel infra-red remote control 1-52 Smoke detector 7-34 Strobe light control 246 Telephone bell 7-64 Teletext power supply 2-21 Time receiver for the Rugby MSF 9-54 TV sound interface 4-25 Water indicator 1-39 Wind sound generator 3-16 16 channels with only 5 ICs 1040 20° + indicator 141 Fun and games Blinky 7-35 Cerberus 11-38 Cubular bell 11-54 Executive decision maker 7-24 Magic running lights 7-76 Rapid loading games 9-20 Talking dice 11-24 Triopodus electrus diclopus 12-17 Generators A dozen and one sounds 1 2-59 Crystal oscillator 7-73 (missing link) 9-53 Digital logarithmic sweep generator 7-20 Frequency generator 7-20 Economical crystal timebase 7-38 Graphic oscillator 748 Inverter oscillator 7-21 Positive triangular waveform generator 7-33 Pulse generator 7-55 Sound effects generator 7-78 Square traingle VCO 747 Stable start/stop oscillator 7-77 Test tone generator 5-14 VCOTA 7-80 Voltage controlled waveform generator 7-75 Voltage controlled waveform generator 7-62 Wind sound generator 3-16 555 pulse generator 7-27 HF and CB A 'mid-fi' receiver 7-52 Active aerial 1046 Automatic squelch 3-26 Compact AM/FM receiver (applikator) 1-56 Compact shortwave SSB receiver 6-16 Converter for varicaps 741 Crystal oscillator 7-73 (missing link) 9-53 DSB demodulator 10-17 Economical crystal timebase 7-38 FET field strength meter 7-38 LED tuning indicator 7-69 Miniature MW receiver 5-24 comulative index 1982 elektor december 1982 — 12-43 "“"tt 1982 Mobile aerials 6-27 Pre-amp for the SSB receiver 1 0-44 RF amplifier for the 10 metre amateur band . . . 7-22 RTTV converter 7-70 Short wave converter 7-59 Simple AGC 7-18 Signal strength meter 7-21 Short wave band shifting for SSB receivers .... 10-36 The principles behind an SSB receiver 6-13 10 W/70 cm amplifier 2-16 150 MHz frequency counter 1-17 Hobby DC motor speed control 7-15 Economical battery tester 7-63 High speed NiCad charger 7-44 I nfra-red remote control transmitter 743 Infra-red remote control receiver 7-74 Lead acid battery charger 3-32 Model train lighting 11-28 Prop tachometer 5-39 Simulated track extender 2-22 Single channel infra-red remote control 1-52 Stop signal override 12-62 Universal NiCad charger 2-30 16 channels with only 5 ICs 1040 Informative articles Blue LEDs 3-36 Colour LCDs 1-58 CX and DNR 241 DSB demodulator 10-17 Electrolytics run dry 10-28 Guitar amplifier philosophy 1-42 Introducing DMOS power FETs 6-52 Measuring AC waveforms 6-24 The principles behind an SSB receiver 6-13 The Z8 family (applikator) 2-33 The 13600, a new OTA 4-50 Toroidal transformers 444 When is an OTA not an OTA 4-35 Miscellaneous A simple window comparator 7-60 AC/DC converter 7-88 Analogue monoflop 7-48 Automatic delay switch 7-36 Biomedical interface 7-80 CMOS switch Schmitt-trigger 7-64 Digital timer 740 Diode dimmer 1-38 Dissipation limiter 347 EX(N)OR opamp 7-51 LED tuning indicator 7-69 Monoflop with a CMOS gate 7-61 Omnivore LED 7-51 OTA Schmitt-trigger 7-29 Phase sequence indicator 7-89 Polystryrene cutter 7-16 Pushbutton interface 7-27 Reciprocal amplifier 7-34 Solid state relay 6-57 Temperature to frequency converter 7-19 The 13600, a new OTA 4-50 When is an OTA not an OTA 4-35 Microprocessors Basic on the Junior Computer 4-52 B iomedical interface 7-80 Calling Junior vectors 7-69 Capacitive keyboard 446 Darkroom computer part 1 9-30 Darkroom computer part 2 10-30 Dynamic RAM card 4-28 Dynamic RAM for SC/MP 7-37 The Elekterminal with a printer 7-91 EPROM eraser 3-50 Floppy disc interface for the Junior 1142 Floppy disc interface for the Junior part 2 .... 1248 F rom the 6502 to the 6809 6-50 High performance video mixer 7-86 High speed printer routine 7-89 Mini EPROM card 4-17 Mini EPROMmer 7-57 NIBL 1200GT 1-50 Program EPROM 742 RAM/EPROM card for the Z80 5-51 (missing link) 9-53 Rapid loading games 9-20 RS 232 interface 7-76 Serial keyboard interface 7-22 Single cycle mode for the Junior Computer .... 7-71 Software cruncher and puncher 5-54 Talking board interface 2-52 Talking clock 6-32 The Elektor connection 9-27 The Junior Computer as a frequency counter . . . 5-26 TRS 80 cassette interface rediscovered 7-30 Z8 family (applikator) 2-33 Z80A CPU card 5-28 2114 RAM tester 440 2716/2732 programmer 1-26 6502 housekeeper 542 (missing link) 6-62 Music Active attenuator 7-23 Adding the finishing touches to the New Elektor synthesiser 3-18 Combined VCA/VCF module 1-22 Dual ADSR and LFO/NOISE modules 2-24 Drum interface . 1 1-20 Guitar amplifier philosophy 142 Guitar tuner 11-33 Mini organ extension 1 1-56 Output unit and keysoft for the Polyformant . . . 7-92 12-44 — elektor december 1982 comulative index 1982 ""‘‘ffi 1982 (missing link) 9-53 Polyphonic synthesiser 3-51 Synthesised sound animation 9-52 The digital keyboard assembly and debounce circuitry for the Polyformant 5-17 The 'Poly-bus' 6-36 Tuning aid 4-24 Universal VCF 7-65 Voltage controlled filter 7-84 12 dB VCF 7-84 Photographic Darkroom computer Part 1 9-30 Darkroom computer Part 2 10-30 Darkroom thermostat 2-58 Slave flash 7-26 (missing link) 9-53 Slave flash trigger 7-19 Power supplies AC/DC converter 7-88 Dissipation limiter 347 Dissipation limiter 7-81 High voltage converter 7-18 Low voltage stabiliser 7-31 Mini high performance voltage regulator 740 Power failure protection 7-83 Precision power supply 12-22 Summer circuits power supply 7-17 Symmetrical opamp supply 7-60 5 V super power supply 7-58 Test and measuring equipment Active attenuator 7-23 AD/DA conversion 342 Connection tester 442 Connection tester 7-87 Economical crystal timebase 7-38 Frequency generator 7-20 Frequency multiplier 1-53 Duty cycle meter 7-53 Logic probe 745 Measuring AC waveforms 6-24 Oscilloscope aid 7-56 Overvoltage protection for meters 7-32 Precision power supply 1 2-22 Simple frequency converter 7-85 Signal strength meter 7-21 Three phase tester 9-60 Tolerance indicator 2-50 The Junior Computer as a frequency counter . . . 5-26 True RMS converter 7-28 Ultra sonic distance measurement 10-24 150 MHz frequency counter 1-17 Missing links Chopper front end for power supplies - E 75/76 1-59 Crystal oscillator - E87/88 9-53 Digital keyboard - E75/76 1-59 EPROM programmer - E81 341 Frequency doubler - E73 1-59 Home telephone system - E89 10-54 Output unit and keysoft for the Polyformant - E87/88 9-53 RAM/EPROM card for the Z80 - E85 9-53 Revolution counter - E77 2-61 Slave flash - E87/88 9-53 Talking board - E80 1-59 Talking board - E 80 2-61 Teletext decoder (2) - E79 1-59 Teletext decoder (3) - E80 9-53 The Elektor Artist - E85 9-53 6502 housekeeper - E85 6-62 Combining the Junior with the Elekterminal - Junior Computer Book 3 2-61 in-car ioniser elektor december 1982 — 12-45 A theory which has been with us for some time and which is rapidly gaining credence relates to the quantity of negative ions in the air. A high concen- tration of such ions is both physically and mentally healthy. One element of scientific thought actually states that the quantity of negative ions contained in the air around areas such as St. Moritz is high, which is one reason for the invigorating effect these resorts have on tourists. There certainly seems some truth in these suggestions as negative ion generators are gaining in popularity. Even institutions traditionally known for their ultra-conservative attitude towards new ideas are now using them. We published a circuit for a domestic ioniser a couple of years ago operated by the mains supply and the idea came to adapt this circuit for use in the car. The circuit design for a suitable power supply is shown in figure 1. It could be loosely termed as a d.c. to a.c. converter. The 555 timer ( 1C 1 ) produces 7.5 kV. The output is then connected to a sewing needle or something similar. As most readers already know the electric field strength around a charged body is greatest where the curvature is also greatest, hence a sharp point. An intense field is therefore present at the tip of the needle with electrons being 'sprayed' onto the air molecules nega- tively charging them. Each batch of negative ions is repelled by the negative charge of the needle point allowing new air molecules to be processed. The result is a constant flow of ions away from the needle which feels very much like a light draught. This in itself will have a refreshing effect upon the driver and passengers without giving consider- ation to the metabolic benefits of an increased concentration of negative ions. Keep in mind that apart from generating negative ions the needle will also pro- ioniser fresh air on wheels The circuit is one way of increasing the concentration of negative ions in the surrounding air, resulting in improved mental concentration, and reaction speed making roads just that little bit safer. At the very least it will refresh the environment. a square-wave signal with a frequency around 85 to 100 Hz. The values of R1 and the combination of PI and R2 have been chosen so that the square- wave produced is symmetrical. This is then fed to transistors T1, T2 and transformer TR1. The result is an a.c. voltage across the two secondary wind- ings of the transformer of approxi- mately 400 V (square wave). Figure 2 shows the circuit diagram of the ioniser which consists of a 27 stage voltage multiplier, in order to step up the voltage from 400 V to around duce ozone (0 3 ). This can on the one hand have certain advantages as it oxidises organic gasses. Carbon monox- ide for instance, can be reconstituted into carbon dioxide which is far less harmful. However, ozone if breathed in large quantities can cause irritation of the respiratory system, because of its corrosive and therefore poisonous nature. We therefore do not recommend using the ioniser near to asthma suf- ferers and please remember that for normal use the ventilation system of the car should be reasonably effective. 12-46 — elektor december 1982 l in-car ioniser Construction The printed circuit board for the power supply is shown in figure 3. There is nothing critical in the assembly and the only calibration needed is to set PI to its mid position. No provision was made for mounting the transformer onto the board as the size and type will depend on what is easily available. Although it is possible by changing Cl for a 330 n capacitor to get a 50 Hz a.c. output, we do not advise it. Basically the peak voltage level produced by the circuit using the specified transformer will be far in excess of 240 V, so that 'blowing up' your razor becomes a distinct possibility. The transistors will need small heat sinks. The transformer should have a 220 V primary and two 6 V secondary windings. Its normal function is reversed in this case. The printed circuit board and com- ponent layout for the ioniser are given in figure 4. Great care is needed to mount the components. Make sure all soldered joints are smooth and neat as any protruding wires or spikes of solder could result in unwanted discharges. This is especially important towards the 'high-voltage' end. Resistors R1 to RIO limit the current flow in the event of the needle being touched. Lowering the value of these or omitting them is unadvisable as it could result in a fatal shock. Any sharp needle will do as long as its connection to the printed circuit board is short and rigid. Obviously the needle should point outwards and to prevent accidents a short piece of 30 mm plas- tic pipe should be mounted coaxially with it. After some use the point will become dirty and possibly eroded, so making the needle removable for cleaning is also a good idea. Safety first is a good motto to follow when mounting the circuit in the car. Use an insulated box to contain the electronics and position the unit within the car so that it is not a hazard to unsuspecting passengers. K 1 77100 02162 1 Figure 1 . With this circuit the ioniser can be used in the car. With a 12 V d.c. input approxi- mately 400 V a.c. is produced. Cl C27 = 33 n ... 47 n/630 V (27 x) 01. 027= 1N4007 (1000 V diode) (27 x) 9823 82162 2 Figure 2. The circuit diagram of the ioniser, consisting of 27 diodes and 27 capacitors. The unit is a voltage multiplier delivering 7.5 kV to the probe or needle. Parts list for the power supply Resistors: R1 = 1 k R2 = 47 k R3.R4 = 470 M'A W PI - 47 k preset Capacitors: Cl = 150 n C2= 10 n C3,C4 - 560 p Semiconductors: T1,T2= BO 139 D1 ,D2 - 1 N4004 D3.D4 = 27 V/400 mW zener IC1 = 555 Miscellaneous: Trl = 2x6 V/0.8 A transformer 2 heat sinks for the BO 139 SI = on/off switch Figure 3. The printed circuit board of the power supply. There is nothing critical in its construction. The transformer uses the 220 V winding as the secondary. Parts list for the ioniser Resistors: R1 . . . RIO = 3M3 Capacitors: Cl . . . C27 = 33 n . . . 47 n/630 V Semiconductors: D1 . . . D27- 1N4007 (1000 V) F = 75 m A fuses 12-48 — elektor dece mber 1982 floppy-disk interface for the Junior Hardware of the DOS Junior In order to upgradethe Junior Computer to a DOS computer, certain hardware modifications are necessary. To allay any fears, let us point out that no tracks need be cut nor need any mechanical changes be made. All that is required is to solder an 1C onto another one in piggyback fashion on the interface p.c.b. of the Junior Computer. In order to connect an EPSON dot matrix printer, an interface is required for the BUSY line. Since this interface only consists of G de Cuyper three resistors, one transistor and one 7 diode, it can easily be mounted in self-supporting fashion next to the V24/RS-232 connector. floppy- disk interface for die Junior . . . and other 6502 computers. Part II the Junior collects its stars and stripes The second and last of these two articles describes the modifications that must be made to the hardware of the Junior Computer, to be able to run Ohio Scientific software on the Junior or any other 6502 computer. A new EPROM is required, to load software from the diskette during initialization (reset) of the computer. Documentation is available showing the source listing for the monitor program in the EPROM. Let us take another look in the circuit diagram of the Junior interface p.c.b. (figure 1). NOR gate N33 is replaced by a NAND gate. Line 8K0 or EX is now no longer active in address range $0000 . . . 1 FFF. The new address range with the NAND gate is E000 . . . FFFF. This addressing affects the memory chips of the Junior Computer as follows: — 48 K bytes of dynamic RAM are located in adress range $0000 . . . BFFF. The advantage of dynamic RAM is that it is cheap and has a low current consumption. Three dynamic RAM cards (see Elektor, April 1982) are sufficient to provide the 48 K byte address range for the RAM. — The address decoder on the standard Junior Computer p.c.b. (IC6) decodes address range $E000 . . . FFFF. The memory chips on the standard Junior Computer p.c.b. are therefore assigned the following addresses: EPROM IC2, type 2708: $FC00 . . . FFFF PIA, RAM, TIMER, type 6532: $FA00 . . . FBFF (documentation) RAM IC4 and IC5type 2114: $E000 . . . E3FF The memory chips on the Junior inter- face p.c.b. have the following addresses: VIA IC1 , type 6522: $F800 . . . F9FF (documentation) RAM IC2 and IC3, type 21 14: $E400 . . . E7FF EPROM IC4 and IC5, type 271 6: $E800 . . . F7FF The second hardware modification con- cerns the interface for the BU3Y line of the EP50N printer. Figure 2 shows how this interface is connected to the Junior. Relay Rel is discharded. LED D4 be- comes the BU5Y indicator in this ar- rangement and operates in parallel with the BU3Y lamp on the EPSON. The drawings of figures 3 and 4 clarify the hardware modifications. After im- plementation of these modifications it is only necessary to plug the type 2708 (E3S 515) EPROM into the socket on the standard Junior Computer p.c.b. The two EPROMs IC4 and IC5 on the Junior interface p.c.b. (PM and TM) are no longer required, because the input/ output programs for printer control are located in the 2708 EPROM. The ad- dress space assigned to IC4 and IC5 is now free for user programs. One more item that must not be forgotten is to fit a wire link between soldering points 'R' and '$' on the Junior interface p.c.b. (WITH). The Junior Computer has now been up- graded to a DO$ computer. Plug at least two or three dynamic RAM cards into the bus card of the Junior Computer. The lines for address decoding on the RAM cards should be wired as follows: RAM card 1: U-0 V - 1 X -2 Y — 3 RAM card 2: U-4 V -5 X -6 Y -7 RAM card 3: U-8 V -9 X-A Y -B At this stage, do not yet plug the floppy disk interface into the bus card of the Junior. Power-up the computer as usual and press the key on the hex keyboard. The display on the Junior should now light up. The command keys , , <+> and have the same significance as previously. Only the key has a new function. Once the disk operating system has been loaded into the computer, the DOS command interpreter can be called with the key. But we shall examine this in more detail later. Table 1 shows the memory map of the DOS Junior Computer. This memory map also applies to all other 6502 com- floppy-disk interface for the Junior elektor december 1982 — 12-49 5 V Figure 1. A part of the circuit diagram of the Junior Computer interface card. Gates N33 and N34 are replaced by a 74LS10 NAND gate. This results in a new memory area for the memories on the standard Junior Computer and on the interface card. puters that are connected to the floppy disk interface. Adresses $C000 . . . FBFF can vary from one computer to another. The only important detail is that the computer must be able to address at least 32 K bytes of contiguous RAM in the lower address range. If, on a compu- ter other than the Junior, the address range FCOO . . . FFFF should already be assigned, the bootstrap software should be relocated to another memory area. This should be relatively simple using the documentation. Software of the DOS Junior The software of the DOS Junior Com- puter is oriented on the latest standards in computer design. This means that the computer is provided with a minimum of ROM intelligence and as much RAM as possible. The advantages of this com- puter system become clear: Since the system software, whether BASIC, FORTH, Assembler or a word processor, has to be loaded rapidly from the diskette into the computer, not much space is wasted for the ROM ad- dress range. The ROM contains only suf- ficient intelligence to allow the compu- ter to operate the hex display and key- board and transmit/receive to and from the Elekterminal. Another function of the ROM intelligence of the Junior is to load track 0 from the diskette into the Junior Computer. Thus BASIC, Assembler etc. are no longer stored in the ROM but are loaded from the diskette into the computer. This is referred to as 'portable software'. The decisive advantage of portable soft ware is that it can be easily modified. Previously, when software errors were detected or when the system was to be updated, it was necessary to plug a new BOOTSTRAP RES. IRQ. NMI 6532: RAM. TIMER, PIA IC5 Interface Board (27t6) IC4 Interface Board <2716> 2k Static RAM Table 1 . The new memory map for the DOS Junior Computer. On account of these extensions the workspace is reduced by 4 K bytes with the OS-65D V3.3. The transient processor overwrites address range $ 2200 . . . 22FF of the DOS software. set of EPROMs into the computer; now with portable software we only need to mount a new diskette on the drive. It is therefore easy to eliminate software errors and the system can be upgraded in keeping with the latest state of the art without difficulty. We make full use of these advantages with the DOS Junior Computer. Let us examine Table 1 once again. Page 0 and the system stack are located in address range SOOOO . . . 01FF. Ad- dress space $0200 . . . 22FF is intended for the 'transient processor’. What is a transient processor? This is the soft- ware that makes the computer system operable. If, for example, you are working in BASIC, the transient proces- sor is the BASIC interpreter. If you are working in machine language, then Assembler or Ohio Scientific's Extended Monitor is the transient processor. The system software for controlling the floppy disk drive and the printer, as well as a memory-mapped video display unit takes up about 4 K bytes of RAM. With OS-65D V3.1 the DOS software occupies memory area $2300 . . . 3278. The work space or memory area in which the programmer's own program is stored begins at address $327E . . . When you type a BASIC program into the compu- ter it is stored from address $327E . . . onwards. The data are written into the floppy disk drive from this address onwards. Data which are read from the diskette into the computer are also stored from this address onwards. With OS-65D V3.3 the work space begins at address S3A7E. The first five bytes of the work space are known as the header. These five bytes contain the following information: 1 ) Start address on the file (2 bytes) 2) End address on the file (2 bytes) 12-50 - elektor december 1982 floppy-disk interface for the Junior 2 Figure 2. This is how the interface is connected to PB6 for the EPSON printer. If PB6 is a logic 1 , the send output (RS 232 connector pin 31 of the computer is inhibited. If no printer is connected, resistor RI1 ensures that the send output is enabled. 74LS27 82187 3 Figure 3. This hardware modification provides the Junior Computer with a new memory arrangement. The drawing shows the 'piggy back' installation of the ICs. 3) Length of the file in pages (1 byte; 1 page = 256 bytes) Next to the work space is the 4 K byte page $DXXX. The memory-mapped video display unit of Ohio Scientific is located in this address range. A video unit of this type differs from the Elek- terminal in that the computer can trans- port data to the monitor screen directly via the data bus. The data interchange between the screen and the computer is much more rapid with the memory- mapped video display unit (VDU) than with a video interface using the V24/RS-232 interface. 2 K bytes of static RAM are located in address range $E000 . . . E7FF. This RAM was previously located in the range $0000 . . . 07FF. In future we shall use this RAM area for the object code that is generated by the Assembler. You can now assemble a source file with the Assembler, print a listing if desired and instruct the computer to execute directly the machine code generated by the Assembler. Please consult Ohio Scientific's Assembler Manual for fur- ther details. Adress range SE800 . . . FFFF to IC4 and IC5 on the interface card of the Junior and to the two interface chips 6522/6532 and the 2708 bootstrap EPROM. The two EPROMs IC4 and IC5 on the interface card of the Junior are now freely accessible to the pro- grammer. Personal software or programs which one does not wish to load from the floppy disk into the computer can be loaded into the two EPROMs (2716) for long term applications. The bootstrap EPROM (ESS515) The bootstrap EPROM is addressed in the range $FC00 . . . FFFF. It contains only 1 K byte of software. The software in the bootstrap EPROM can be subdivi- ded into eight groups: 1) Hex display monitor. This program is very similar to the original monitor of the Junior. The commands , , , <+> have kept their significance. Only the key has a different function. The DOS command interpreter can be called directly from the hex keyboard with the key. The main purpose of the hex display moni- tor is to be able to modify the software on the Ohio Scientific diskettes. This allows Ohio software to be converted to Junior software. Important start addresses can be called with the hex display monitor and star- ted with the key. - RESBAS * $FF17 - RESDOS * $FF34 - VONE * $FFE2 - VTHREE * SFFE8 2) Loading the BASIC interpreter from the diskette. You can load BASIC from the diskette into the Junior Com- puter with the command FF17 (RUBOUT). To distinguish between command keys on the hex key- board and Elekterminal keys, we shall floppy-disk interface for the Junior elektor december 1982 — 12-51 4 Figure 4. The EPSON interface is placed on the interface card of the Junior Computer as shown here. write the command keys of the hex key- board between <> symbols and com- mand keys on the Elekterminal between () brackets in future. Once BASIC is loaded and the computer has responded on the terminal the tran- sient processor is ready for program interrupts. During output, you can interrupt a BASIC program with the (BREAK) key on the terminal. If the (BREAK) key was pressed during a LIST command, the BASIC interpreter prints 'Break'. If you have started a BASIC program with the RUN com- mand and wish to stop output of the program, simply press the (BREAK) key. The BASIC interpreter responds with 'BREAK IN LINE X'. All program variables and pointers are then saved in the stack. You can restart the program after the interrupt with the CONT state- ment. The indirect jump vector for the (BREAK) key is automatically set by the computer via address SFF17 when the BASIC program is started. The break vector is stored in address SFA7C and SFA7D. 3) Loading of DOS software from the diskette. In future you can load ELEKTOR's own software into your Junior Computer with the command FF34 (RUBOUT). The address DOS is intended for future ex- pansions and 'non-Ohio-software'. 4) Adapting an OS-65D V3.1 Ohio diskette. An OS-65D V3.1 Ohio Scientific dis- kette can be adapted to your Junior Computer with the command FFE2 (RUBOUT). When the monitor is started at the address VONE * $FFE2 the following occurs: - The computer positions the read/ write head of the drive on track D. — The computer reads the data on track 0 and stores these data at address $2200 ... in the memory. — The computer positions the read / write head of the drive on track 1 . — The computer reads the data on track 1 and stores these data at address $2A00 ... in the memory. The entire disk operating system is now loaded in the computer and can be mo- dified by the programmer with the hex keyboard. — Once track 0 and track 1 have been loaded into the computer a jump is made to the hex display monitor and the computer responds with *T rack 0&1 * on the printer. 5) Adapting an OS-65D V3.3 Ohio tutorial disk 5. An OS-65D V3.3 Ohio Scientific tutorial disk 5 can be adapted to your Junior Computer with the command FFE8 (RUBOUT). For track 1 and track 1 the same applies as under 4). However, once track 0 and track 1 have been loaded into the computer the fol- lowing occurs: — The computer moves the read/write head of the drive to track 6. The data on track 6, sector 2 are read into the memory from address $3200 . . . on- wards. Sector 2 is one page long. — The data on track 6, sector 3 are read into the memory from address $0000 . . . onwards. Sector 3 is one page long. — The computer removes the read/write head of the drive to track 1 3. The data on track 13, sector 1 are loaded into the computer from address $3A79 . . . on- wards. Sector 1 of track 13 is eight pages long. When all tracks have been loaded into the computer the message "Track 0&1* is issued, as with OS-65D V3.1, and a jump is made to the hex display monitor. 6) The key You can leave the hex display moni- tor and jump to the DOS command interpreter with the key. Printer I/O is initialized but no new baud rate measurement is made. The computer responds with the prompt A* or B *, etc.. 7) Printer input routine. The printer input routine receives a character from the terminal. The re- ceived ASCII character is in the accumu- lator of the CPU after the return from the input subroutine. Bit 7 of the charac- ter is always a logic zero. The contents of the index registers are not affected Transfer of Control Commands Figure 5. The command transfer to the different transient processors: from BASIC to ASSEMBLER, from ASSEMBLER to EXTEND MONITOR, etc. and vice versa. The 'control centre' is the DOS. It allows the BASIC, ASSEMBLER and EXTENDED MONITOR transient processors to be loaded from the diskette into the computer. 12-52 — elektor dscembsr 1982 floppy-disk interface for the Junior by calling this subroutine. The start address of the printer input routine is: RECCHA* $FE1B (call: JSR RECCHA). 8) Printer output routine. The printer output routine sends the character in the accumulator of the CPU to the terminal. The transfer for- mat is: — one start bit — seven data bits — no parity bit — two stop bits The contents of the index registers are not affected by calling this subroutine. The start address of the printer output routine is: PRCHA* $FEA3 (call: JSR PRCHA). How does the bootstrap routine operate? Now that we know what the software in the bootstrap EPROM consists of, we should discuss operation of the boot- strap routine. Once you have entered FF1 7 into the computer, the following opera- tions take place: 1) The computer sets the indirect jump vector for the (BREAK) key on the terminal. The NMI vector points to the hex display monitor. 2) The computer initializes the I/O on the floppy disk interface p.c.b. (6821 and 6850) and the serial I/O for the ter- minal (6532). When all I/O lines have been initialized the computer waits for the RUBOUT character from the ter- minal, to measure the baud rate. The baud rate is stored in memory cells SFA5A . . . FA5D (see floppy disk documentation). 3) Once the baut rate of the terminal has been measured, the computer loads track 0 (= 2 K bytes of software in machine language). The steps re- quired to load track 0 into the compu- ter are as follows: — Position the read/write head of the drive over track 0. A light barrier in the drive informs the computer when the head is positioned over track 0. — The computer then emits a head load pulse to the drive. The read/write head is lowered onto the surface of the diskette. The computer then waits for the index hole in the diskette. Another light barrier generates the index pulse as soon as the index hole of the diskette passes the light barrier. — Once the index pulse has decayed, the control register is set in the ACIA (6850). The transfer format of the ACIA is: one start bit, eight data bits, one even parity and one stop bit. — The computer reads the first byte from the diskette. This byte is the most significant start address of the memory area in which track 0 is stored (= $22). The second byte from the diskette is the least significant byte of the start address (= $00). Both bytes are loaded into the bump pointer (load FFE2 (RUBOUT) •TRACK Bil* 2217 4C <♦> 40 <+> 22 2245 4C <♦> 76 <+> 22 2283 4C <♦> A6 <♦> 22 A A .1200 = 13,1 Table 2. Adapting an OS-65D V3.1 diskette to the Junior Computer (Part 1 ), if only one drive is available. A *IN ARE YOU SURE7Y A*GO020O - DISKETTE UTILITIES - SELECT ONE: 1) COPIER 2) TRACK 0 READ/WRITE ? 2 - TRACK ZERO READ/WRITE UTILITY - COMMANDS: Rnnnn - READ INTO LOCATION nnnn. Wnnnn/gggg,p - WRITE FROM nnnn FOR p PAGES WITH gggg AS THE LOAD VECTOR E - EXIT TO OS-65D COMMAND? W2200/2200 , 8 - TRACK ZERO READ/WRITE UTILITY - COMMANDS: Rnnnn - READ INTO LOCATION nnnn. Wnnnn/gggg,p - WRITE FROM nnnn FOR p PAGES WITH gggg AS THE LOAD VECTOR E - EXIT TO OS-65D COMMAND? E A* i H0ME A *SA 31.1-2A03/8 Table 3. Adapting an OS-65D V3.1 diskette to the Junior Computer (Part 2). A *CA 4000-02,1 A*CA 4000-10,1 A*CA 4000-18,1 < u < 4000-26,1 A*CA 4800*03,1 A*CA 4800-11,1 A*CA 4800=19,1 A*CA 4800-27,1 A*CA 5000=04,1 A*CA 5000-12,1 A*CA 5000-20,1 A*CA 5000-28,1 A*CA 5800=05,1 A*CA 5100-12,2 A*CA 5800-21,1 A*CA 5800=29, 1 A*CA 6000-06,1 A*CA 5200-12,3 A*CA 6000-22,1 A*CA 6000-30,1 A*CA 6800-07,1 A*CA 5300-12,4 A*CA 6800-23,1 A*CA 6800-31,1 A*CA 7000-08,1 A*CA 5800-13, 1 A*CA 7000-24,1 A*CA 7000-32,1 A*CA 7800-09,1 A*CA 6000-14,1 A*CA 7800-25,1 A*SA 26,1=4000/8 A*SA 02,1-4000/8 A*CA 6800-15,1 A*SA 18,1=4000/8 A*SA 27,1-4800/8 A*SA 03, 1=4 800/8 A*CA 7000-16, 1 A*SA 19,1-4800/8 A*SA 28, 1=5000/8 A*SA 04,1-5000/8 A*CA 7800-17,1 A*SA 20,1-5000/8 A*SA 29,1-5800/8 A *SA 05,1-5800/8 A*SA 10,1=4000/8 A*SA 21,1-5800/8 A*SA 30,1=6000/8 A*SA 06,1-6000/1 A*SA 11,1=4800/8 A*SA 22,1-6000/8 A*SA 31,1=6800/8 A*SA 07,1-6800/8 A*SA 12, 1-5000/1 A*SA 23,1-6800/8 A*SA 32,1-7000/8 > V) > 08,1-7000/8 A*SA 12,2-5100/1 A*SA 24,1=7000/8 A* A*SA 09,1-7800/5 A*SA 12,3-5200/1 A*SA 25, 1-7800/8 A*SA 12,4-5300/1 FF17 A*SA 13,1-5800/5 (RUBOUT) A*SA 14,1-6000/8 A*SA IS, 1-6800/8 A*SA 16,1-7000/8 OS-65D V3.0 OSI 9 DIGIT BASIC A*SA 17,1=7800/8 COPYRIGHT 1977 BY MICROSOFT 36225 BYTES FREE OK Table 4. Adapting an OS-65D V3.1 diskette to the Junior Computer (Part 3). pointer). Thus the bump pointer points to address $2200. The computer then reads the third byte from the diskette. This byte indicates the number of pages on track 0 (= $08). — The next bytes are 2 K bytes of soft- ware in machine language. These bytes are loaded from the diskette into the computer. The memory area is $2200 ...29FF. — Once track 0 has been stored in the memory, the read/write head is raised from the surface of the diskette and a jump to address $2200 takes place. $tarting with this address, the computer finds further instructions con- cerning the address into which the tracks and sectors are to be loaded. Nor- mally the two K bytes of software on track 1 are loaded into address range floppy-disk interface for the Junior elektor december 1982 — 12-53 RUN "BEXEC * * BASIC EXECUTIVE FOR OS-65D V3.1 JUNE 2S, 1980 RELEASE FUNCTIONS AVAILABLE: CHANGE- ALTER WORK- SPACE LIMITS DIR- PRINT DIRECTORY UNLOCK- UNLOCK SYSTEM FOR END USER MODI- FICATIONS FUNCTION? UNLOCK SYSTEM OPEN OK RUN BASIC EXECUTIVE FOR OS-65D V3.1 JUNE 2S, 1980 RELEASE FUNCTIONS AVAILABLE: CHANCE- ALTER WORK- SPACE LIMITS DIR- PRINT DIRECTORY UNLOCK- UNLOCK SYSTEM FOR END USER MODI- FICATIONS FUNCTION? DIR LIST ON LINEPRINTER INSTEAD OF DEVICE f 1 ? NO OS-6SD VERSION 3.0 — DIRECTORY — FILE NAME TRACK RANGE 0S65D3 0-12 BEXEC* 14 - 14 CHANGE IS - 16 CREATE 17-19 DELETE 20 - 20 DIR 21 - 21 DIRSRT 22 - 22 RANLST 23-24 RENAME 25 - 25 SECDIR 26 - 26 SEQLST 27 - 28 TRACE 29-29 ZERO 30 - 31 ASAMPL 32 - 32 50 ENTRIES FREE OUT OF 64 OK Table 5. Trial run' of an adapted OS-65D V3.1 diskette. Utility program BEXEC* is called when the BASIC interpreter has been loaded. S2A00 . . . 31FF. Track 0 and track 1 jointly contain 4 K bytes of software in machine language, which make the en- tire disk operating system operable. - Once the disk operating system (DOS) has been loaded into the com- puter, the BASIC interpreter is loaded from the diskette into the memory. With OS-65D V3.1 the BASIC interpre- ter is located on tracks 2 ... 4. With OS-65D V3.3 the BASIC interpreter and various extensions of the editor are lo- cated on tracks 2 ... 4, track 6 and track 13. When the interpreter has been loaded a jump is made to the cold-start entry of the BASIC interpreter ($20E4). The system then responds with the prompt 'OK'. - The system is not yet ready for the statements LIST, CONT etc.. Even a 3ASIC file cannot yet be created. The OK LIST 10 REM DIRECTORY UTILITY FOR OS-65D VERSION 3.0 2 0 REM 30 NF-0 40 PN-11897 50 DEF FNA (X)«10*INT (X/16)*X-16*INT (X/l 6 ) BREAK OK NEW OK LIST ON LINEPRINTER INSTEAD OF DEVICE • 1 ? NO OS-65D VERSION 3.0 — DIRECTORY — FILE NAME TRACK RANGE 0S65D3 0-12 BEXEC* 14 - 14 BREAK IN 11110 OK CONT CHANGE 15 - 16 CREATE 17-19 DELETE 20 - 20 DIR 21 - BREAK IN 11100 OK Table 6. Checking the (BREAK) key with the LIST command and the RUN statement. A *CA 0200»13 A *GO 0200 - DISKETTE UTILITIES - SELECT ONE: 1) COPIER 2) TRACK 0 READ/WRITE - DISKETTE COPIER - FROM DRIVE (A/B/C/D)? A TO DRIVE (A/B/C/D)? B STARTING TRACK? 2 ENDING TRACK (INCLUSIVE)? 32 READY (Y/N) ? Y 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Table 7. If two floppy disk drives are available (drive A and drive B), an OS-65D V3.1 diskette can be adapted very easily. Copying is performed automatically by the computer. only BASIC statement that the compu- ter recognizes is RUN. If you wish to generate a BASIC file, load the program BEXEC* into the computer using the command: RUN 'BEXEC*' The computer then presents a menu of options. With OS-65D V3.1 select the option 'UNLOCK' and with OS-65D V3.3 select option '9'. The computer responds with the prompt 'SYSTEM OPEN'. Now enter a 'NEW' statement and the work space of the Junior is formatted for your BASIC file. The DOS and its capabilities The Junior Computer can easily be con- verted to various transient processors with Ohio's disk operating system, which is located on track 0 and track 1 of the diskette. Figure 5 shows how the individual connections between the DOS and the processors can be made and cancelled. If, for example, you start the DOS at address $FF17 the Junior becomes a BASIC computer. If you want to convert the Junior to an Assembler computer, you must enter EXIT in order to leave the BASIC inter- preter. After the EXIT command the Junior responds with the DOS prompt A* or B*, etc. on the printer. If you then enter AS or Assembler the Junior converts itself to an Assembler Compu- ter. You can then create an Assembler file and write on the diskette with the DOS. However, you can also assemble the source file and program the object code directly in the EPROM with the ELEKTOR EPROM programmer, with- out having to type in a single byte. The computer performs this task automati- cally. For further details regarding operation with Assembler, please con- sult Ohio Scientific's Assmbler Manual. Note that the Assembler and Extended Monitor are not on the diskette in the case of OS-65D V3.1 . Only OS-65D V3.3 has the Assembler and Extended Monitor (EM) on the diskette as standard. Adapting an OS-65D V3.1 diskette for one drive If you have only one drive, V3.1 can only be adapted to the Junior Computer with a considerable amount of typing. Tables 2, 3 and 4 show how this is done. First we start the system with the hex keyboard. A program with which a V3.1 diskette can be loaded into the Junior Computer to facilitate modifi- cation begins at address SFFE2. When the (RUBOUT) key has been pressed the computer loads track 0 and track 1 into the memory and responds with ’TRACK 0&1*. The display lights up, the vectors for the input and output routines are automatically loaded and the programmer can now modify some bytes at addresses $2217, $2245 and $2283. Once the bytes have been modi- field in accordance with Table 2, a jump to the DO$ command interpreter is made using the key. The DO$ responds with the prompt A*. The command CA 0200-13,1 loads from track 13, sector 1 of the diskette. The data are stored in the computer from address 0200 onwards. The program that we have just loaded into the Junior Computer is a disk copier and track 0 read/write utility. You can write the modified DO$ onto track 0 of your dis- kette with the track 0 read/write utility program. But first you must initialize your diskette. Remove the Ohio dis- kette from the drive and insert your dis- kette, then close the door of the drive. Table 3 shows the next steps: type the word IN. This command initializes the diskette. The system asks: 'ARE YOU $URE? '. Your reply is Y ( ES) . The drive ceases clicking when the formatting in- 12-54 — elektor december 1982 floppy-disk interface for the Junior FFE8 (RUBOUT) •TRACK 041* 2217 4C <♦> 40 <+> 22 2245 4C <+> 76 <♦> 22 2285 8E <♦> C6 <+> 2A <♦> 4C <♦> B3 <♦> 22 2E84 4C <♦> B0 <♦> 2E Table 8. Adapting an OS-65D V3.3 diskette. BASIC is loaded via a jump to the DOS command interpreter (GO 2276) and copied with the utility program BEXEC*. Finally a 'trial run' is executed to ensure that the diskette was adapted correctly. A*CA 0200 «06 , 4 A*GO 0200 - TRACK ZERO READ/WRITE UTILITY - ATNENB 35 - 35 COLORS 36 - 36 MODEM 37 - 38 COM PAR 39 - 39 46 Entries free out of 64 Depress RETURN to continue ? FFE8 (RUBOUT) •TRACK 0&1 • A*GO 2276 OS-65D V3.0 OSI 9 DIGIT BASIC COPYRIGHT 1977 BY MICROSOFT 33921 BYTES FREE Ok RUN "BEXEC*" OS-65D Tutorial disk five - Sept. 16, 1981 1 > Directory 2 > Create a new file 3 > Change a file name 4 > Delete file from diskette 5 > Create blank data diskette 6 > Create data diskette with files 7 > Create buffer space for data files 8 > Single or dual disk drive copier 9 > Enter OS-65D system Type the number of your selection and depress RETURN 7 8 COMMANDS: Rnnnn - READ INTO LOCATION nnnn. Wnnnn/gggg ,p - WRITE FROM nnnn FOR p PAGES WITH gggg AS THE LOAD VECTOR E - EXIT TO OS-65D - Diskette copier - Copy from which drive (A/B/C/D) ? A Copy to which drive (A/B/C/D) 7 A COMMAND? W2200/2200 , 8 What is the last track to be copied (Inclusive) <0-39> 7 39 - TRACK ZERO READ/WRITE UTILITY - Are you ready to start copying (Y/N) 7 Y COMMANDS: Rnnnn - READ INTO LOCATION nnnn. Wnnnn/qqqg ,p - WRITE FROM nnnn FOR p PAGES WITH gggg AS THE LOAD VECTOR E - EXIT TO OS-65D Insert master diskette — press 7 Reading — Insert blank diskette — press 7 COMMAND? E Initializing — A*SA 01 , l=2A00/8 Track Track 01 02 _ 01/08 01/08 A* Track 03 - 01/08 Track 04 - 01/08 Track ab - 01/08 FF17 Track 06 - 01/01 - 02/01 - 03/01 - 04/02 Track 07 - 01/08 Track 08 - 01/08 (RUBOUT) Track 09 - 01/08 Track 10 - 01/08 Track 11 - 01/01 - 02/01 - 03/01 - 04/01 - OS-65D V3.0 Track 12 - 01/01 - 02/01 - 03/01 - 04/01 OSI 9 DIGIT BASIC Track 13 - 01/08 COPYRIGHT 1977 BY MICROSOFT Track 14 - 01/08 34177 BYTES FREE Insert master di skette - - press 7 05/01 - 06/01 - 07/01 Ok RUN "BEXEC*" Reading — OS-65D Tutorial disk five - Sept. 16, 1981 1 > Directory 2 > Create a new file 3 > Change a file name 4 > Delete file from diskette 5 > Create blank data diskette 6 > Create data diskette with files 7 > Create buffer space for data files 8 > Single or dual disk drive copier 9 > Enter OS-65D system Type the number of your selection and depress RETURN 7 1 Directory utility Directory of which drive 7 Type A, B ,C or D and depress RETURN 7 Do you want to list the directory to the printer (Yes or No) 7 — Directory — File name Track range OS65D3 0 - 13 BEXEC* 14 - 16 COPIER 17 - 18 CHANGE 19-20 CREATE 21 - 22 DELETE 23 - 23 DIR 24 - 24 RANLST 25 - 26 RENAME 27 - 27 SECDIR 28 - 28 SEQLST 29 - 30 TRACE 31-31 ZERO 32 - 33 ASAMPL 34 - 34 Insert blank diskette — press 7 Track 15 _ 01/08 Track 16 - 01/08 Track 17 - 01/08 Track 18 - 01/08 Track 19 - 01/08 Track 20 - 01/08 Track 21 - 01/08 Track 22 - 01/08 Track 23 - 01/08 Track 24 - 01/08 Track 25 - 01/08 Track 26 - 01/08 Track 27 - 01/08 Track 28 - 01/08 Insert master diskette — press 7 Reading — Insert blank diskette -- press 7 Track 29 - 01/08 Track 30 - 01/08 Track 31 - 01/08 Track 32 - 01/08 Track 33 - 01/08 Track 34 - 01/08 Track 35 - 01/08 Track 36 - 01/08 Track 37 - 01/08 Track 38 - 01/08 Track 39 - 01/05 - 02/02 floppy -disk interface for the Junior elektor december 1982 — 12-55 6 2 x Figure 6. These two hefty power supplies are more than adequate for powering the DOS Junior Computer. They can both be constructed on Elektor printed circuit boards (EPS 825701. They are sufficiently rated to power a complete Junior Computer, including four disk drives. formation has been written on all 39 tracks of the diskette. After initiali- zation the diskette is ready for read/write operations. Now proceed according to Table 3. First you write eight pages of software on track 0 with the command A/2200/2200,8. The start address is S2200 and the load vector for the boot- strap function is also $2200. The first naif of the DOS is thus written on the diskette. The command SA 01,1=2A00/8 saves a data block of a length of eight pages on track 1, sector 1, beginning with start address $2A00. You save the second part of the DOS on the diskette with this command. Since, with the V3.1 copier it is not possible to copy from drive A to drive A, copying must be performed track by track and manually. Table 4 shows how this is done. It should be noted that with the 'CA' commands the Ohio diskette is mounted in drive A, whilst with the "SA' com- mands your diskette is mounted in drive A. When tracks 2 ... 32 have been copied from the Ohio diskette onto your diskette, a trial run can be started. Switch off the supply voltage of the computer. Wait a few seconds and switch the supply voltage on again. Mount your diskette in drive A. Start the bootstrap function at address $FF17 as shown in Table 4. The system responds with a message and the prompt 'OK'. The number of free bytes depends on the capacity of the RAM. With a 48 K byte dynamic RAM you have access to 36225 memory locations. Start the utility program BEXEC* with the command RUN 'BEXEC*'. This program is written in BASIC. The com- puter prints a menu of options. First select the option UNLOCK; the compu- ter is now ready for all BASIC state- ments. Proceed as shown in Table 5. Start the directory utility program. The computer then prints the entire directory. Finally, a check should be made to establish whether the (BREAK) key of the termi- nal is functioning correctly. Table 6 shows how a LIST command is first interrupted and a program is then termi- nated with the (BREAK) key. You can continue the program from its point of interruption using the CONT command. Adapting an OS-65D V3.1 diskette for two drives It is considerably easier to adapt V3.1 to the Junior Computer when two 12-56 - elektor december 1982 floppy-disk interface for the Junior drives are connected. Modify and copy track 0 and track 1 in accordance with 7 Table 2 and Table 3. When these two tracks are both on the diskette, remove your diskette from drive A and mount it in drive B. Now mount the Ohio diskette in drive A. Proceed according to Table 7. The computer now automatically copies all tracks from the Ohio diskette in drive A onto your diskette in drive B, beginning with track 2 and ending with track 32. After this procedure your dis- kette is adapted to the Junior Computer with the Ohio software. Adapting an OS-65D V3.3 diskette The V3.3 diskette is much easier to adapt than the V3.1 diskette. Only one drive is required. Table 8 shows how to proceed. First we load track 0 and track 1 into the computer. Since the DOS is located in the memory, we load the BASIC interpreter with the com- mand GO 2276. The prompt OK shows that the interpreter is loaded. Then we load the utility program BEXEC* with the command: RUN 'BEXEC*' The computer prints a menu of options and we select the option '8'. With this option we are selecting the copier. When all the tracks have been copied, some modifications must be made to the disk operating system. Press the key again and start the program at address SFFE8. Tracks 0 and 1 are then loaded into the computer again. However, memory area $2200 . . . 22FF is no longer overwritten by the BASIC inter- preter. Implement the modifications at addresses $2217, $2245, $2285 and $2E84 in accordance with table 8. Once these modifications have been implemented, load the track 0 read/write program from track 6, sector 4. Write the modified DOS onto track 0 of the diskette again using this program. The command SA01,1=2A00/8 saves eight pages of DOS software on track 1 . You now have a V3.3 diskette which is adap- ted to the Junior Computer. You can start BASIC at address $FF17. The DOS and BASIC interpreter are automatically loaded into the Junior Computer after you have pressed the (RUBOUT) key — see Table 8. Figure 7. The power supplies for the dynamic R AMs and the printer interface can also be built quite simply with Elektor printed circuit boards (EPS 9968-5a). The DOS command interpreter As already mentioned, the disk ope- rating system has its own command interpreter. We shall now examine the most important commands. If a com- mand is not entered correctly an error message is issued. Whenever the computer has printed the DOS prompt A* or B* you can enter a DOS command after the prompt. The computer only recognizes the first two upper-case letters of each command. With a SAVE command, for example. it makes no difference whether you enter SA, SAV or SAVE. Command AS or ASM The computer loads the Assembler and Extended Monitor from the currently selected drive. Once this program has been loaded in machine language, a jump is made to the cold-start entry of the Assembler. The Assembler is linked with a line-oriented editor. Command EM The computer loads the Assembler and Extended Monitor from the currently se- lected drive. Once this program has been loaded in machine language, a jump is made to the Extended Monitor. This is a program with which a machine language program can easily be checked. The Extended Monitor has its own com- mand interpreter. The most important commands are: I8TRING Send 'STRING' to the DOS command interpreter as a command. aNNNN Open the memory cell with the address NNNN for execution of the following subcommands: (LF) open the next memory cell (CR ) close the currently addressed memory cell (D)(D) write the data into the currentlt addressed memory cell (") print the ASCII character of the currently addressed memory cell (/) prepare the currently ad- dressed memory cell for data input n open the previously addres- sed memory cell for data in- put BN, LLLL Set the break point with number N at address LLLL. The numeric range of the break point number N is 1 ... 8. EN Eliminate the break point with num- ber N. A Print the accumulator contents resulting from the last break point. C Start the program after the last break point. DNNNN.MMMM Make a memory dump from address NNNN up to and excluding address MMMM. EX Leave the Extended Monitor and return to DOS. FNNNN,MMMM=DD Fill the memory area beginning with address MMMM and ending with address MMMM— 1 with the data DD. floppy-disk interface for the Junior elektor december 1982 — 12-57 GNNNN Jump to address NNNN and process the program there. HNNNN,MMMM(OP) Call of the hexadecimal arithmetic unit. The arithmetic unit prints the result NNNN (OP) MMMM, where (OP) can be equal to +, — , *, /. Addition, subtrac- tion, multiplication and division of hexadecimal numbers can be executed. MNNNN=MMMM,LLLL Move memory area MMMM . . . NNNN— 1 into the memory area that begins at address NNNN. RMMMM=NNNN,LLLL The Extended Monitor has a so-called relocator. This allows machine language programs to be moved to another me- mory area. The computer performs the correcting of all absolute addresses: 'Relocate' the memory area between NNNN . . . LLLL— 1 into the memory area that begins at address MMMM. The Extended Monitor has some other commands which you can find in the Ohio Manual. Command BA The command loads the BASIC inter- preter from the currently selected drive. Once the BASIC interpreter has been loaded, a jump is made to the cold-start entry. The interpreter provides infor- mation regarding the number of free memory locations in the system and responds with the prompt 'OK'. Command CA NNNN=TT,S or CALL NNN= TT,S Load the data from track TT, sector S of the diskette into the computer. Store the data in memory area NNNN... The numeric range for TT is 1 ... 39 and the numeric range for S is 1 ... 8. Command SA TT,S=NNNN/8PP or SAVE TT,S=NNNN/P Save the data in memory area NNNN . . . for P pages onto the diskette. The track number is TT, the sector number is S and the number of pages to be written onto the diskette is r. I he numeric range for TT is 01 . . . 39, for S it is 1 ... 8 and for P it is also 1 ... 8. Command Dl TT or DIR TT This command allows a check of the numbers of sectors on track TT to be made. The numeric range of TT is 01 . . . 39. Command IN or INIT The command IN initializes a new dis- kette on which no data have been writ- ten yet. If you wish to erase a recorded diskette, you can do so with the IN command. Command IN TT: This initializes track TT only. Figure 8. These modifications constribute greatly to the proper functioning of the dynamic RAM card. floppy-disk in terface for the Junior 12-58 — elektor december 1982 FC0O: FC 1 0 : FC20: FC 30 : FC40: FC50 : FC60: FC70: FC80: FC90: FCA0: FCB0 : FCC0 : FCD0: FCE0: FCF0: FD00 : FD10: F D 2 0 : FD30: FD40: FD50: FD60: FD70: FD80: FD90: FDA0 : FDB0 : FDC0: FDD0: FDE0 : 0 1 A9 IE FC D0 C9 13 01 85 12 D0 4C FD 0A 0A CA D0 85 F9 20 B 8 00 A 9 8C 82 CB FC 80 FA 60 A2 60 A0 03 18 00 10 FD 20 00 C0 C0 8C FB D0 CE FD 18 20 C0 30 10 C 0 85 FF D0 F 2 FF CA 78 A 9 A2 FF 8D 83 FB 20 D0 07 FD DC 09 E6 FE C9 0A 0 5 F9 A5 A 9 7F FC A5 FF 8E •r-A 68 29 8E 82 21 A 0 FF 0A 69 07 08 03 CF FD A 9 04 03 C 0 09 A 9 29 F7 BA FD FB AD 20 C5 A0 00 86 FE D0 F6 67 8D 8E 5B FA A 9 7C FC A 2 FF 14 C 9 FA D0 15 10 FF 91 FA 05 8D 81 F9 20 82 FA 80 49 0F 20 FA A0 01 20 B 0 03 CA D0 46 21 20 4 F 8D 01 88 8C 02 2C 8D 02 F 0 DD 00 C0 FD 85 20 C5 A9 FF 60 AD 82 FA FA EA 01 85 F0 FB 9A D8 11 D8 02 E6 F 2 85 FA 4C FF 85 FA A2 B8 FC E8 E 8 FF 60 CB FC FF 88 A 1 FC C8 10 FA 60 06 0E FD 60 C0 A 9 02 C0 00 C0 C0 20 A2 7F 10 FB FE AA FD 91 8D 02 10 C0 A9 00 A9 7F 8 9 FD A 2 20 7C 6C FA 06 A 9 F3 4C FF A 4 49 FC FA 4C 08 A 5 A 9 00 2D 80 48 84 A4 FC D0 FD D0 07 FA 8A 40 79 20 IE A0 00 4 0 8D 8E 03 F0 1C CE FD 8E 02 A9 03 20 C 5 FD C8 C0 60 4A 90 8D 80 8D 81 A B FF 9A FC F0 00 C9 00 85 0D FC FD D0 A2 04 49 FC FB 20 8D 81 FA 88 FC 4A 60 A 8 88 8C E 0 27 29 0F 24 30 FD 6C 8C 01 00 C0 C0 8C A 9 FF 09 08 C0 20 8D 10 FD 85 DO F8 A0 F8 FA AD FA A2 FA 4A 78 D8 F6 20 10 D0 FD F0 C9 14 0D B 1 06 FA A0 00 B8 FC FA A0 DO F5 4A 4A B 9 08 82 FA D0 F5 4A AA 19 12 FD 00 C0 A 9 A 2 e4 02 C0 8D 02 8D 02 D7 FC C0 A 9 FD 20 E6 FE 88 D0 11 C0 FC 8E 8D 83 20 7C El f; 06 A 9 0A C 9 D0 f3 FA CA 26 FB Bl FA A5 FA 03 A 2 A0 06 4 A 20 FD 8D E8 E8 A 9 15 98 10 02 78 20 28 4 0 8D 8E 01 60 A 9 C0 20 C0 A2 AD 00 58 8D C5 FD C6 FF FD 55 60 D8 5A FA FA A2 FDF0 FE00 FE 1 0 FE20 FE30 FE40 FE50 FE60 FE70 FE80 FE90 FEA0 FEB0 FEC0 FED0 FEE0 FEF0 FF00 FF10 FF 20 FF 30 FF40 FF50 FF60 FF70 FF80 FF9C FFA0 FFB0 FFCC FFD FFE FTP 03 8E 5E FA 20 72 8E 61 38 6E 20 81 AD 5A 2C 80 FA 60 0C AD FA E9 B0 EB AD 82 90 30 AE 59 F2 2C 7C FA 4C 51 81 FA FA A9 8D 25 8D 7A FA 20 26 A 9 21 23 8D 03 8D C4 BC 26 29 A 9 74 85 0A 2k er fc rr Pr 59 FA AD 5E FE 20 FA A2 62 FA FE AD FA 69 FA 10 AD 5C 5A FA 01 8D 60 8E FA 29 AD 82 FA AD 80 FA AD 82 2A A 9 4A 8D 00 8D 23 60 FA A 9 IE FD 2A 85 8D 22 23 A 9 2 3 8D 20 67 01 8D FE 20 54 52 I.C 2C 80 FA 30 FA 8D 5C FA 2B FE C9 7F 08 20 72 FE CA D0 FI F 0 62 FA 29 7F 01 8D 5A FA EA AD 5A FA FA 8D 5E FA 8D 5E FA AD 5E FA AD 5F 60 FA 8D 62 FE 8D 82 FA FA 09 01 8D 82 FA 09 01 10 04 AE 60 FA 29 FE 8D 27 8D 82 FA 83 FA 60 A 9 7A FA A 9 FC A 9 51 8D 7C FC 8D 7B FA A 9 28 8D A3 FF 20 54 27 23 8D C 6 2 A A 2 8D 11 23 12 23 8D 14 29 EE 5E 26 5E 26 A9 13 54 27 20 67 41 43 48 20 FF 4C C9 FF ff n rr fi FB 20 AD 5F D0 B5 20 81 07 18 8D 63 AD SB 8D 5E AD 5D 5B FA FA E9 FA AD 20 81 82 FA 8D 82 FA 60 82 FA A 9 00 2E 8D 8D 7B FA A 9 4C 18 26 A 9 86 FE 20 C6 8D 13 23 60 A 9 BC 20 BC 29 20 30 26 20 5! FF r 4F FE FA 8D 60 2C FE 2C 6E 62 23 AE FA 69 FA AD FA 8D 8D 5F 00 8D 82 FA FE A2 20 81 FA 20 2C 80 18 90 8D 80 7C FA FA 4C 2A 8D FD 6C 01 8D 20 67 29 A 9 23 A 9 EE 5E 85 FE 26 A 9 61 27 31 2A FF 4C 4F FF 4E 5F 5D FA 80 FA 80 FA FA CA 61 FA 0 0 8D 5B FA 5F FA FA 38 5F FA 29 40 07 4E FE CA 81 FE FA 10 CD 20 FA A 9 A9 FF 18 FD 7D FA 7A FA 5E 26 29 A 9 1A 8D FE 8D 26 A 9 85 FF 32 85 20 73 0D 0A 9A FF rc fc FA 6E A2 08 30 FB 10 09 D0 E 8 60 18 SB FA 8D 5F 38 B 0 AD 5E EA EA D0 F9 62 FA D0 ED CA DC FB 6C 03 FF 7F 8D 8D 7D A 9 C 3 A 9 C0 6C 7E 20 BC 01 8D 01 23 02 23 06 20 20 67 FF A 9 2D C > 00 4 FF I ' 4L r Table 9. Hex dump listing for the bootstrap EPROM ESS515. The source listing is available as part of the documentation. Command SE X or SELECT X One of four drives can be selected with this command. The computer only operates with one selected drive. X = A, AB, Cor D. Command LO FILE or LOAD FILE A file with a name can be loaded into the main memory with this command. However, the name of the file must be specified in the directory of the dis- kette. A file name can be generated in the directory (track 12) with the CREATE utility program. Further de- tails can be found in the manuals sup- plied with the diskettes. The file name must begin with an alpha character and can be 1 ... 6 characters long. Command PU FILE or PUT FILE A file with a name can be written from the main memory onto the diskette with this command. Before the file can be written on the diskette the file name must be specified in the directory. Commands PU TT and LO TT With this command a file in the work space of the computer can be written onto the diskette or read from the diskette into the computer without a file name. However, the file in the work space must not be longer than 2 K bytes. It is advisable not to use these two commands because it is easy to overwrite existing software on the diskette without being warned by an error message. Command RE or RETURN With RETURN commands you can re- turn from the DOS interpreter to the present transient processor: RE AS return to Assembler RE BA return to BASIC RE EM return to Extended Monitor You can find other commands in the Ohio Manual. We shall now end the soft- ware adaptation and discussion of the disk operating system and consider some special hardware requirements of the DOS Junior Computer. Hardware requirements of a 6502 DOS computer Computers on which a disk operating system is implemented require a good power supply. The power supply chosen for the DOS Junior Computer should have sufficient reserve capacity and should have a pulse-free output. If you already have an extended Junior Com- puter, you can continue to use the exis- ting power supply. It is only necessary to build a new 12 V/4 A power supply for the drive. If, however, you have not yet got a power supply you can use the circuits in figure 7. These power supplies are intended to power the en- tire DOS computer; they are easy to build and are reliable. The entire circuit can be accommodated on only four ELEKTOR printed circuit boards: 2-times ESS 82570 and 2-times ESS 9968-5a. Two transformers are required with the following ratings: 9 ... 10 V/ 10 A and 15 V/4.4 ... 5 A. Both trans- formers should be of the toroidal type. Toroidal transformers are more expens- ive than the laminated type but are lighter and generate less of an inter- ference field. The temperature rise of a toroidal transformer is also limited, even under full load. A third transformer with 2-times 15 V/ 1 A provides the supply voltage for the dynamic RAM cards and the printer interface. No particular demands are made of this transformer. With the power supply circuits of figures 6 and 7 you will never have problems with the DOS Junior or any other 6502 compu- ter, because they are generously over- rated and short-proof. The power sup- plies can be accommodated in a flat case of the dimensions 300 mm x 70 mm x 200 mm. Conductors with a cross-section of 1 .5 square millimetres should be used for the wiring. The dynamic RAM card Experience has shown that the dynamic RAM card 'refuses' to operate with cer- tain 6502 processors. Neither was the card very fond of the 6809 processor so far. For this reason we have made a few modifications to the dynamic RAM card which make it extremely reliable in con- junction with any 6502/6809 system. Figure 8 shows all the modifications that we made to the circuit diagram and p.c.b.. We wish you continued enjoyment with the DOS Junior Computer which is still the lowest-cost and favourite computer for home construction, with a disk operating system. H a dozen and one sounds elektor december 1982 — 12-59 No matter how unusual or irrelevant a sound may be, a use is always found for it and the number of people wanting to produce sound effects is simply amazing. To a certain extent noises of one kind or another are part of everyday life. In artificial environments such as recording studios, discos, concert halls and quiet front rooms the lack of birds chirping and explosions is noticeable. The only way to introduce well known and seemingly well loved (?) noises is to produce them synthetically. This is where the SN76477 1C comes into its own, because it contains all the in- gredients of a BBC effects laboratory, producing a large number of refreshing sounds. This Texas 1C is not new especially to Elektor readers. The main advantage of using a well proved compo- a dozen and one sounds... an assorted box of sound effects nent is that it is available at a relatively low price. The circuit will be of great interest to experimentors musicians and recording enthusiasts alike. A single chip with a few surrounding components realises a circuit which makes a respectable job of reproducing sounds like, rain, explosions, cars crashing and so on. Ideal for mixing new film sound tracks. The 1C The article is aimed at the practical aspects and not at the theory so we will stick to a brief survey of the most important features. The internal layout of the complex sound generator is given in figure 1. A closer look shows that there are three fundamental signals pro- duced. These are obtained from: the super low frequency oscillator (SLF), the voltage controlled oscillator (VCO) and the noise generator. The SLF provides two output signals: a square-wave processed by the mixer stage and a triangular wave form used to control the VCO by way of the external VCO/SLF select section. The oscillator frequencies of each stage are determined by the various external RC networks. It is clear from just a rudimentary description that the SN 76477 is a very versatile 1C. The circuit Because of the enormous capabilities of the 1C, to create an endless reper- toire is theoretically possible. But, for practical reasons a circuit able to produce a dozen or more different sounds is by far a better investment and a good basis for further experimen- tation. Figure 2 shows the complete circuit in block diagram form. The potentiometers set the frequency of each stage. The VCO consists of an oscillator which is dependant on the input voltage. This control voltage can either be the SLF output signal or an externally applied one. 12-60 — elektor december 1982 a dozen and one sounds SN76477 ANALOG LEVELS LOGIC LEVELS 81112 • 1 82176 - 1 Figure 1 . The block diagram of the internal components of the complex sound generator 1C. It basically contains an SLF oscillator, a VCO and a random white noise generator. 2 Figure 2. The complete circuit of the effects box in block diagram form. The output signals from the three stages are fed to the mixer. Depending on the logic levels presented to the mixer select inputs, one, or a combi- nation of the three is passed on. Figure 3 illustrates the practical circuit design of the effects box. Apart from the complex sound generator 1C nearly all the other components are just potentiometers and switches. Transis- tors T1 and T2 constitute a simple complementary pair AF amplifier with P4 as the volume control. Because of the number of switches and potentiometers the easiest way to explain their function is to list them. • PI adjusts the clock frequency of the pseudo-random white noise gener- ator. • P3 sets the SLF oscillator frequency. • SI determines whether the VCO is controlled by the SLF signal (pos- ition 2) or not (position 1). • S2 is used to switch from one SLF frequency range to another, (1 = high, 2= low). • S3, S4, S5 are connected to the select inputs of the mixer stage. • P2 determines the VCO frequency. a dozen and one sounds eiektor december 1982 — 12-61 S3, S4,andS5can be individual switches or combined into one 8 way 3 pole wafer (S7) as shown in figure 3. Table 1 itemises exactly what effect each position of S7 has. Each new sound produced can be further modified by rotating any one or all of the potentiometers. Constructors wanting to experiment further can try new values for R3, R4 and R6. Construction •Vith the accent on experimentation to produce a ready-made printed circuit board would be inappropriate, and anyway, as most of the components are switches and potentiometers it would be a waste of money. In fact vero board or something similar is ideal here. Any 4 to 8f2 speaker is suitable provided it can handle at least 100 mW. Substi- tuting the amplification stage (T1, T2) with the circuit shown in figure 4 provides a line feed to any Hi Fi system or external power amplifier. A single 9 V battery is sufficient. This is because the 1C contains its own internal voltage regulator (not shown in figure 1) which derives a stable 5 V from the original input voltage (pin 14 9 V in, pin 1 5 5 V out)). The current consumption will depend on the output volume but it should not exceed 50 mA. 12-62 — elektor december 1982 stop-signal override for model railways stop-signal override Km* model mihvajs Model railway enthusiasts know the problem: stop-signals disconnect the supply voltage from a section of track as soon as the signal indicates 'stop'. Any train approaching the signal is obliged to stop on the dead section of track. It can only continue its journey when the stop-signal indicates 'line clear' or 'reduce speed'. The problem arises when a train is sup- posed to approach the stop-signal from the opposite direction during shunting or at rural stations. This is not possible with the usual stop-signal circuit which disconnects the supply voltage, thus preventing traffic in both directions. What we need is a circuit that allows the stop-signal to operate almost like a diode: in the normal direction the signal stops trains but allows trains to travel in the reverse direction. As usual, we have found a simple solution. In addition to the stop-signal we need two rail contacts at the two ends of the controlled section (see figure 1). These are contact A at the end and contact B at the beginning, when viewed in the contact closes and bridges the circuit that was originally disconnected by the stop-signal. The train can travel in re- verse along the controlled section of track. As soon as it reaches contact B the flip-flop is reset and the original state is restored. A train travelling in the normal direction first reaches con- tact B, thus causing the flip-flop to be reset and enabling proper func- tioning of the stop-signal. LED D6 (drawn with dashed lines in the circuit) lights up when the signal override is effective. The power supply of the override cir- cuit is generously rated and can power several override circuits of this type. The current drawn by one of these circuits depends on the relay used. Transistor T1 can supply a maximum relay coil current of 100 mA. Instead of using a separate transformer for the power supply, the AC voltage can also be obtained from the 'lighting output' of the model railway trans- former. If a 12 V relay is utilized the AC voltage required is 15 ... 18 V and a 7812 volt- age regulator is needed for IC2. With a 5 V relay a type 7805 regulator must be used for IC2, in which case the trans- former secondary voltage should be approximately 8 ... 12 V. With a relay voltage of 5 V, the value of R5 should be 120 SI. H shunting possible in spite of the stop-signal normal direction of travel. The train travelling in the reverse direction reaches contact A first. The contact closes and sets flip-flop IC1. The Q- output of the flip-flop goes to logic 'V and energizes the relay via T1 . The relay advertisement elektor december 1982 — 12-63 elektor switchboard JUNIOR computer + interface + Elekterminal + ASCII-keyboard + power supply uncased £ 1 5. Write 1 . Japan St., Cheetham, Manches- ter. WANTED 5%" disk drives (Teac or Olivetti but Shugart con- sidered). Tel: Great Yarmouth 62102 after five. TESTEQUIPMENT oscilloscope D43 DC 15 MHz @ 100 mV/cm £110, digital voltmeter PDM35 £ 24, avo valve tester £ 30, bar- gains! P.E. Coughlin, 19, Victoria Cres- cent, Chelmsford, Essex CM 1 1QF ALL WANTED complete years of "Elektors", computer and radio test equipment, unbuilt kits, H.F. triband beam, 041-884.4988. ELEKTOR PROJECT WANTED active aerial, built or in kit includ- ing all components and printed circuits boards. Apply to: Mr. Samuel Wong, 37 South Rich- mond Street, Dublin 2, Ireland. CREED 15 teletype with inter- face and control program for atom. All leads and new roll of paper. Plug in + print £ 40. K.Y. Chang, 10, Ashley Street, Glasgow G3 6HW. SURPLUS COMPONENTS 1%" panel mounting fuseholders. 40p ea. Varley type 4-C/O relays. 0 k at 12 V 55p ea. p&p 20p C.W.O. F.R. Box, 23, Regency Gardens, Yardley Wood, Birming- ham B14 4JS CIRCUIT or handbook for Marco- ni CR100/Mod8 receiver. I.R.Turk. 11 Medway Crescent, North Hykeham, Lincoln. PE RANGER CB 6CHS pair £60 Ring T.C. Choy 01-505-9137 eves or write: London, 9 Peel Road, South Woodford El 8. PAST ISSUES OF ELEKTOR Electronics today intnl. in one complete lot. 0632 525791 be- tween 6 p.m. and 8 p.m. STEREO DISCO CONSOLE, complete except for record decks, great sound, autofade, VU-meters etc. £ 100 O N. O. Tel. 0706 50223. WANTED Microprocessor evalu- ation kit(s) ($DK85) ($DK86) (MEK6800D2) MMDI Losmac Texas univ. module 9900 E.T.I. Fair price paid. Tel. 043-879-262. FOR SALE entire students stock or exc. transistors BC, BD, BF, TIP, etc., etc. COS/MOS, TTL, etc., etc. Reason for sale com- putor. Mr. B.P. Watson, 142 King Street, Great Yarmouth Norfolk. NR30 2PQ. WANTED 7 and 3.5 Mcs crystals CW ends also holders FT243 type or WHU. J.W. Mackay, 11 Lans- downe Grove, Whitehaven. ELEKTOR JUNIOR computer built and working but not in case. £ 60 O.N.O. Tel. 52998 Basing- stoke, K.W. Pearson. CREED 75 teleprinter. Ideal for cheap listings and built to last! Just £20 O.N.O. + carriage. Phone (0202) 693646. FOR SALE Junior computer with power supply & books 1 & 2 per- fect WKG. Order £ 60. Phone 0603 411199. MINI MIXER for sale. High quality, special design. Electronics according to Elektor, with some adds. Price £ 80. Urosevic Predrag, Palmira Tol- jatija 24/19, 11070 H. Beograd, Yugoslavia. HYBRID G.E.C. P4E EMOCTVS & panels. Cheap to clear or ex- change for frequency meter. GDO or other test gear. Tel. 0742 311191 BC221 WWII frequency meter with calibration cards complete. Also Valves 6CH6, 6AK5. 6F33 what offers. Tel . Romford 41118. CASIO FX-602P programmable calculator 51 2 steps/88 memories. 3 months old new £ 75, yours for £60. Phone after 7 p.m. Jeffrey, Tel: 01-458-3025. ELECTRONIC MUSIC SYNTH equipment on eurocard modules for sale. Each card has 2 VCO, 1 VCF, 2 TG, 2 VCA, PS @ £ 170. Paolo Bozzola, via A. Molinari 20, 251 24 Brescia, Italy. FREE I am a private reader. I have read your rules and I enclose a valid switchboard voucher. Please place the following advertisement, free, in the next available space. advertising for our renders! Rules: • Private advertisers only. No trade, no business. • Full address or private telephone number; no post office boxes. • Items related to electronics only. Software only when related to Elektor computer systems. • Maximum length: 114 characters — letter, numeral, comma, space, etc. (not including address and/or telephone number). • One advertisement per reader per month. To enforce this rule, a switchboard voucher will be printed each month. • Elektor cannot accept responsibility for any correspondence or transaction as a "esult of a 'switchboard' ad, nor as a result of any inaccuracy in the text. • Ads will be placed in the order in which they are received. • We reserve the right to refuse advertise- ments, without returning them. Name and address: Send to: Elektor switchboard, Elektor House 10 Longport Canterbury CT1 2BR All advertisements must include the voucher printed here. They must be post -marked within the month indicated. 12-64 — elektor december 1982 market Sealed switch EECO have introduced the environmentally sealed MINI-DIP switch. This model features a flexible, clear cover and an epoxy bottom seal. The switch can be actuated directly through the flexible cover. This configuration offers a maximum protection during the wave soldering and cleaning processes, as well as when used in hostile environmental con- ditions. The seal will maintain its integrity when subjected to a range of temperature and solvents. The 2400 series MINI-DIP offers standard gold-on-gold true wiping action contacts and strong .508 mm x .355 mm terminals to vir- tually eliminate insertion and alignment prob- lems. Standard 2.54 mm x 7.62 mm centres make MINI-DIP compatible with other major brands. EECO Sales, Telephone: 0954.8025 (2477 M) Small instrument enclosures The K-series of instrument housings is now available from Alusett UK Limited, based in Winchester, Hampshire. These enclosures comprise front and rear satin anodised panels that attach to a tubular blue powder -dipped aluminium cover using brass fixing bushes. Six sizes, from 90 x 42 x 100 mm to 138 x 66x200 mm, are available and separate chassis are offered which can be secured to the front and rear panels using 3.5 mm screws. Special sizes, configurations and colours, can be supplied by Alusett to suit non-standard requirements. Alusett UK Ltd., Whiteshute Lane, Winchester Telephone: 0962.68673 (2509 M) Alphanumeric keypads A range of cost-effective keypads, formed from Shin-etsu conductive silicon rubber is now available from N.S.F. Advantages of this range include . . . • extended operational life • reduced assembly costs • one-piece construction • elimination of spring and plunger assembly • low chatter and bounce • impervious to many contaminants • non-discolouration of pads Most conventional keyboards for computer data entry incorporate metal contacts and mechanical actuation. Dirt, dust and other contaminants will, of course, render this type of construction liable to failure should they get between the contacts. Also, economically, such keyboards tend to be costly due to the use of gold plated contacts and to a more labour intensive manufacturing and assembly system. In comparison, the new keyboards are formed with conductive rubber instead of metal for the key contacts and silicon rubber domed membranes instead of mechanical switches. N.S.F. Limited, Switches and controls, Keighley, West Yorkshire BD21 5EF Telephone: 0535.61 1 44 (2508 M) BCD switches The model CBS switches from N.S.F. are available in both binary and binary coded complement outputs. They are provided with various terminations including pcb and wire- wrap. The moulded indexing mechanism is fully enclosed giving effective protection for the rotor connections to the printed circuit tracks. Four indexing configurations are available: 12 positions at 30° 16 positions at 22.5° 24 positions at 15 32 positions at 1 1 .25 Stops can be fitted as specified or made adjustable for customer adjustment. The model CBS switch has a switching ca- pacity of 3 VA with a resistive load, switches a maximum current of 100 mA, a voltage of 60 V rms and has an insulation resistance of not less than 50,000 megohms. N.S.F. Limited, Keighley, Yorkshire BD21 5EF Telephone: 0535. 61144 Subminiature binary coded wafer switch The CLSA from NSF is a small, moulded, binary coded, single wafer, rotary switch for pcb mounting. The compact and low profile of this switch, manufactured to the EBE system, accommodates both the index mech- anism and the switching contacts in a single moulding, 18.6 mm dia. and only 10 mm deep, and also has the position numbers moulded into the face. Model CLSA has a switching capacity of 3 VA max. with a resistive load, carries 1 A and has an insulation resistance of 10,000 megohms. Contacts incorporate a gold flashed finish. N.S.F. Limited, Keighly, Yorkshire, BD21.5EF. Telephone: 0535.61 144 (2480 M) market elektor december 1982 — 12-65 This unit replaces conventional mechanical thermostats without additional wiring and it has an additional advantage; the hysteresis may be programmed at choice. It is also possible to control the unit manually without disturbing any of the pre-selected-programs. The main savings are obtained by a more accurate measuring of time and temperature and precise ON and OFF switching, elim- inating mechanical tolerances. The unit is built around a specially designed 'single chip microcomputer' and based on the most advanced technologies. It is available in kit-form and also as a built and tested unit. VeHeman (U.K.) Limited, P.O. Box 30, St. Leonards-onSea, East Sussex TN37 7NL Telephone: 0424.753246 (2507 M) mm m m Voice response unit The VR8 Voice Response Unit provides clear, human speech output from existing com- puters over the dialled telephone network. The VR8 automatically accepts incoming calls from telephones with multi-frequency tones or standard telephones with pocket sized 'Minitone' keypad attached, and passes the data to a computer in the normal manner. The speech facility is activated by commands from a computer and the Unit's own micro- computer then delivers speech down the telephone line. Although the system is expected to operate mainly over the telephone, the unit may also be used to provide speech in conjunction with or as an alternative to computer output, normally associated with Video displays, where speech may be more convenient than the printed word. This device provides an inexpensive, simple to use method of communicating with a com- puter for the non technical user. The Voice Response Unit can easily be added to virtually any business computer and is compatible with View Data Systems. Speech which may be recorded by the user is held in ROM. Other manufacturer's speech systems, such as National Semiconductors 'Digitalker' with large vocabulary, are standard options on the VR8. For users who require a lengthy or constantly changing vocabulary there is the highly sophis- ticated, multi track, cassette tape option, the VR5. This unit enables the user to change or add recordings by using a conventional plug in microphone. The VR5 has the unique ability to search for speech during data entry and of having more than one recording at a given location which, together with its high search speed, virtually eliminates response delays. Applications responding to the same computer program but in more than one language are also within the capabilities of this unit. The VR5 and VR7 may be com- bined providing both tape and solid state memory in a single unit. The Unit is available with a standard RS 232C interface and both IBM and ICL interactive protocols are also catered for. Medway Data Limited, Victoria House, Grover Street, Tunbridge Wells, Kent TNI 2QB Telephone: 0892.44462 (2502 M) Heating controller This unit is designed to control the tempera- tures inside buildings enabling central heating systems (oil, gas, electricity) to work more economically and therefore save direct energy. It provides a 4 program daily cycle condol- ing the temperature at any given period. These programs are totally independent and therefore it is possible to select day and night temperatures separately as desired. The dis- play functions as a clock as well as a ther- mometer. The different programs may be called, verified and changed individually. 12-66 — elektor december 1982 market High-specification, low-cost test equipment Sifam Ltd of Torquay is now marketing a range of test instruments and accessories. Available now (all with one-year guaran- tee, prices exclusive of VAT) are: Hand- held 3V& digit (1999) multimeter, model DMM2200B, priced at £ 43.43, offers 21 ranges in five modes: DC & AC voltage and current, and resistance. Said to have a basic accuracy of 0.3 per cent (DCV), the instru- ment will operate continuously for 1 000 hours from a standard 9 V radio-type battery, and has overload protection, autozero and auto- polarity facilities as well as over-range and low-battery indications. Robustly constructed, two rotary switches provide clear and simple colour-coded selec- tion of measurement mode and range, and the read-out is by 12 mm high LCD digits. It measures 165 x 110 x43 mm, weighs 360 grammes with battery, and is supplied complete with test leads, spare fuse, battery and operators' manual. 3% digit (1999) bench multimeter, model DMM2500, priced at £66.04: 24 ranges in the same five DC/AC measurement modes, with the same order of accuracy and oper- ational features, but with push-button func- tion/range switching and 2000 hours battery life and circuit-breaker overload protec- tion. It has a built-in bench stand/handle and measures 1 55 x 1 20 x 57 mm, weighs 383 grammes with battery. It is supplied with test leads, battery and manual. Digital logic probe, model DLP50, compatible with DTL, TTL and CMOS standards, this too, says Sifam, offers exceptional value for money at £ 39.09. It has an input frequency range of DC to 50 MHz, a minimum detect- Epson EA series LCD modules are available with character formats ranging from 16x1 line to 80 x 4 lines and a dot format of 84 columns x 32 rows. Operating on a 5 V DC power supply, with a consumption of 0.6 to 3 ma, EA and EG LCDs have a high contrast display for easy readability. Backlighting can be provided. EA and EG series modules have a built-in Data RAM (80 Bytes) and character generator (160 ch based on ASCII code). The modules will operate at any temperature from 0 to 50°C and have a response time (turn on/turn off) of 150 ms. The design life is 50,000 hours. Datac Limited, Tudor Road, Altrincham, Cheshire WA 14 5TN. Telephone: 061 .941 .2361 /2. (2521 M) able input pulse width of 10 nanosecs, high input impedance of 1 0 megohms, power range of 4.5 to 30 V DC with input protection (including an audible warning) up to ± 1 20 V DC or AC. Three-colour LEDs signal: high (red), low (green), open-circuit/bad level (yellow) and pulse/memory (red). The audible alarm sounds if an input signal exceeds the operating voltage of the circuit under test, or when a voltage in excess of 30 V DC is applied to the probe input, if the power lead is connected in reverse or with AC line. The probe is fitted with a 800 mm long power lead, and has a consumption of 50 mA maxi- mum at 5 V DC. Dimensions of the probe itself are 195 x 26x 16 mm and it weighs 70 grammes. Supplied in a moulded carrying case, it comes complete with ground and 1C clip leads and operating manual. REA publicity, 6 St. James Square, Cheltenham, Gloucestershire GL50 3PR Telephone: 0242. 514418 ( 2505 M ) 20 A relay Introduced as the Magnacraft Class 389, this new power relay features extremely high contact ratings for its size: it can be supplied in open style, which measures 2.1 in. long x 1.45 in. wide x 1.28 in. high; or with a clear plastic flange-mount cover, measuring overall 2.9 in. long x 1 .5 in. wide x 1.4 in. high. Its base incorporates quick-connect /solder ter- minals. LCD modules A new range of dot matrix liquid crystal dis- play modules, is now available from Datac Limited. Compact in size and light in weight, The most important feature of the Class 389 is its power switching capability. The relay is available in single, double or triple throw contact configurations rated at 20A per pole. A further version, designated 389 DZ, features double break and double make contacts rated at 30A. Contact horsepower ratings range from 'A hp to VA hp, according to contact configur- ation and the voltage of the motor to be controlled. Coil voltages range from 6 V a.c. to 277 V a.c. or 6 V d.c. to 220 V d.c. Diamond H Controls L td„ Vulcan Road North, Norwich NR6 6 AH. Telephone: 0603.45291/9. (2514 M) 12-68 — elektor december 1982 advertisement A selection of circuits which give \ as much enjoyment in building them as actually playing the games. The cir- cuits are fascinating although the elec- tronics involved are not complex and therefore anyone with a good soldering iron will find this book satisfying. These are electronic games that do not need a TV screen, and as a result can be played just about anywhere elektor Elektor publishers Ltd. Elektor house 10 Longport Canterbury CT1 2BR advertisement elektor december 1982 — 12-69 new books from elektor 301 circuits The book follows the theme, and is a continuation of our popular and very successful 300 circuits publication. It is composed of 301 assorted circuits ranging from the simple to the more complex designs described and explained in straightforward language. An ideal basis for constructional projects and a comprehensive source of ideas for anyone interested in electronics. In a nutshell something to please everybody. THE NEW MAPLIN CATALOGUE FOR 1983 BRINGS YOU RIGHT UP- TO-DATE IN ELECTRONICS & COMPUTING Nearly 400 pages of all the most useful components and a whole big new section devoted to home computers and personal software. As always the catalogue keeps you up-to-date with the latest technology — even our ordinary miniature resistors are now superb quality 1% tolerance metal film, yet they're still only 2p each. As well as our usual quality products at low prices, now we're offering quantity discounts too. So pick up a copy of our catalogue now — it’s the biggest and the best! See us at the UK's new electronics exhibition — The Electronic Hobbies Fair — at the Alexandra Pavilion from 18th to 21st November. (Special bus from Alexandra Palace BR station and FREE car park in Alexandra Palace parkl). The exhibition covers electronics, computing, amateur radio, CB, practical hi-fi and radio control modelling. ^^Posi this coupon now for your copy of our 198^^ catalogue, price £ 1 .25 + 25p p&p. If you live outside the UK send £1.90 or 10 International Reply Coupons. I enclose £1 .50. Name . . Address V E '2/62 J mpuirn ELECTRONIC 1B SUPPLIES LTD P.0. Box 3, Rayleigh, Essex SS6 8LR. Telephone: Southend (0702) 55291 1 /5541 55 Shops at 159 161 King Strew, Hammersmith. London W6 Tel (01) 748 0926 Lynton Square. Perry Barr. Birmingham Telephone (021) 356 7292 284 London Road. WestcliH on Sea. Essex Tel (0702) 554000 Att shops closed Mondays