nr. 44 december 1978 UK. 50 p U.S.A.|Can.$1.50 elektor up-to-date' electronics for Idb and leisure : Liu* -, : -•*»•« .r- . '!>•'** c ?|1M rt’tf t#r<- .vrui >»*•*«*•. /i ’ j .«.*!• ..u r. ■ I •:>um. , *vu! . ' Jlr -.IlScCK •»)( » « ..»* ‘ •'**“** ’ ••••; - TV typewriter Austria Belgium Denmark S. 33 France Germany Netherlands F. 7 DM. 3.80 DFL. 3.25 Norway Kr. 9 Sweden Kr. 9 Switzerland F. 4.40 F. 55 Kr. 9 elekterminal UK 4 — elektor december 1978 decoder etekar Volume 4 44 Number 12 decoder Editor Deputy editor Technical editors Subscriptions W. van der Horst P. Holmes J. Barendrecht, G.H.K. Dam, E, Krempelsauer, G.H. Nachbar A. Nachtmann, K.S.IVL Walraven Mrs. A- van Meyel international head offices: Elekluur Publishers Ltd, Bourgognestr, 13a Seek (L), Netherlands Tel, 04402-4200 Telex: 56617 Elekt Nt U.K. editorial offices, administration and advertising: Elektor Publishers Ltd,, Elaktor House, 10 Longport Street, Canterbury CT1 TPE, Kent. U.K. Tel,: Canterbury (0227)54430, Telex: 965504. Please make all cheques payable to Elektor Publishers Ltd, at the above address. Bank: 1. Midland Bank Ltd,, Canterbury, A/C no. 1 1014587 Sorting code 40-16-11, Giro no. 3154254. 2. U.S.A. only: Bank of America, c/o World Way Postal Center, P.0, Box 80689, Los Angeles, CA 90080, A/C no. 12350-04207. 3- Canada only: The Royal Bank of Canada, c/o Lockbox 1969, Postal Station A, Toronto, Ontario, M5W 1W9, A/C no. 160-269-7. Assistant Manager : R.G. Knapp Advertising Manager : N.M. Willis ELEKTOR IS PUBLISHED MONTHLY on the third Friday of each month. 1. U.K. and all countries except the U.S.A. and Canada: Cover price £ 0.50. Number 39/40 (July /August), is a double issue, price £ 1, — , Subscriptions for 1978, January to December incl., £ 6.75 (surface mail) or £ 12.00 (air mail) . 2. For the U.S.A. and Canada: Cover price $ 1.50. Number 39/40 (July/August), is a double issue, price $ 3. — , Subscriptions for 1978, January to December inch, $ 18.— (surface mail) or $ 27. — (air mail). AM prices include post & packing. CHANGE OF ADDRESS, Please allow at least six weeks for change of address. Include your old address, enclosing, if possible, an address label from a recent issue. LETTERS SHOULD BE ADDRESSED TO the department concerned: TQ = Technical Queries; ADV = Advertisements: SUB = Subscriptions, ADM “ Administration; ED = Editorial (articles submitted for publication etc.}; EPS = Elektor printed circuit board service. For technical queries, please enclose a stamped, addressed envelope or a self -addressed envelope plus an IRC. THE CIRCUITS PUBLISHED ARE FOR domestic use only. The sub- mission of designs or articles to Elektor implies permission to the publishers to alter and translate the text and design, and to use the contents in other Elektor publications and activities. The publishers cannot guarantee to return any material submitted to them. All drawings, photographs, printed circuit boards and articles published in Elektor are copyright and may not be reproduced or imitated in whole or part without prior written permission of the publishers, PATENT PROTECTION MAY EXIST in respect of circuits, devices, components etc. described in this magazine. The publishers do not accept responsibility for failing to identify such patent or other protection. National ADVERTISING RATES for the English-language edition of Elektor and/or international advertising rates for advertising at the same time in the English, Dutch and German issues are available on request. DISTRIBUTION in U.K.: Seymour Press Ltd., 334 Brixton Road, London SW9 7AG, DISTRIBUTION in CANADA: Gordon and Gotch (Can.) Ltd., 55 York Street, Toronto, Ontario, M5J IS 4, Copyright ©1978 Elektor publishers Ltd — Canterbury. Printed in the UK by Thanet press. MIMIC* QT THE *U&«T igALAU G* tl»tlAATIDNi What is a TUN? What is 10 n? What is the EPS service? What is the TQ service? What is a missing link? Semiconductor types Very often, a large number of equivalent semiconductors exist with different type numbers. For this reason, 'abbreviated' type numbers are used in Elektor wherever possible: • "741 ' stand for M741 , LM741, MC641 , MIC741, RM741 , SN72741 , etc, • TUP' or TUN' (Transistor, Universal, PNPor NPN respeet- iveiyl stand for any low fre- quency silicon transistor that meets the fo Mowing specifi- cations: UCEQ, max 20V 1 C, max 100 mA hfe, min 100 Ptot, max 100 mW fT, min 100 MHz Some TUN'S are: BC1G7, BC108 and BC109 families; 2N3856A, 2 N 3859, 2 N 3860, 2N39G4, 2N3947, 2N41 24. Some TUP's are: 6CT77 and BC178 families: BC1 79 family with the possible exeptton of BC1 59 and BC179; 2N2412, 2N3251, 2N3906, 2N4126, 2N4291. * 'DUS' or 'DUG' (Diode Univer- sal, Silicon or Germanium respectively) stands for any diode that meets the following specifications: DUS DUG UlR, max IF, max IR, max Ptot, max Op, max 25V 100mA 1 jjA 2S0mW 5pF f20\T 36mA 100 pA 250mW lOpF Some 'DUS's are: BA127, BA217, BA218, BA221, BA222, BA317, BA318, BAX13, BAY61, 1N914, 1N4148. Some 'DUG's are: 0A85, QA91 , OA95, AA116. * 'BC1 07B\ 'BC237B', 'BC547B' all refer to the same 'family' of almost identical better -quality silicon transistors. In general, any other member of the same family can be used instead. BC107 (-8, -9) families: BC 107 (-8,-9), BC147 (-8, -9) BC207 (-8, -9), BC237 (-8, -9} BC317 (-8, -9), 8C347 (^B, -9) BC547 (-8, -9), 0C171 (-2. -3), BC182 (-3, -4) p BC382 (-3. 4), SC437 (-8, -91, BC414 0C177 |-8, -9) families: 0C177 (-8, -9), BC157 (-8, -9), BC204 (-5, -6), BC307 (-8, -9), BC320 (-1.-2). BC350 M, 2), BC557 (-8,-9), BC251 (-2, -3), 8C212 (-3, -4), BC512 (-3, 4), BC261 (-2,-3), BC416, Resistor and capacitor values When giving component values, decimal points and large numbers of zeros are avoided wherever possible. The decimal point is usually replaced by one of the following abbreviations: P (pico-J = IQ- n (nano ) - 10 - M (micro-) - 10- 1 m Imilli-i = 10 k (kilo-) = 1 0 3 M (mega-) - 10* G (giga-) = 10 s A few examples; Resistance value 2k7: 2700 H. Resistance value 470: 470 n. Capacitance value 4p7: 4.7 pF H or 0.000 000 000 004 7 F . . . Capacitance value 10n: this is the international way of writing 10,000 pF or .01 pF, since 1 n is 10 " 9 farads or 1000 pF. Resistors are % Watt 5% carbon types, unless otherwise specified. The DC working voltage of capacitors (other than electro- lytics) is normally assumed to be at least 60 V. As a rule of thumb, a safe value is usually approxi- mately twice the DC supply voltage. Test voltages The DC test voltages shown are measured with a 20 kH v instru- ment, unless otherwise specified. U, not V The international letter symbol 'U' for voltage is often used instead of the ambiguous 'V r . 'V' is normally reserved for Volts', For instance: U b -10 V, not V b = 10 V, Mains voltages No mains (power line) voltages are listed in Elektor circuits. It is assumed that our readers know what voltage is standard in their part of the world 1 Readers in countries that use 60 Hz should note that Elektor circuits are designed for 50 Hz operation. This will not normally be a problem; however, in cases where the mains frequency is used for synchronisation some modifi- cation may be required. Technical services to readers • EPS service. Many Elektor articles include a lay-out for a printed circuit board. Some — but not alt — of these boards are avail- able ready-etched and predrifled. The 'EPS print service list' in the current issue always gives a com- plete list of available boards. • Technical queries. Letters with technical queries should be addressed to: Dept. TQ. Please enclose a stamped, self addressed envelope; readers outside U.K, please enclose an IRC instead of stamps, • Missing link. Any important modifications to, additions to, improvements on or correct ons in Elektor circuits are general y listed under the heading M ssing Link' at the earliest oppo^_nity. contents elektor decern be r 1978 — UK 5 A (repetitive) tone-burst signal is an extremely useful aid for testing audio equipment Basically, this type of test signal is obtained by switching the output of a sine- wave oscillator on and off at regular intervals. The tone- burst generator utilises a novel design approach that simplifies the circuit considerably and only involves one minor reduction of the capabilities. p. 12-10 The most convenient and elegant method of displaying data from a micro- processor is on a Visual Display Unit (VDU), When used in conjunction with an ASCII keyboard, a video interface circuit forms a complete video data terminal. The elekterminal can be used with the Elektor SC/MP system or any microprocessor system possessing a serial input/output facility, p. 12-16 Although the basic version of the TV scope is an extremely useful tool when testing low-frequency circuits, the extended version offers vastly superior display possibilities. Any signal within the frequency range of the scope (DC to 100 kHz) can be displayed as a clear and stable trace on the screen of any domestic TV receiver. p. 1 2-34 contents selektor 12-01 PWM audio amplifiers 1 2-04 Th h article takes a look at the principles of ciass-D amplifiers and examines the circuit of a commercial design. singing SC/MP It's possible to make a yP sing a tune i as demonstrated in the SC/MP program described here. With Christmas in the air, a well-known tune was chosen: 'Silent Night'. Programs for several other well-known melodies are also included on an ESS record. 12-08 tone-burst generator 12-10 disco drums (L.witkam) 12-14 A large number of contemporary pop records are aimed at the so-called 'disco scene'. With the aid of the circuit described here, records from the 'pre-disco era' or one's party tapes can be lent an authentic disco 'flavour' by the addition of a synthesised percussion track. elekterminal 12-16 RAM diagnostic (h. Huschitt) 12-25 The more RAM one adds to a juP system, the more difficult it becomes to trace any faults which may occur on one of the memory boards. The following program can be used to test any block of memory larger than Yz K and will indicate the location of any faults which exist. The complete program is also recorded on the Elektor Software Service record ESS-001 . reading Elektor 12-27 cumulative index volume 4 - 1978 12-28 TUP/TUIM /DUS/DUG . . . 12-30 variable fuzz-box 12-31 This super simple circuit can produce a large variety of manually controlled sound effects, by employing signal clipping techniques. TV scope — extended version . , , , 12-34 This month's cover illustrates the close relationship between the ASCII keyboard, published last month, and the Elekterminal described in this issue. Although they are two independent units, each useful in its own right, the combination is equal to a TV typewriter' with a microprocessor-compatible interface. extending the elektor piano 12-51 This article provides readers with details of the changes in component values and of the additions to the filter circuits which will be required to extend the piano with one, two or three octaves. missing link 12-54 market 1 2-54 tdektor eEektor december 1978 — 12-01 sensing unit designed for the precise Major development in telephone engineering A fivefold increase in the capacity of existing telephone lines might be possible within five years as a result of a new speech processing system being developed in Britain, The UK Ministry of Defence says a top military communications expert and a team from the University of Bath’s School of Electronic Engineering in Western England has solved a speech transmission problem that has baffled scientists for years. The result, it is claimed, is a system that transmits speech by wire or radio with greater clarity and efficiency than at present. It will be marketed, it is hoped, at less than a tenth of the cost of existing systems. Patent applications have already been filed and experts are now predicting that worldwide commercial sales could be worth millions of pounds sterling. It is not only the world's telephone systems that stand to gain from the development. Others who could benefit, in the long-term, include deep sea divers and the partially deaf. In the case of divers the system could improve the quality of helium-affected speech, while hearing aids could be made more efficient. Terminal smaller than shoebox The Ministry of Defence says details of how the new system works are classified. But it is known that it reduces speech patterns with the aid of computers to a small number of basic ‘shapes/ It is possible to break down and reconstruct combinations of these shapes like jigsaw puzzles. The system's terminal will be smaller than a shoebox. The military communications expert involved, Brigadier Reg King, did some research into speech processing eight years ago but it was not until he joined forces last year with Professor William Gosling, head of Bath Universi- ty's School of Electronic Engineering, that major progress was made. Brigadier King said: Tn the end, the answer was so simple that I had trouble in convincing myself that it really worked. I had the system checked on a second computer in case we had overlooked some vital detail, but it came up with the same result'. Pr o f e sso r G o sling c omme n t e d : The re is no doubt whatsoever that this represents a revolutionary concept in communications engineering. The mathematicians have always said that it could not be done. I think it was mainly Brigadier King's ability to tackle the problem obliquely and his dogged military persistence in refusing to accept defeat that enabled him to solve the seemingly insoluble'. A prototype of the system now being built should be ready for evaluation in some 3 8 months time. It is said that a fully engineered version could be in general use within the next five years. (361 S) New Finds A new sm all er-dia meter version of the Ferranti inertial directional surveyor (FINDS)* known as the Model 1 063, has been developed. The outside diameter of the new tool is 1 0 s /b inches (270 mm) which will enable it to be used to survey 1 3 3 /s inches casings down to depths of 8,000 ft Preliminary trials have indicated that the FINDS 1063 tool can produce survey measurements to an accuracy of 6 inches in all three channels (North, East and vertical) at the maximum depth at which it has been used to date — 4,800 ft. The trials have been conducted in association with Shell Expro in the Brent and Dunlin fields. The first experimental surveyor designed by Ferranti had a diameter of 1 7H inches and was used down to depths of 1 ,500 ft. very successfully. The significantly smaller diameter of the new 1063 tool has been achieved by repackaging the inertial sensor assembly and its associated electronic units. This will enable oil and gas wells to be surveyed very accurately over a signifi- cant proportion of their total depth, FINDS - principle of operation FINDS is an adaptation of an inertial navigation of aircraft. The unit has been proven in service over many years and is exceedingly robust. The inertial sensing unit comprises a gyroscopically stabilised platform which is maintained within one minute of arc of a fixed attitude in space regardless of changes in the orientation of the vehicle in which it is being carried. Three precision accelerometers are mounted on the stabilised platform with their axes aligned mutually at right angles to one another. One axis is automatically gyro-compassed to face north. The accelerometers detect accelerations along all three axes and the output signals are integrated twice to derive displacements from the known starting datum. These functions are carried out whilst the tool is in motion and the derived co-ordinates recorded in a semi- conductor store every 0.1 second. On recovery of the tool these results are transferred to a computer store for immediate analysis, the results being produced at specified depth intervals and presented in a chosen format. Battery powered, the entire system is self-contained and can operate for several hours independently of external services. Survey runs, however, are of short duration since the tool can be lowered at the maximum wireline (or sandline) speed and has been tested in free-fall conditions at 1,500 ft /min. It need only be stopped for ten second fixes every one to two minutes and this transit interval can be increased, if required, at the expense of accuracy, Ferrari ti Offsh ore Sys te ms L fd. Ferry Road Edinburgh EH 5 2XS (358 S) selektor 12-02 — el ski or december 197B Telephone with telescribble People will soon be able to speak and write to each other via the telephone at the same time. Philips Telecommunicatie Nederland has developed a system, called the Scribofoon, comprising an electronic note pad and a visual display unit, which can be used to transmit graphic information via telephone lines. The system can be used either by two people, by several people if they are taking part in a telephone conference, and in mo bilop hone communications. A number of pilot projects will be set up in the Netherlands next year (1979) to evaluate the system. Once the Scribofoon has become widely accepted, its cost is expected to be about the same as that of a colour television set. Before that, however, it will cost more because of the small series which will be made. Philips Telecommunicatie Nederland based the Scribofoon’s development on research done at Delft University of Technology where the staff of the Information Transmission Laboratory developed a method of transmitting pictures via radio. The work done by Philips Telecommunicatie Nederland, in cooperation with the Dutch Post Office’s Dr. Neher Laboratory, aimed at applying this method to achieve two- way picture transmission via telephone lines and at developing the equipment. A very small part of the telephone frequency band, the conversation band, is reserved for transmitting the pictures without any unfavourable influence on the telephone conversation. This means that picture and speech can be trans- mitted simultaneously so that one can talk and write via the telephone at the same time. Thus, a telephone conver- sation can be clarified with drawings, sketches and maps for example. When one person is writing, the other cannot do so because the facility is blocked. The Scribofoon consists of a screen and a writing pad. This pad contains printed wiring: vertical wires on top of a plastic layer and horizontal wires underneath it. An electric pulse scans this network of wires. When a pen touches the paper covering the pad it acts as an antenna and signals the moment when the rapidly scanning pulse is maximal, in other words nearest to the pen. The system then knows the pen’s position because the time-lag between the start of the pulse and it being seen by the pen is specific for a certain position on the pad. These positions are then trans- mitted via the telephone line and made visible on the other subscriber’s screen. This transmission is made at normal writing speed. When a group of people are using the system, the graphic information appears on everyone’s screen at the same time. Members of the group can propose additions or corrections because an erase facility is a feature of the Scribofoon. The pen can also be used to point out a detail of the picture without this mark being incorporated into the picture. Both the telephone drawing and the spoken conversation can be recorded on a normal cassette recorder for later reproduction. In the Scribofoon’s application with mobilophones, bodies such as the police and fire brigade could use the system for quickly transmitting a situation sketch, for example. International interest has already been shown in the system, also for its possible uses in education. A pre-recorded, language-teaching cassette could be used to pass on spoken and written information. Another possible use is the transmission of cartoons. Pre-recorded audio cassettes could contain the story on one track and the pictures on the other. Philips Telecommunicatie Nederland’s experimental equipment included oscilloscopes for presenting the pictures. However, television screens are used in the current version. In addition, integrated circuits will be applied in future so that the quipment will then be smaller and cheaper. The main cost factor is the filter in the central unit which goes next to the telephone. Ph Hips Telecommun icatie , Ne th erlands. (393 S) On the trail of interference The Automobile Equipment Group 1 of Robert Bosch GmbH has recently acquired a new weapon with which to hunt down annoying sources of radio interference. A computer-controlled VHF - UHF test assembly supplied by Rohde & Schwarz will be used for investigating interference suppression in cars, car parts of the company's own production — such as ignition systems, electric motors and generators film cameras marketed under the Bauer label and emergency generator units. The main component of the test assembly, the VHF - UHF Test Receiver ESU 2, measures over the frequency range of 25 to 1000 MHz in compliance with the VDE guidelines 0875 to 0877 and 0879, the CISPR Publications 2, 4 and 12 to 14, and European Community directives. The Basic-programmable desktop computer 4051 from Tektronix presents the test results in graphic and tabular form on its large storage screen. Programmed evaluation of the results by the computer is also provided, as is permanent documentation on the Hard Copy Unit 463 1 . This automated test assembly can be used for measurements on useful radio signals as well as on interference, and is suitable for operation in the field, in the laboratory or in a test vehicle. The assembly comprises the VHF - UHF Test Receiver ESU 2, the Frequency Controller EZK and two Code Converters PCW from Rohde & Schwarz, the Graphic Computing System 4051 and Hard Copy Unit 4631 from Tektronix, a digital voltmeter and a frequency counter. For frequencies from 25 to 1000 MHz the receiver has a voltage measurement range of — 1 0 to + 1 20 dB (jiV) and can measure two-port transfer constants from —90 to +40 dB, - * P M j mt ■ %■ W Ml 8 ' >•' : ■ 4: :4: :4^ m Pi i i i * #: • Jill! m . f 5 -« ' ms! rawnift & Wi W ms WM 44444 ■■ 1 B& .S&., M : * ' v M , ■!!■ | , •£; tv ’P' i#.. II jifj-il- rfl f| »T' f ■4: 4' m m : P m m ■ 1 :■ .. & # 4 4 ; 4: 4 ::J p i:4 ♦ v 4? 4? 4- 4 • m- < Pf- fpf- : :: • «« gg ™ w , :S 4 * i|: # *1% |PWg| IP || P* ♦ 9 §P | M # \ 1 I % 4 * -v I# W mm mm w m mm m m m . mmm. H mmem _ ' : :£. | H • *■. j| (379 S) selaktor elektor decern ber 1978 — 12-03 of liquids caused by different Using heat to cool the Sahara A small communications station in the north African desert is constantly heated up by the dissipation of its radio relay amplifiers. And yet even in the extreme midday heat the temperature inside the insulated shelter is about ten degrees below that outside. The cooling effect is produced without pumps and fans, entirely without the use of energy, by me an so f an ingenious ‘rectifier 5 system of air and water cycles devised in the Siemens development laboratories. Working roughly on the principle of gravity 'Circulation hot -water heating, the system extracts heat from the interior during the day, stores it in a water tank and discharges it to the environment during the night. In this way the system acts as a ‘refrigerator' during the day. I he cooling system was developed in an effort to protect communications equipment in extreme climates from excessively high temperatures, without additional energy having to be supplied for this purpose. The shelter ( 1 *3 m 3 volume), equipped with heating elements (150 W), fulfilled all expec- tations the very first time it was tested in the climatic chamber. The equally good results obtained in the test under extreme outdoor conditions encouraged Siemens to extend the tests to larger shelters with a volume of about 20 m 3 ♦ The unpowered cooling system is based on the differences in density and weight temperatures. It consists of three cycles. The warm air rising from the equipment in the shelter is first guided by a baffle to heat excharger A where it cools and consequently flows downwards again to the equipment. The second cycle is formed by the water in exchanger A. It heats up, expands and, as a result of its lower specific weight, passes through an ascending pipe to a central tank inside the shelter, while a downpipe with cooler water feeds exchanger A, Heat exchanger B, which links the tank with the exterior, is warmer during the day, so that no cycle is created. At night, however, as soon as the outside temperature drops below the temperature of the tank, the third and most important cycle is started up: the water from exchanger B cools the contents of the tank — via an ascending pipe and a downpipe again — until the densities have become equal or until the ambient temperature rises above that of the tank again. The reservoir cooled in this way is then used for fully automatic cooling of the interior during the day, Siemens A G $ Munchen, W-Germany. (395 S) 'Charged' cars are no danger You occasionally see them: cars dragging a hail" of metal braiding and hard rubber behind them, intended to discharge the static charge of the vehicle. The idea comes from the big tanker trucks, which always drag along a metal chain touching the ground. But does this ‘discharge band 5 really serve any useful purpose? Engineers at the Siemens high-voltage laboratory in Berlin investigated to what extent cars are charged while being driven. Friction between the wheels and the road surface is the main cause for the electrical charge. The rubber-tired vehicle is electrically insulated as it rolls on the roadway; it is - from a physical point of view — a capacitor with a capacitance of about 100 pF, When the weather is dry and the insulation resistance of the tires is thus sufficiently high the ‘autocapacitor' will be charged to approximately 10,000 V. The energy stored equals about 0.005 Ws. This amount of energy is so small that it can be discharged via the human body without causing any harm whatsoever. If a ‘charged 5 car is touched with the hand, the energy will almost completely discharge in about a microsesond. In a darkened room this discharge could be perceived as very thin blue sparks, but in daylight or with street lighting it rema ins invisible. The whole effect is thus similar to the static charge experienced when one walks on synthetic carpets: it is a nuisance, but completely harmless. The electrical voltage between the car body and the road surface can always be measured in terms of tens of thousands of volts even at high speeds since pointed or sharp-edged parts of the car body provide for local discharging and prevent higher voltages from being attained. It is questionable w T hether conductive bands can prevent cars from becoming charged. The vehicle can only be charged when both the road and tires are dry — in the rain a good electrical discharge is always available. A dry road surface acts as a high electrical resistance, making it difficult to bring the charge from the car to the ground. For the owner of a passenger car there is really only one piece of advice: since it is impossible to protect oneself against spark discharges resulting from static charging it is better to accept them without a fuss. If one is prepared for the tingling sensation it is easier to take. People are not endangered by the charge on a car — regardless of whether or not one attaches a ‘lightning conductor 5 to the car or not. Siemens engineers once again advise motorists caught in a thunderstorm to drive to the nearest parking space and remain in the car. Since the vehicle is a closed Faraday cage passengers are best protected against lightning inside the car. Siemens AG, Munchen, W-Germany. (394 SI 12-04 — elektor decern ber 1978 PWM audio amplitii PWM audio amplifiers Although the concept of the pulse-width modulated or 'Class-D' audio amplifier has been known since the late nineteen- forties, there has been little practical exploitation of the technique apart from a few short-lived designs by Sinclair and Mu Hard in the nineteen- sixties. However, recent advances in semiconductor technology, power FETs in particular, have made the class-D amplifier a more feasible proposition, and the PWM concept has been resurrected by some Japanese hi-fi firms. This article takes a look at the principles of class-D amplifiers and examines the circuit of a commercial design. The problem with conventional audio amplifier particularly when called upon to deliver large amounts of power, is that they are inefficient. This is a direct consequence of the fact that they operate in a linear fashion, be. the output voltage at any instant is equal to the input voltage multiplied by a constant - the voltage gain of the amplifier, Since the output voltage of an amplifier must be derived from the supply voltage via the output transis- tors) it follows that the difference between the output voltage and the supply voltage must be dropped across one of the output transistors, resulting in wasteful dissipation of energy as heat in the output transistors. The output transistors in a linear amplifier can be likened to variable resistors whose resistance is adjusted by the input voltage so that the voltage developed across the amplifier load is always equal to the input voltage multiplied by the voltage gain of the amplifier. To take a simple example, suppose an amplifier is delivering a voltage equal to half supply into a load, R t The effective resistance of the output transistor is thus also R, and the same power is dissipated in both the load and the output transistor, representing an efficiency of only 50%. The highest efficiency that a conven- tional class-B amplifier can hope to achieve is obtained when the peak output voltage (neglecting losses due to Table 1 . Output power: (both channels driven simultaneously) Damping factor: Harmonic distortion: IM distortion: (60 Hz/7 kHz; 4:1 ) Frequency response: Signal-to-noise ratio: Input sensitivity: Loudspeaker impedance: Total consumption under full load: transistor saturation voltages) is equal to supply voltage, be. no voltage is dropped across the output transistors on signal peaks. For a sine wave input signal the efficiency achieved at this level is around 70%* Of course, in a practical situation with a music input the amplifier rarely achieves full output and the average efficiency is much lower. Clearly the efficiency of a conventional amplifier is limited because the output transistors are used as ‘rheostats* to regulate the voltage developed across the load. The only way that this can be avoided is to operate the output transis- tors as switches. This means that a transistor would either be turned hard on, passing current but with very little voltage developed across it, or cut off, in which case it would have full supply voltage across it but would be passing little current. In either case very little power is dissipated in a transistor. Ho tv can simple on-off switching of transistors be translated into a continu- ously variable analogue waveform? Consider the circuit of figure i, in which the output transistors are rep- resented as switches. If the switches are opened and closed alternately, then the output will be connected alternately to positive and negative supply. If each switch is closed for an equal length of time then the average output voltage will be zero, since the output spends the same amount of 2 x 160 w 20 (8 n, 1 kHz} <0,5% at < 2 x 160 W < 0,1% a t < 2 x 160 W 5 . . . 40,000 Hz ± 0.5 dB > 1 10 dB (inputs shorted) 1 ,4 V RMS into 50 k 8 ... 16 n 550 W (total efficiency over 58%) Specifications of the TA-N88 etektor decern ber T97B - 12-05 Figure 1. Block diagram of a pulse width modulated amplifier, the basic constituents of which ere: the analogue input signal, a sym- metrical squarewave, a pulse width modulator, a switched output stage and a passive low pass filter. Figure 2, This figure illustrates how the duty- cycle of the squarewave pulse train is varied in accordance with the amplitude of the ana- logue input signal. The diagram should be 'read' from top to bottom, one line at a time. The conversion from amplitude to time (i.e, duration of the pulse] can be clearly seen. time at the positive potential as it does at the negative potential. If the upper switch is closed for a longer time than the lower switch, i.e. the duty-cycle is greater than 50% then the average output voltage will no longer be zero but will be positive. Conversely, if the duty-cycle is less than 50% then the average output voltage will be negative. Since one output transistor is always turned hard on and the other is turned off, there is little power dissipated in the output transistors. These are the essential principles of a PWM amplifier. The output transistors switch the load alternately to the positive and negative supply at a rate much greater than the highest audio frequency. The audio input signal is used to control the duty -cycle of the switching waveform such that the average output voltage is proportional to the audio input voltage. This prin- ciple is illustrated in figure 2, where 23 consecutive cycles of the switching waveform are shown underneath one another. The duty-cycle modulation for a sine wave input is clearly visible. To retrieve the audio signal free from he switching waveform, all that is required is to interpose a lowpass filter between the output and the load. To minimise losses in the filter it must, of course be a passive LC filter con- sisting of low-loss inductors and capacitors. Class-D building blocks So far only two sections of a PWM amplifier, the switched output stage and the output lowpass filter, have been considered. The most important section of a PWM amplifier is the pulse-width modulator, which converts the audio input signal into a variable duty-cycle squarewave. This is not such a compli- cated procedure as it might at first ippear, and the "building blocks’ of a pulse-width modulator are quite standard circuits which should be familiar to many readers. A more detailed block diagram of a PWM amplifier is given in figure 3, The first step is to generate a square- wave with a 50% duty-cycle at the PWM audio amplifiers 12-06 — elektor decern her 1978 switching frequency. This is then fed through an integrator to produce a triangular waveform of the same frequency. The triangle waveform is summed with the audio waveform and the resultant signal is fed to a zero-crossing detector, i.e. an analogue voltage comparator which has one input tied to zero volts. If the audio signal level is zero then the signal fed to the comparator is simply the triangle waveform whose zero- crossing points occur precisely in the centre of the waveform, so the output of the comparator is a squarewave with a 50% duty-cycle. If the audio wave- form is going positive then the triangle waveform is also displaced positive so that it spends more time above zero than it does below. The duty-cycie of the comparator output therefore increases. Conversely, when the audio waveform is swinging negative the triangle waveform is displaced negative and the duty -cycle of the comparator output decreases. The output of the comparator is therefore a pulse train whose duty-cycle is proportional to the audio signal level. This signal is used to control the switched output stage. The distortion level of such a simple PWM amplifier would depend upon the linearity of the triangle signal and the accuracy of the voltage comparator, since the system Is open-loop, he, has no negative feedback. However, it Is not too difficult to construct linear integrators and accurate comparators, so that the open-loop distortion of a PWM amplifier can be quite low. By adding negative feedback the distortion can be reduced still further to ‘hi-fi’ levels. The application of negative feedback involves taking the output signal of the amplifier, inverting it and summing it with the input signal. Any difference between the two, e.g, distortion, pro- duces an error signal which is fed back into the amplifier in such a sense as to correct the error. In a PWM amplifier it might be thought that the feedback signal would be taken from the output side of the lowpass filter. However, this filter is optimised for low energy loss rather than for maximum rejection of the switching frequency, and consequently the output waveform is somewhat ‘spiky’. This has no audible effect, since the switching frequency is too high to be heard, and in any case is filtered out by the mech- anical inertia of the loudspeaker. How- ever, the output signal is not sufficiently clean to be used as the feedback signal. For this reason the feedback signal is picked off before the output filter and is fed back to the input via a precision inverting integrator. So that the feedback signal is summed with a similar input signal the audio input signal is also fed through an identical integrator. This arrangement is shown in figure 4a. It will be seen that three integrators are involved, one for PWM audio amplifiers m / vVWW comparator ©n + S— > switched output stage > 7901 1-3 mbo — 19 F7 08 0C45 C809 ST COUNT 1 32 9D 05 0047 080 A ST COUNT V 5A 04 01 0C49 0601 LD@ 1 (2) 78 6E 03 0048 C804 ST COUNT 2 6A 04 03 0C4O 9000 JMP S 1 60 AF 01 0C4F 00 4 counter bytes BF 93 02 0C50 00 A9 6D 0C 0C51 00 0C52 00 singing SC/MP elektor decern bar 1978 — 12-09 TABLE 2 Note hex code frequency A# C7 466 B 88 494 C A9 523 C# 9C 554 D 8F 587 D# 83 622 E 78 659 F 6D 698 F* 63 740 G 5A 784 G# 51 831 A 49 880 A# 41 932 B 39 988 C 32 1047 C# 2B 1109 D 25 1175 D# IF 1245 E 19 1319 F 14 1397 F# 0F 1480 G 0A 1568 G# 06 1661 A 02 1760 Figure 1, Only three components are required for a 'loudspeaker interface'. Figure 2. In case of availability problems, the SC 517 can, of course, be replaced by a standard Darlington configuration. Table 1, Complete program listing for 'Silent Night 1 , Table 2, Using the information given in this Table, it is a fairly easy matter to program the SC/MP for other melodies. The SC/MP can, of course, sing quite happily to itself. However, if the per* formance is intended for the benefit of a human audience, a loudspeaker will be required. In computer jargon; a loud- speaker interface’, A suitable circuit was described in the recent 'Summer circuits' issue {July/August 1978, circuit no, 12, ‘software Kojak siren*). It consists of a Darlington-transistor amplifier and a loudspeaker, connected to a 'Flag* output of the SC/MP (figure 1 ; an alternative circuit is given in figure 2). If the Flag is set and reset rapidly, a tone is produced; the more rapidly the Flag changes state, the higher the output frequency. A melody consists of a succession of 'tones 1 with different frequencies. To make the SC/MP ‘sing’, it must be programmed to set and reset one of its Flags at a frequency that is determined by a list of numbers somewhere in its memory. Furthermore it must be told, by means of a second list of numbers, how long each note should last. In other words, a program is required that will combine two lists of numbers (one for tone pitch and one for tone duration) to produce a melody. A suitable program is given in Table 1 , The output frequency generated at any given moment is determined by a hexadecimal number XX as follows: f = ^ Hz 556 + 8(XX)ig ’ where XX is limited to the range G The values of C2 and C3 are fairly critical, and two positions are reserved on the p.c, board for each of these capacitors so that the desired value can be approximated fairly accurately by connecting two capacitors in parallel. For instance, if an exact I kHz sine wave is desired, C2 and C3 would have to be 1 6 nF ; this can be obtained by con- necting a 1 5 n and a i n capacitor in parallel. The value of C 1 is not so critical, since the frequency of the clock generator can be set correctly by means of PL The highest frequency obtainable is 20 kHz. The current consumption of the circuit is quite low: 1 2 ... 15 mA. The calibration procedure is extremely simple - only one potentiometer (PI) requires attention. The idea is that the clock frequency must coincide exactly with the centre frequency of the active filter, as otherwise the toneburst will not start and stop at the zero-crossing of the sine wave. The adjustment can be carried out quite easily with the aid of an oscilloscope. Photos 4 and 5 illus- trate two incorrect settings; correct adjustment will produce the result shown in figure 6. M Figure 4. Complete circuit for a tone-burst generator that is more suitable for home construction. Figure 5. Printed circuit board and com- ponent layout for the tone-burst generator (EPS 79017). Photo 1. Tone burst, 1 sinewave period with a 16-period gap. Photo 2. Tone burst, 16 sinewave periods with a 1 -period gap. Photo 8. Tone burst, 16 sinewave periods with a 16-period gap. Photos 4 and 5. Clock signal and tone-burst with PI incorrectly adjusted: the burst does not start and stop in the zero -crossings of the sinewave. Photo 6. Tone-burst output after PI has been accurately adjusted. 1 l ■ I tone-bunt generator elektor decern ber 1978 — 12-13 Parts list Resistors: Capacitors: Semiconductors: R1 = 39 k Cl * = 33 n IC1 . . . IC4 = CD4015 R2.R6* = 8k2 C2a/b*,C3a/b* = 15 n + 1 n IC5 = CD 4049 R3,R8 . . . R14= 10 k C4 = 22 p/16 V IC6 = CD 401 1 R4 = 1 M C5‘ = 10 m/25 V IC7.IC8 = 741 R5 = 22 k IC9 = CD 4066 R7 = 470 k R15.R16 = 27 k Miscellaneous: R17* - 100 k S1,S2 = single pole, 16-way PI = 10 k preset switch* * see text 12-14 — elaktor decern ber 1978 disco drumi disco drums A large number of contemporary pop records are aimed at the so-called 'disco scene'. A distinctive feature of most disco music is a heavy, repetitive and fairly complex percussive sound, which often has a characteristically artificial or 'funky' quality. With the aid of the circuit described here, records from the 'pre-disco era' or one's party tapes can be lent this authentic disco 'flavour' by the addition of a synthesised percussion track. L. Witkam It goes without saying that the ‘drumbox’ must have some means of detecting the beat of the piece of music being played, in order that the per- cussion be in time with the music. The simplest way of doing this is to utilise the fact that, almost invariably, the bass line in pop music lays down the beat for the rest of the instruments. Further- more, the bass part tends to heavily accentuate the basic beat, making it easy to detect and follow. Thus the circuit operates by monitoring the level of the low frequency signal com- ponents and adding in the desired disco sounds at the appropriate points. The drum sound is produced by a noise signal with a typically percussive attack-decay envelope. That is to say, initially the amplitude of the noise signal rises sharply to its peak value, then is made to die away more slowly (exponentially). The block diagram of the drumbox is shown in figure 1. As can be seen, the music signal is first fed to a lowpass filter which eliminates all but the bass frequencies. The resulting signal is then rectified and used to control an opto- coupler. When the latter is actuated, a noise signal is fed to the summing circuit, where it is mixed with the original music signal. Circuit diagram The complete circuit diagram of the unit is shown in figure 2. The lowpass filter is formed by T1 and T2. The filter slope is 18 dB per octave (i.e. the filter is third-order) and the turnover fre- quency is 40 Hz. The filter output is rectified by diodes D1 and D2, and smoothed by capacitor C7. This capaci- tor, which has a very short charge time, stores the instantaneous value of the bass part of the music signal. Thus when the voltage across C7 is sufficiently large, T3 is turned on. C7 then dis- charges via R7 and the base-emitter junction of T3, until, after a short time, this transistor is once more turned off. However, during the time that T3 remains turned on both LED D3 and the opto-coupler LED are lit, thereby turning on the opto-coupler transistor. This means that the noise signal, which is applied to the base of this transistor, is fed from its emitter to T4, where it is amplified before being mixed with the original music signal via R17. The noise signal itself is produced by transistor T6, which is connected as a zener diode, and is amplified by T5. The volume of the noise signal can be varied by means of P3, whilst PI alters the sensitivity of the circuit (i.e. the level of the bass part at which a ‘drumbeat’ is produced). P2 should be adjusted until a suitable noise signal is obtained. Varying the position of this preset will have some effect upon the timbre of the noise; one therefore has a certain element of choice regarding the type of sound produced. The noise level will also vary considerably depending upon disco drums dsktor decern ber 1978 — 12-15 Figure 1. Block diagram of the 'disco drumbox'. Whenever the circuit detects a pronounced bass note in the passage of music, it adds the desired percussive effect in the form of a noise signal. Figure 2* Complete circuit diagram. The operation of the circuit can be checked by testing the voltages at the points indicated. The values given in the table apply to quiescent conditions. the sample of transistor used. Thus it is recommended that the noise generator be built first and tested using different transistors. Construction of the circuit should not present any special problems, A T1 1 1 or TU2 from Texas Instruments can foe used for the opto-coupler. Both types are available in a six-pin DIL package. The pinout details given in the circuit diagram apply to both versions. The quiescent current consumption of the circuit is roughly 3 mA; maximum consumption (during each 'foe at 1 ) is approximately 40 mA, which means that, if desired, the unit can be powered by battery. The operation of the circuit can be checked with the aid of the measurement points and test voltages indicated in the circuit diagram. Although in theory the unit can be inserted at any point in the (pre-) amplifier chain, it is recommended that it be placed before the volume control of the amplifier. The reason for this is that otherwise it would foe necessary to alter the setting of P 1 every time one adjusted the volume control, W T1 ... T6 = BC 5498 IC1 = TIL 111, TIL 1.12 (T> 0,8 V 6.3 V 9,0 V (&■ 0 V 5,6 V <£► 5,6 V ©9 V 550mA 73001 - 2 12-16 — elektor decamber 1978 elekterminai elekteminal Low cost video terminal for juP/TV typewriter applications There is no doubt that by far the most convenient and elegant method of displaying data from a microprocessor is on a Visual Display Unit (VDU). When used in conjunction with an ASCII keyboard (such as that described in fast month's issue of Elektor) the video interface circuit described here forms a complete video data terminal which can be used with the Elektor SC/MP system or any microprocessor system possessing a serial input/output facility. The video terminal described in this article is of the serial (i,e, non memory- mapped) type, in which the video RAM used to store the characters to be generated on the screen is not shared by the microprocessor. There are several advantages of this type of system: firstly the terminal can be used indepen- dently (he. is not tied to a micro- processor) as a ‘TV typewriter’. Sec- ondly, the unit is TTY compatible, and in conjunction with a MODEM, can be employed to transmit /receive data over the telephone line. Thirdly, since most microprocessor systems already possess serial input/output routines, it means that the terminal can be used with the vast majority of different juFs and that the necessary device driver software is for the most part already present. The Elekterminal uses one of the new single-chip CRT controllers, the SF,F 96364 from Thomson-CSF (Sescosem), Due in part to the large number of functions assumed by this one chip, the complete video interface for the terminal uses only 21 ICs, is accommodated on a board little larger than Euro card format, yet offers the following comprehensive features: — 1024 characters per page, formatted as 16 lines x 64 characters — plug-in option allow r s character memory to be expanded to 1 6 pages — choice of six different Baud rates: 7$, 110, 150, 300, 600 or 1200 — programmable serial interface charac- teristics: i.e. choice of 6- or 7- bit ASCII code, even, odd or no-parity, 1- or 2 stop bits generation. — choice of TTL or RS232C voltage levels elekterminal elektor decern bar 1976 — 12-17 — normal (white on grey) or inverted (black on grey) video signal - sophisticated cursor control and screen scrolling functions provided in hardware Block diagram The block diagram of the Elekterminal is shown in figure 1 . A detailed descrip- tion of the ASCII keyboard was con- tained in last month’s issue, thus the remainder of this article is devoted to the video interface card proper, The ASCII output of the keyboard is fed directly to the UART. UART stands for Universal Asynchronous Receiver/ Transmitter. This is an LSI IC which accepts data* whether in serial or parallel form, from a peripheral device (keyboard, modem) and transmits this data with appropriate serial-parallel or parallel-serial conversion to the CPU or video interface. Basically the UART allows the keyboard* VDU and CPU to communicate with one another. A more detailed description of this important IC is contained later in the article. To be able to transmit data at different speeds, a programmable Baud rate generator is required. As was explained in the article on the cassette interface (Elektor 36), the Baud rate is defined as the total number of bits including control bits such as stop and parity bits which are transmitted in one second. The programmable Baud rate generator generates a number of frequencies which are 16 x the desired Baud rate. The output frequencies are derived by dividing down a clock input obtained from the crystal oscillator of the CRTC, The CRTC is without doubt the most important component in the entire circuit. CRTC stands for Cathode Ray Tube Controller, however it might be more accurate to describe the chip as a video processor*. However one calls it, the device is another LSI IC which performs a wide variety of control functions with a minimum of peripheral hardware. In the past, video interface cards required a veritable mountain of discrete logic ICs to perform the tasks which are now assumed by this single chip. Among other things the CRTC generates the line and field sync pulses of the video signal, is responsible for the addressing of page memory, and con- trols the character generator. The chip also provides cursor control and screen scrolling in hardware. Like the UART, a more detailed discussion of the CRTC is contained later in the article. The page memory, which holds the data to be displayed on the screen, is formed by a number of static RAMs. The entire memory is scanned once every frame (20 ms). The ASCII code which is stored in page memory is converted into a form suitable for display with the aid of a character generator; after parallel- serial conversion it is then added to the horizontal and vertical sync pulses in the video combiner. The latter provides a 5 Vpp signal which can be fed direct to a video monitor or else via a VHF/UHF modulator (such as that contained in the October issue of Elektor this year) to the aerial input of a domestic TV receiver. The only section of the circuit which remains to be discussed is the CTL decoder. This is basically a ROM which decodes the ASCII character trans- mitted by the UART and informs the CRTC whether it is a control signal (non-printing code) or a character to be displayed on the screen. UART, character generator and CRTC Since many readers may be unfamiliar with these important devices, it is worth while taking a closer look at just exactly how they operate, UART The block diagram of the AY-5-1013 UART which is used in the Elekterminal is shown in figure 2. In fact the UART can be thought of as two ICs (a trans- mitter and a receiver) which are housed in the same package and which combine certain functions in order to economise on the number of terminal pins. The UART is basically a device which pro- vides asynchronous control of data communications, that is to say it is capable of both receiving and transmit- ting data at different rates, as well as performing parallel- serial and serial- parallel conversions, adding or deleting the necessary control and error detecting bits as required. Figure 1. Block diagram of the Elekterminal. Figure 2. Simplified internal block diagram of the UART. m 2 > zz r m Ui > H 1 C « in - m 12 tr > *-( r r eft 3 8 a 05 o x o H * m z > cn 30 m m X W -I fn m H 33 Z > n m I z > □ m 30 O W H ? C > b r- o > H 2 c 171 Figure 3, Characters are generated by means of an S x 5 dot matrix. The top row of dots remains permanently extinguished. elekterminal 12-18 — elektor december 1978 The transmitter and receiver have independent reference clock inputs, which determine the rate at which the data flow occurs, the UART can thus be used for both code- and speed conver- sion. The Baud rates are determined by the output frequency of a programmable Baud rate generator contained on-chip* The output frequency of the Baud Generator is 16 x the Baud rate. Data are fed to the transmitter section of the UART (e.g, the ASCII output of the keyboard) in parallel form. The UART converts the parallel code into a serial data stream, adding the necessary start-, stop- and (if desired) parity bits. The system user can program the format of the transmitted or received serial character to suit his own requirements* That is to say that he has the choice of one or two stop bits, odd, even or no parity bit, and of selecting a 5, 6* 7 or 8-bit data word. The receiver section of the IC does exactly the opposite of the transmitter, he. it deletes the start and stop bits from the received serial charac- ter, checks for parity errors (which are flagged by setting the parity-error output), and presents the data in parallel form at the data outputs. When using the UART as a Baud rate- or code converter the data at the receiver out- puts are fed to the data inputs of the transmitter; code conversion is per- formed by means of a ROM decoder connected between the two sections. Character generator Less complex than the UART, but just as important is the character generator. This 1C is responsible for translating the ASCII code stored in the video RAM into a format which can be used to generate the equivalent alphanumeric characters on the screen. In general the characters are formed by means of a dot matrix, the most common types of which are 5 x 7 and 7x9. Both types have their advantages and disadvantages. Because of the greater resolution available the 7x9 matrix produces characters which are more attractive and have greater detail. However the larger number of dots in the matrix requires a corresponding increase in the bandwidth of the video signal. With 64 characters per line this bandwidth is several Megahertz too large for conventional TV receivers, and results in poor picture definition. For this reason the 7x9 matrix is generally reserved for use with video monitors* Although the characters generated by a 5x7 matrix are somewhat simpler, it is still possible to obtain excellent definition using a normal domestic TV receiver which has a video input. Even with the unavoidable picture degra- dation caused by a VHF/UHF modu- lator, the legibility of the resulting display is still perfectly satisfactory. The format of the matrix produced by the character generator is illustrated in figure 3. The information contained in 4 Vcc GND r PT CO Cl C2 GfJD V cc *1 INI Q ( Qq SYNC 9966 5 each row is stored in a ROM which is addressed by the 6-bit ASCII code (stored in the video RAM) together with a 3-bit row address which is supplied by control logic in the IC* Figure 4 shows the simplified internal block diagram of the character gener- ator. The total of nine address bits allows up to 512 different 5-bit rows to be selected; since 8 are required to form one complete character, the total number of characters available is 64, Depending upon the ASCII code, the correct data word for each row address is put on the five data outputs. With the aid of the output inhibit pin the data outputs can be placed in the high impedance state (tri-state mode), I thereby allowing two or more character Figure 4* The character generator is basically nothing more than a specially programmed ROM* The only difference between it and a normal ROM is die shorter word length of 5 bits. Figure 5* This simplified block diagram of the CRTC illustrates the large number of functions performed by the one 1C. Table 1. Depending upon the slate of the control inputs, CO, Ct and C2 r the SF.F 96364 will execute the following control functions. Table 2. An overview of the division factors required to obtain the various Baud rate clock frequencies. Rounding up the figures fisted in table 2a it is possible to obtain a low-cost Baud rate generator which is still accurate to within 1% (table 2b], elekferminal elektor december 1978 — 12-19 Table 1. C2 Cj Co Execution time ms Page erase and cursor home (top-left) 0 0 0 132 End of line erase and cursor return {at left) 0 0 1 8.3 Line feed (cursor down) 0 1 0 8,3 Inhibition of the character sent 0 1 1 8,3 Cursor left {one position) 1 0 0 8,3 Erasure of cursor-line 1 0 1 8,3 Cursor up {one position) 1 1 0 8,3 Normal character 1 1 1 8.3 Table 2a Baud rate f UART division factor division factor 1000MHz 1008 MHz 75 1 200 Hz 833,33 840 110 1760 Hz 568.18 572.73 150 2400 Hz 416,67 420 300 4800 Hz 208.33 210 600 9600 Hz 104.17 105 1200 19200 Hz 52.08 52.50 Table 2b Comparison of division factors Baud rate 1 MHz 1008 MHz 75 64 x 13 64 x 13 (+8) 110 44 x 13 (-4) 44 x 13 150 32 x 1 3 32 x 13 (+4) 300 16 x 13 16 x 13 (+2) 600 8x13 8 x 13 (+1) 1200 4x13 4x13 generators to be connected in parallel so as to provide the remaining 64 ASCII characters, lower case letters and special symbols. CRTC Almost all the major microprocessor manufacturers have already brought out a CRTC or are in the process of doing >o. Most CRT Controllers are designed ro be used in conjunction with a micro- processor, and some are even tied to a particular family of processor. The device used here, however, is an exception to this rule, and the video rterface card, of which it is the heart, :m be used to form an independent YDU/TV typewriter as well as an : utput terminal for any microprocessor which has a serial output. The device in question is the SF.F 96364 from Thoms on-CSF, which, as we shall see, provides all the control and timing signals required for screen display, as well as providing a number of sophisti- cated screen management facilities (cursor control and screen scrolling etc.) A simplified block diagram of the SF.F 96364 is shown in figure 5, One of the most important tasks of the CRTC is to generate the sync pulses required to display a video signal. With the aid of a simple on-chip crystal oscillator the SF.F 96364 provides a close approximation to the CCIR standard sync signal. Line and field sync pulses are both combined in the one sync waveform. The sync generator also drives the display counter which is responsible for the addressing of the character generator (he, providing the correct row address) and of the video RAM (page memory). In addition, the display counter provides information for the cursor- and page-end compara- tors, The cursor comparator supplies a signal which ensures that the cursor appears at the correct point on the screen. The page-end comparator allows the amount of addressable video RAM to be extended to include extra pages, since the RS output is used to enable the VDU to ‘turn a page 7 in mid -screen. The RP output clocks the counter used to address the additional pages of character memory. A detailed explanation of how the video RAM is extended will be contained in a subsequent article describing an add-on circuit which will permit the use of up to 1 6 pages of memory. The above mentioned functions of the CRTC are of course indispensable, however the most important features of the device are almost certainly the screen formatting control functions which are available in hardware. Many less sophisticated video interface cards rely upon software routines to provide control of cursor and screen scrolling, which means that they are necessarily tied to a microprocessor. The SF.F 96364, however, can provide fairly sophisticated screen control options on-chip, allowing it to function indepen- dently. As was stated in the description of the block diagram in figure 1, the CTL decoder provides a 3-bit instruction code which informs the CRTC that the ASCII character transmitted by the UART is in fact a non-printable control character. Depending upon the code applied to pins Co, C i and C 2 » the CRTC will perform one of the cursor control functions listed in table 1. Certain control instructions take a comparatively long time to execute, since they have to be carried out during blanking intervals so as to minimise display distortion. As will be explained later, the number of control functions can be extended by manipulating the W-( Write- )signal; this facility is exploited in the Elekterminal. Circuit The 21 ICs and associated components which are shown in figure 6 form the complete circuit of the Eiekterminal, All that is required to render the terminal operational is the addition of an ASCII encoded keyboard. Page memory, which holds the ASCII version of the characters to be displayed on the screen, is 6 bits wide and is formed by six 2102A4 IKxl RAMs. The 4 4 7 in the type number indicates the access time of the device, which in this case is 450 ns. If character memory is to be expanded to several pages, it may well be w r orth investing in low power 12-20 — elektor dec ember 1978 el ek terming elektermina! elektor decern ber 1978 — 12-21 Table 3. PIN LEVEL TRANSMITTED OR RECEIVED FORMAT i m 35 \m ■mMm ■ : yy ; mu i ■ mj 11 iNlI r 39 1 No parity bit 0 & "" f 1 " 1 T ran smitted par i ty b I t Cty&n rt&ritv Hr” 0 Odd parity v*:-- *:•& 36 T.' ,'Z' f ZZv 7 m m m m 1 2 stop bits 1 '*• iWi -•* ••• M Mi •* 8 0 1 stop bit 37 38 0 0 5 bits/character 37 38 0 1 6 bits/character 37 km- i § i 1 § It & # 7 bits/character § 37 38 1 1 8 bits/character Table 4* Function Key corresponds to line-feed LF CTL J carriage-return + CR CTL M erase to end of line cursor up VT CTL K cursor down LF CTL J cursor left BS CTL H cursor right HI CTL 1 home cursor FS home cursor + FF CTL L page erase scroll up ESC CTL i {cursor down) carriage return -- (GS) CTL ] (no erasure) erase current line — (SUB) CTL Z memories (2102AL4), since they can lead to a saving in current consumption of up to roughly 30%. Since each character is formed by 8 rows of 5 bits, the ASCII code stored in the page memory is read out 8 x every frame. With 64 characters per line, the memory is scanned in blocks of 64 words, IC10, he. the CRTC, ensures that the same block is scanned 8 times in succession, and that the character generator is provided with the correct row addresses. The outputs of the memory are not connected directly to the character generator, but to an intermediate latch (IC9). The memory address can thus remain one step ahead of the position on-screen, which means it has ample time to set up the following ASCII code on its outputs. The 5- bit parallel ‘row' data from the character generator is fed to a shift register (IC12), where it is converted into serial form and thus becomes suitable for video display. This shift register is driven by a ‘dot-clock' with a frequency of approx, 1 1 MHz, The dot-clock generator is formed by N17, N18 and N26. Since all eight rows of a character must be positioned directly beneath one another, the dot-clock generator is synchronised by the CRTC. This is achieved with the aid of the INI line (see figure 5) which goes low after the 64th character, thereby stopping the dot-clock generator until the follow- ing line sync pulse. All memory addressing is clocked by the dot-clock, since the ‘character- clock 1 , which, via the 1 -input, drives the address counter in the CRTC, is derived from the dot-clock signal via a dmd e-by-eight counter (IC 1 3), The frequency of the dot-clock, which can be varied by means of C2, deter- mines the width of a character: the lower the frequency the wider the character. The minimum usable fre- quency is determined by the available space on the TV screen. If too low a fre- quency is chosen, the characters will run off the edge of the screen. On the other hand, if the frequency is too high, the characters will be compressed into one portion of the screen, with a consequent loss of definition and legibility. Thus it is important that C2 be adjusted so as to obtain an optimal picture on-screen. The spacing between character lines is regulated by the CRTC, which blanks the video signal for the period of four line scans. Between successive character lines there is therefore a space of half a line. Spacing between the actual charac- ters is provided by the shift register, IC12, Since this is an eight-bit shift register and the character generator output is only five bits wide, each character can be preceded by two unmodulated dot columns and followed by one such column. Between each character there is thus a space three dots wide. The fact that, including spaces, each character is in fact eight bits wide explains why the character clock is derived from the dot-clock via a divide- by-eight counter. The serial data stream output by the shift register is also available in its inverted form. One can therefore choose between a positive (white on grey) or negative (black on grey) video signal. The video combiner is built round N22 , , , N25, Depending upon the position of S3, N22 . . , N24 provide a video signal of appropriate polarity, N23 not only inverts the signal at S3, but also inhibits the inverted video signal during the sync period. This causes the current through N25 during the sync pulses to be limited to an acceptable value. The voltage divider network formed by R14 and R15 determines the ratio between the amplitude of the video signal and that of the sync signals. With the values shown the black level is around 30%. Table 5, Address Positive □ 3 O 2 logic 0 , Oq 0 to 1 27 1 0 0 0 128 to 135 0 0 1 1 136 0 1 0 0 137 0 1 1 1 138 1 0 1 0 139 0 1 1 0 140 1 0 0 0 141 1 0 0 1 142 to 153 0 0 1 1 154 1 1 0 1 155 0 0 1 0 156 0 0 0 0 157 0 0 0 1 1 58, 1 59 0 0 1 1 1 60 to 254 1 1 1 1 255 0 0 1 1 Figures, Complete circuit diagram of the video interface for the Elekterminal. With the addition of an ASCII keyboard, which is connected to the K strobe- and KBO . , , KB7 lines, the circuit forms a com- plete video data terminal. Table 3. The programmable serial interface characteristics of the UART. This table applies both to the AY-5-1013 and to the MM 5303. A recommended bit format is shown shaded; this format corresponds to the arrangement of wire links illustrated in figures 6 and 8. Table 4. In addition to the control functions listed in table 1 ( the 4-bit PROM decoder offers several extra possibilities. These extra control functions can be generated either by special keys or with the aid of the control key and one of the data keys. Table 5. The program for the PROM decoder IC7. 12-22 — efektor dec ember 1978 elekterminaf The video combiner is followed by a buffer stage which has an output impedance of 68 £2, and to which one can directly connect a length of coaxial cable. Assuming that the cable is termin- ated with the correct impedance it is possible to use lengths of over 10 m without any problems. The buffer stage has the effect of increasing the black level to around 35%; this can of course be corrected by altering the values of R14 and R15, however the increase in sync signal level does no harm and the adjustment is not really worth the bother. So much for the circuitry which is responsible for actually generating the video signals; there remains the circuit which allows the unit to communicate with peripheral devices such as the CPU and/or keyboard. The most important interface element is of course the UART, the basic operation of which has already been described. As was stated then, the rate at which data is trans- mitted by the UART is determined by a clock signal, whose frequency is 16 x the desired Baud rate. Normally a monolithic Baud rate generator is employed to provide the clock signal, however these ICs are still fairly expens- ive and require a 1 MHz crystal to supply the basic clock frequency from which the fi x 16 5 clock signals are obtained by means of frequency division. An obvious alternative is to make use of the crystal oscillator for the CRTC to provide the necessary clock fre- quencies, This can be done quite simply by first of all amplifying the signal at the output of the oscillator (with the aid of Ml 4), and then feeding it to a programmable divider (IC14 and 1C 15), The accuracy of the clock frequencies obtained thereby is better than 1 %, Generally speaking, the methods adopted for the transmission of data allow a reasonable tolerance in the accuracy of signals, so that a deviation in the region of 1% remains a perfectly acceptable figure. Table 2 lists the relationship between the UART fre- quencies and the (theoretical) division factors (2a) obtained with crystal frequencies of 1 MHz and 1008 kHz, The manufacturer of the SF.F 96364 recommends a crystal frequency of 1 008 kHz in order to avoid interference from the mains frequency. In practice, however, a 1 MHz crystal is perfectly satisfactory. As can be seen from table 2, keeping to an accuracy of 1% and rounding up to the nearest whole figure produces identical division ratios for both crystal frequencies. The result is a Baud rate generator at roughly 20% the price of a monolithic type. Results obtained in practice have failed to indicate that this approach has any untoward effect upon the performance of the circuit. Once provided with suitable clock frequencies, the UART can receive and transmit data at any one of six different Baud rates, which are selected by means of switch S2. An important point about most UARTs is that the output logic levels are often not TTL-compatible. The most common voltage levels used are the so-called RS232C and V24 norms. These two norms, which are virtually identical and more or less interchangeable, have the advantage of minimum signal levels of +5 V (logic M*) and -5 V (logic ‘0’) and maximum levels of + and -15 V respectively. Thus they clearly have a much greater noise immunity than TTL logic levels. The circuit described here attempts to reach a compromise solution by em- ploying a discrete interface design which is compatible with both RS232C/V24 and TTL voltage levels. If the output signal is required to drive TTL loads, then D4 should be included. This diode limits the output voltage to — 0.6 V. Without the diode the output voltage can swing between +5 V and —12 V. The output impedance is deliberately held low in order to facilitate matching with the cable. As was mentioned earlier, the format of elekterminal ale k tor decern ber 1978 — 12-23 Part$ list to figure 6; Resistors: R1 . , . R3 = 4k7 R4 = ms R5 ^ 2k 2 R6 = 560 H R7 = 270 SI R8 . * „ R12 - 1 k R13 = 10 M R 1 4 — 150 n R 1 5 = 390 n R16 “ 220 n R1 7 - 68 H R 1 8 = 390 H R19 = 680 O R20 = 4M7 Capacitors: Cl = 27 pF (see text) C2 - 45 pF trimmer C3 - 10 pF C4 . . , C8= 100 n Semiconductors; D1 . . * D4= DUS T1 = BC 107B, BC 5476 or equ. T2 = BC 1 776 r BC 557 B or equ . T3 - 6F 451 T4 = 2N2219 ICs: IC1 . * . IC6 = 2102-1 ,21G2A4, 2102AL4 IC7 = SFC 71301 E 1-0 (pre- programmed) or equivalent eg 74S387 (to be programmed as shown in table 51 IC8 - AY-5-1 01 3 r MM5303 ICO = 74LS174 I CIO = SF.F 96364 (Sescosem) IC11 = R 0-3-251 3 IC12 = 74LS165 IC1 3 - 74 LSI 63 (C14JC15 = 4024 (C16 = 4011 1017,1021 = 4081 IC18 = 74LS04 IC19 = 74LS00 IC20 = 74 LSI 25 Miscellaneous; 51 = SPST switch 52 = 2 pole 6 way switch 53 = SP changeover switch P«c.b. connectors (female): type ITT-Cannon G09 1 x 22 way (keyboard) 1 x 26 way (extension card) For keyboard cable (male): type ITT-Cannon G09 1 x 22 way XI - 1008 kHz or 1000 kHz crystal Figures 7 and 8 « Track pattern and com- ponent overlay of the p.c.b. for the Elekterminal (EPS 9966 K the serial output /input signal can be programmed by the system user* The number of stop/start bits, choice of parity bit and of data word length can be selected by making the appropriate connections to pins 35 , . , 39, The details are listed in table 3, where a preferred format (7 -bit code with even parity) is shown shaded. If desired the parity bit can be omitted (no parity) since although the UART checks for parity errors in the received signal , the parity-error output is not brought out externally. Thus the parity bit is only of use to the device receiving a character transmitted by the UART, The Elekterminal is capable of being operated in both the full-duplex and half -duplex mode. In a full-duplex system, where the terminal is linked to a t*, the CPU and terminal com- municate in both directions simul- taneously, That is to say that the computer is programmed to echo what is transmitted (from the keyboard via the UART) back to the terminal display. In half-duplex systems the terminal is normally wired up so that the screen 12-24 — elektor decern ber 1978 alektermiPz responds directly to the keyboard. Switching between half- and full-duplex is accomplished with the aid of SI, which is included between the serial output- and input pins of the UART\ The UART relays the ASCII code from the keyboard (or CPU) onto the data bus lines BO . , , B6, where it is picked off by the CRTC and character mem- ory, Before the data reaches the RAMs, however, it is converted from 7- to 6-bit ASCII code; bit 5 is ignored and bit 6 inverted. In addition, gates N1 ...N7 offer the possibility of forcing the 'space 7 code (100000) onto the data inputs of the RAMs, so that, if the appropriate control code is applied to the CRTC, an entire line or the complete screen can be erased. These are in fact only two of the many control functions which the Elekterminal possesses. The 7-bit ASCII control codes are detected and decoded by a 256 x 4 ROM (IC7), which forms the CTL decoder of figure 1. The ASCII code is placed on the address inputs of this ROM and the code which appears at the data outputs is fed to the CO, Cl and C2 control lines of the CRTC. Table 1 has already listed a number of the control functions offered by the CRTC, However by utilising the read/ write line of the RAMs it is poss- ible to extend these* Table 4 provides an overview of all the various cursor control and screen scrolling functions which are provided by the Elekterminal, The majority of these functions can be selected by an individual key on the Elektor ASCII keyboard which was published last month. However, with the exception of 'home cursor’, all the above functions can also be obtained using the control key and the appro- priate data key, which means that the video interface is compatible w r ith keyboards other than the Elektor modeL The PROM decoder for the CRTC is programmed as shown in figure 5. This device is available from a number of manufucaturers under the type number 74S387. Since only 128 combinations are possible with a 7-bit code, only half of the PROM is used. Printed circuit board The printed circuit board for the video interface card (see figures 7 and 8) is not much larger than Eurocard format, but nonetheless is single-sided. Because of this there are a considerable number of through connections to be made (approx, 60 in all), however the extra effort required is more than compen- sated for by the lower cost of a single- sided board. The p.e.b. has been specially designed to accomodate an extension board to increase the number of memory pages to 16* The latter (which will appear shortly) simply plugs into the video interface card with tne aid of edge connectors. At the left hand end of the video interface board is a simple connector to accept the ribbon cable from the ASCII keyboard. All the connections to the keyboard, i.e, including supply lines, can be made via the connector socket. This connector is also used to bring out the data lines from the UART, These connections will be required when incorporating the extension board. The second p.e.b. connector provides access to all the address and data lines of the character memory and to two lines which are used to enable the RAMs to be addressed properly. This con- nector is designed to accomodate the memory extension board. It should be emphasised that the Elekterminal as presented here represents a complete output peripheral, which can be extended to accomodate more memory by plugging in one or more additional cards. The only on-board modification required is the removal of one through connection. The UART (IC8) is programmed by means of wire links. The connections indicated on the component overlay correspond to the recommended format listed in table 3 . Connection to a TV Not every TV has an input for an unmodulated video signal, but in those cases where one is present, or if a video monitor is being used, the ampli- tude of the terminal output signal has to be adjusted to suit the sensitivity of the input in question. This is best done as follows: The video signal is fed to the TV receiver or monitor via a length of coaxial cable (50 . . ,75 £2). The 'receiver-end’ of the cable should terminate in a low impedance. A 100 £2 potentiometer is ideally suited for this purpose. The potentiometer can then be used to adjust the signal amplitude to a suitable value. Naturally, an alternative solution will have to be found in the case of sets which are provided with an internal terminal impedance. If an input for unmodulated video signals is not available the output must be fed to a VHF/UHF modulator such as that published in the October issue of Elektor this year. Due to the large bandwidth of the video signal a certain degradation of picture quality is inevi- table, however the resultant definition is still quite acceptable for the type of application for which the Elekterminal is intended. With or without a modulator, the signal ampli- tude should be adjusted so that the picture is ‘sync-ing’ with both positive and negative polarity video signals. This can easily be checked by changing the polarity several times in succession. One should also first ascertain that the line oscillator of the TV receiver is correctly tuned. Supply Using normal memory ICs the current consumption of the circuit is around 750 mA (5 V). If low power memories are used, however, consumption drops to around 550 mA (5V). The -12V supply draws well under 100 mA. This means that the circuit could be powered using the spare capacity of an SC/MP system, assuming the latter was not driving a large amount of additional memory. Alternatively, a better idea might be to use the SC/MP power supply design to form a separate supply for the Elekterminal, especially if one bears in mind that extending the num- ber of character memory pages will push up the current consumption. H RAM diagnostic stektor decern ber 1978 — 12-25 RAM diagnostic An SC/MP program to test large sections of RAM The more RAM which one adds to ajiP system, the more difficult it becomes to trace any faults which may occur on one of the memory boards. Think, for example, of how tedious it would be to have to step through every location of a 4 K RAM card! Fortunately, however, with the aid of the right software, we can make the microprocessor itself do all the hard work. After all, computers are supposed to be ideally suited to taking over boring and repetitive tasks from humans. The following program can be used to test any block of memory larger than % K and will indicate the location of any faults which exist. H. Huschitt Figure 1. Flow diagram for the RAM- diagnostic program, showing the order of the different tests. Table 1, For reasons of space, the program is listed in condensed form. The complete program is also recorded on the Elektor Software Service record ESS-OOT A static Random Access Memory basi- cally consists of a large number of flip- flops and address decoders. These decoders are either integrated on the RAM chips themselves, or else are formed by discrete (TTL and CMOS) logic ICs, However, if somewhere within a section of memory there is a faulty 1C, a short circuit or a loose contact, there is a good chance that the program which is contained in that section of memory will fail to function properly. A suitable debug program must therefore be able to check for such a malfunction. The flow diagram of the RAM- diagnostic program is showm in figure 1 . 12-26 - alektor deaamber 1978 RAM diagnostic RAM - DG 0C 00 C4 5E C9 06 C4 3D C9 05 0C08 C4 00 09 FF C9 00 C4 80 0C10 C9 04 C9 03 C9 02 C9 01 0C18 C4 3E CA ID C4 00 37 C4 0C20 55 33 3F C2 01 CA 0A CA 0C28 10 C2 02 CA 0B CA 1 1 3F 0C30 C2 01 CA 13 C2 02 CA 14 0C38 C4 00 CA 12 C2 10 CA 0E 0C40 C2 11 CA 0F C4 00 37 C4 0C48 B4 33 3F C4 55 CA 12 C2 0C50 0A CA 0E C2 0B CA OF C2 0C58 01 CA 13 02 0B CA 14 C4 0C60 OD 37 C4 B4 33 3F C2 0E 0C68 E2 01 9C 08 C2 0F E2 02 0C70 9C 02 90 22 C2 01 CA 13 0C78 G2 02 CA 14 02 C2 0E F4 0C80 01 CA 0A CA 0C C2 0F F4 0C88 00 CA 0B CA OD C4 0D 37 0C90 C4 DC 33 3F 90 A2 C4 00 0C98 CA 12 C2 10 CA 0E C2 1 1 0CA0 CA 0F C4 0D 37 C4 B4 33 0CA8 3F C2 10 CA 0C C2 11 CA 0C80 0D C4 0D 37 C4 DC 33 3F 0CB8 C4 FF CA 12 C2 10 CA 0E 0CC0 C2 11 CA 0F C4 OD 37 C4 0CC8 B4 33 3F C2 10 CA OC C2 0CD0 11 CA 0D C4 OD 37 C4 DC 0CD8 33 3F C2 01 CA 13 C2 02 0CE0 CA 14 C4 AA CA 12 90 IB 0CE8 C2 0C CA 10 C2 0D CA 11 0CF0 C2 12 E4 55 9C 0A C4 00 0CF8 CA 13 C2 0F CA 14 90 03 mm 40 CA 14 C4 01 31 C4 07 0D08 35 C2 12 01 40 E4 AA 9C 0D10 0A C4 54 C9 00 C4 5C C9 0018 FF 90 24 40 E4 55 9C 0A 0D20 C4 06 C9 00 C4 00 C9 FF 0D28 90 15 40 98 0A C4 4F C9 0D30 00 C4 00 C9 FF 90 08 C4 0D38 5B C9 00 C4 00 C9 FF C4 0D40 79 C9 06 C4 50 C9 05 C9 0D48 04 C9 02 C4 5C C9 03 C4 0D50 00 C9 01 C4 0E 37 C4 1C 0D58 33 3F C4 A0 CA 1 D C2 10 0060 CA 01 C2 11 CA 02 C4 00 0D68 37 C4 55 33 3F C4 00 C9 0D70 FF C9 00 C9 05 C9 06 C4 0D78 0E 37 C4 1C 33 3F C2 13 0D80 CA 01 C2 14 CA 02 C4 00 0D88 37 C4 55 33 3F C4 00 C9 0D90 FF C9 00 40 E4 AA 98 0B 0D98 40 E4 55 98 06 C4 00 C9 0DA0 01 C9 02 8F FF 8F FF C4 0DA8 0E 37 C4 1C 33 3F C4 0D 0DB0 37 C4 02 33 3F C2 0E 31 0DB8 C2 0F 35 C2 12 C9 00 C2 0DC0 0E E2 13 9C 06 C2 0F E2 0DC8 14 93 0F 02 C2 0E F4 01 0DD0 CA 0E C2 0F F4 00 CA 0F 0DD8 90 DB 3F 90 D8 C2 00 31 0DE0 C2 OD 35 Cl 00 E2 12 C9 0DE8 02 01 C2 12 E4 55 98 05 0DF0 40 98 0C 90 03 40 9C 07 0DF8 C4 0C 37 C4 E7 33 3F C2 0E00 0C E2 01 9C 09 C2 0D E2 0E08 02 9C 03 3F 90 CF 02 C2 0E10 0C F4 01 CA 0C C2 0D F4 0E18 00 CA 0D 90 C0 C4 08 C8 0E20 08 0E28 00 8F FF B8 04 9C FA 3F Once started, the program writes ‘00 5 (i6) (° r ‘FF 7 he)) into each byte of the section of RAM to be tested. The latter can be selected by entering the start and end addresses of the section of RAM in question. The program then writes the byte ‘55 5 (i^ into the first Va K of the suspect block of RAM (the Elektor 4 K RAM card is structured in blocks of % K). The program next tests to see whether the byte ‘55’ appears anywhere else in the memory. If that is the case, then it means that there is obviously a fault in the address decoding of the RAM. If the program fails to locate the byte ‘55* elsewhere in memory, then the program repeats the same test, but this time with the byte in the second l A K of RAM to be tested. The cycle is repeated until the entire suspect block of memory has been checked. When an error is detected, this is regis- tered on the displays and the test program is stopped. It can only be continued if the fault in question is rectified. The next stage in the program consists of loading the byte ‘00’ into every location in the memory and testing to check the contents. If every- thing is in order, the test is repeated, but this time for the byte ‘FF 5 If all these tests prove negative, then one can safely assume that the malfunction of a program loaded into the area of memory in question is not due to a fault in the memory hardware. Program The complete condensed listing of the RAM -diagnostic program is given in table 1. The program is loaded from location 0C00 and started from that address. The byte ‘FF* he) for the first test can be loaded into address 0039. When the program is started the text ‘DG . 5 should appear on the displays, whereupon the start address of the section of memory to be tested should be entered. As is apparent from the flow-diagram, this should be the initial address of a l A K RAM IC, and the length of the section of memory under test must be a multiple of Va K (x 8). The start address should be immediately followed by the end address of the block of suspect memory (e.g. 1000 to 1FFF). The program will start as soon as the end address has been entered. To test an entire 4 K RAM card lakes approximately 2 minutes. It is perfectly possible to test larger amounts of memory, thereby crossing page boundaries. The time taken increases exponentially with the size of the memory to be tested. During the test the last entered address will remain on the displays. The program can terminate in one of four different ways: ‘Error no 5 . When this appears on the displays it indicates that the RAM in question has been given a clean bill of health. This text is always immediately followed by first the start and then the end address of the section of memory which has just been tested. ‘Error r. For this text to appear the program must have found the byte the wrong section of mem- ory, This indicates a fault in the address decoding of the memory, caused, for example, by a short in the CE line of the RAMs in question. This text is immediately followed by the ‘block address 5 of the faulty % K of RAM. If the precise address of the faulty store is desired, this can be obtained by using the ‘MODIFY 5 routine to load ‘FF 5 into address 0C38 and then restarting the program. The exact address where the error is found will then be displayed immediately after the start and end addresses of the faulty ‘block 5 , ‘Error 2\ This text indicates that one or more bits have not been reset (loaded with 5 0 5 ). The first address where this is the case is registered on the displays, followed by a two-digit number which indicates which bit(s) in the byte are false. For example, ‘0F 5 would mean that bits 0, 1, 2 and 3 of the byte in question contained a ‘l 5 instead of a Possible causes for such an error include a faulty RAM IC, bad solder joints, faulty IC sockets, a break in the CE or Read/Write lines. ‘Error 3 5 , This is basically the same fault as in the previous case, except that here each bit is tested for a 4 1 5 . Once again the first address where a fault is detected is displayed. In conclusion, it should be remarked that an ‘Error 5 indication need not always denote a fault in the memory hardware. It is perfectly possible that the program will detect an ‘error 5 , which if the program is run a second time, will promptly disappear. In such a case the ‘fault 5 is probably due to such factors as, e.g., unsatisfactory supply voltage(s), faulty decoupling capacitors or bus- drivers with too small a fan-out, H reading elektor elektor decern ber 1978 — 12-27 reading elektor TUNs and TUPs Nowadays most low-frequency, small- signal silicon transistors from reputable manufacturers meet the following mini- mum specifications: UCEG, max 20 V lC r max 100 mA frfe, min 100 Ptot, max 100 mW fT, min 100 MHz When a transistor of this type is required, it is referred to in Elektor as a TUN (Transistor, Universal, ISIPN) or a TUP (Transistor, Universal, PIMP). Some TUNs are the BC 107, BC 108 and BC109 families, and the 2N3856A, 2N3859, 2 N 3860, 2N39G4, 2N4124 and HEP S001T. Some TUPs are the BC 177, BC 178 and BC 179 families, and the 2N2412, 2N3251, 2N3906, 2N4126, 2N4291 and HEP S0G1 3. DUS and DUG Similarly, for many small-signal appli- cations the only really important difference between all the available diodes is that some are silicon and some are germanium. When a general- purpose small-signal diode is required in an Elektor circuit, it is often desig- nated DUS (Oiode, Universal, Silicon} or DUG (Diode, Universal, Germanium). However, it should be noted that even a DUS or DUG should meet minimum specifications: DUS DUG Ur, max 25 V 20 V *F, max 100 mA 35 mA l R, max 1 juA 100 juA Ptot, max 250 mW 250 mW CD, max 5 pF 10 pF Some DUS's are: BA127, BA217 BA 218, BA 221 , BA222, BA317, BA 318, BAX 13, BAY 61, 1N914and 1N4148. Some DUGs are: OA85, OA91, OA95 and AA1 16. Resistors Unless otherwise specified, resistors are % Watt 5% tolerance carbon types. Higher power ratings are, of course. permissible (e.g, the recently introduced fl /3 Watt' types), provided they fit on the p.c, board — if this is to be used. A 10% tolerance type is usually also permissible, with only a minor effect on the performance. The resistance values are specified using V for 1000 ft and 'M'for 1,000,000 ft; the decimal point is replaced by either 'ft', 'k' or W. For instance, 4ft7 = 4.7 ft; 4k7 - 4700 ft; 4M7 = 4.7 Mft. Capacitors The DC working voltage of capacitors (other than electro I ytics) Is normally assumed to be at least 60 V, unless otherwise specified. Generally speaking, of course, a DC working voltage equal to (or greater than) twice the supply voltage is sufficient. In most circuits where electrolytic capacitors are used, a working voltage equal to the supply voltage plus 20% is safe; very often, a lower voltage is sufficient, in recent Elektor circuits, the lowest DC working voltage permissible is often specified, regardless of availability; in practice, any higher voltage type can be used — bearing in mind that a higher voltage rating involves greater physical size, so the available space on the board should be watched. For instance, in a circuit operating off a 9 V battery, an elco might be specified as 1 ji/16 V — even though the normally available types are 1 ju/63 V, Capacitor values are specified using 'p' for 10~ 12 , 'n' for ] 0~ 9 and for 1G -6 . As with resistors, the decimal point is replaced by one of these letters. This means that 4700 pF, for instance, is written as 4n7 — not as 0.0047 juF! Voltages The international letter symbol 'U' for voltage is normally used instead of the ambiguous 'V'. 'V # is normally reserved for 'volts'. For instance: Up = 10 V, not Vp = 10 V. The DC test voltages shown in circuits are measured with a 20 kft/V instru- ment, unless otherwise specified. No mains (power line) voltages are given in Elektor circuits. It is assumed that our readers know what voltage is standard in their part of the world ! International problems Although it is the intention that Elektor circuits can be built and used all over the world, some problems are unavoid- able. Readers in countries that use 60 Hz should note that Elektor circuits are designed for 50 Hz operation. This will not normally be a problem; however, in cases where the mains frequency is used for synchronisation some modifi- cation may be required. In some cases, the necessary modifications are given; in others, especially where modification involves drastic re-design, the circuit idea may be of use — even though the circuit itself is not. Circuits intended for use with domestic television receivers may also run into difficulties. As many of our readers will know, Television Standards' are anything but 'standard'! However, by avoiding sound and colour wherever possible, a reasonable degree of inter- national compatibility can be achieved — provided the preset adjustments are given a suff iciently wide range. Finally, especially for our readers in Canada and the USA, a list of equivalents for some of the commonly- used transistor types may prove useful: AF 239: G0003 BC 140, BC 141 , BC 142: S3011 BC 160: S3012 BC546: no direct equivalent known to us, however in most cases an S0024 can be used BD 241: TIP29, S5000 BD 242: TIP30, S5006 BF 259, BF 494, BFT66: no direct equivalent known to us; basically, these are low-voltage, low-current high-frequency devices. Some further equivalents differ only in that the emitter and collector connec- tions are transposed, so that they must be mounted 'back-to-front' on the EPS p.c, boards: TUN: RS2010, RS2016 TUP: RS2022, RS2034 BC 107 or equ.; SOOT 5 BC 177 or equ.: S0019 BC 109B,C or equ,: S0G24 BC 517: S9100 BC 556: S0026 H 12-28 — eleklor decern ber 1978 index Audio AGO microphone preamp 7-74 amplifier for low-2 headphones , , 7-05 analogue delay line . . . . .......... 7-01 analogue reverberation unit 10-44 automatic mono/stereo switch 6-34 compander ........ 4-14 consonant control amplifier. ....... t * . .7-38 digital audio mixer 7-58 digital delay line 7-57 digital reverberation unit 5-08 disc preamp t ....... 7-71 disco drums 12-14 DJ Killer 7-68 electronic input selector .4-06 elektor equaliser 1-08 elektornado power amplifier 4-07 FET audio amplifier .7-25 limfter/compressor ........................ 7-87 loudspeaker connections 4-39 I um inant level indicator ..................... 7-48 micropower amplifier ................. ..... 7-1 5 moving coil preamp . 4-02 peak programme mater t- 17 preconsonant disc preamp 7-34 PWM audio amplifiers 1 2-04 stepped volume control . 4-12 stereo pan pot 4*17 stereo width control 7-22 third octave filters (ejektor) 1-32 variable fuzz box .12-31 vocoders { 1 ) 4-27 vocoders (2) . . . .5-21 18 dB per octave high/I owpass filter 7-18 Car & bicycle automotive voltmeter . 7-10 bicycle speedometer 7-32 brake efficiency meter ...................... 7-85 car ammeter . . . . 7-79 car battery monitor 7-28 car lights failure indicator 3-11 car rip-off protection 4-26 car start booster , . 9-30 car voltage /current regulator .................. 7-67 flasher bleeper . , . 7-27 power flasher 7-26 Design Ideas bandwidth limited video mixer 7-83 cheap crystal filter ........................ 7-58 constant amplitude squ a rewave to sawtooth converter . . .6-14 electronic variable capacitor . 7-62 hum filter using electronically simulated inductor ...... 7-22 improved 723 supply .7-19 no-noise preamp ......................... 7-76 selective bandpass filter ..................... 7-30 simple CMOS squarewave generator .............. 7-07 simple LS TTL squarewave generator ............ .7-06 simple TTL squarewave generator 7-06 speedy rectifier 7-24 stable-start -stop-squarewave . .......... 7-31 super-simple touch switch . 7-09 TAP-tip 1-16 temperature-compensated reference voltage . . . 7-89 TTL-LC-VCO 7-13 variable capacitance multiplier 7-62 voltage mirror ........................... 7-23 zero crossing detector ...................... 7-17 18 dB per octave high/I owpass filter 7.18 723 as a constant current source . .4-18 Domestic and hobby automatic aquarium lamp 7-74 automatic shutter release .................... 7-81 back and front doorbell 7-23 bat receiver 7^32 burglar alarm .............. 7-1 2 cackling egg-timer 11-02 central alarm system ...................... 10-20 cold shower detector . 7-62 development timer ........................ 2-24 electronic gong 7-09 electronic open fire , .1-16 Infra-red light gate 2-02 Infra-red receiver 7-14 inf ra-red transmitter . . 7-20 liquid level alarm 7-24 loudspeaking telephone 11-31 percolator switch 5-26 proximity detector 10-09 safety first 3-34 slide-tape sy nchroniser . 7-92 slow on/off fader . 247 solid-state thermostat 7-65 touch dimmer 7-90 ultrasonic alarm indicator .................... 7-30 ultrasonic alarm receiver 7-29 ultrasonic alarm transmitter . .7-28 Games and model building coulourTV games . 3-12 electronic gong 7-09 electronic maze , . 3-23 glowpfug regulator 7-64 Joysticks ji-ig many hands make light work 7-93 mode! railway lighting ....... . 7-82 monopoly dice 6-18 pocket bagatelle 1 1 -38 puffometer .9-02 quizmaster . 7-68 ring the bell and win a prize .11-22 servo polarity changer 6-16 super cheapo NiCad charger 7^02 tag! . . . 11-12 traffic light controller .6-36 Generators AM/FM alignment generator .................. 7-80 CMOS function generator .................... 2-20 CMOS noise generator 1 -05 digital spot sinewave generator ................. 7-03 frequency synthesiser 7-93 master tone generator . 9-09 sawtooth oscillator 7-12 simple function generator 1-40 signal injector 7-27 simple CMOS squarewave generator .............. 7-07 simple LS TTL squarewave generator ............. 7-06 simple TTL squarewave generator . . 7-06 sine-cosine oscillator ....................... 7-08 squarewave oscillator 7-88 stable-start-stop-squarewave .................. .7 31 tone-burst generator 12-10 TTL-LC-VCO 7-13 15 duty-cycles at the turn of a switch ............ It -30 288 MHz generator . . . 7-04 index elektor decern bar 1978 — 12-29 HF aerial systems for satellite communication cheap crystal filter cheap r.f, amplifier colour modulator .............. FM IF strip ................. frequency synthesiser controller mini short-wave receiver .......... modulatable power supply ........ programmable call generator ....... TV sound modulator ............ VHF preamp VHF/UHF TV modulator ......... wideband RF amplifier 2 m transmitter 283 MHz generator . 5-04 . 7-58 „ 7-07 , 5-28 .7-84 . 7-86 ,5-17 . 7-73 .6-26 ,6-29 .7-17 10-27 , 7-16 . 7-63 * 7-04 Informative articles aerial system for satellite communication . .5-04 an introduction to the TV scope 10-03 applikator, complex sound generator ............ .9-32 applikator, light sensing chip 10-42 applikator, RPM and dwell meter 1 1-34 buffered/unbuffered CMOS . . 943 circuit boards and soldering ,5-34 ejektor, third octave filters 1 -32 ejektor, TV scope using bucket brigade memory ....... 5-20 elektor software service 6-20 one chip does not make a piano , .9-08 PWM audio amplifiers 1 2-04 safety first 3.34 throwing some light on LEDs ... . .2-06 vocoders (1 ) . 4-27 vocoders (2) 5-21 metronome . , 7-10 oscilJographics , , 9-06 pseudo random running lamp .................. 7-72 saw-song 7-33 simple video mixer 7-70 simple video sync generator ................... 7-19 squarewave-staircase converter . . 7-16 stereo vectorscope 7-66 temperature-control led soldering iron 9-24 touch controller 7-59 Music disco drums 12-14 easy music 640 extending the elektor piano .................. 12-51 formant - the elektor music synthesiser 17) , . .1-34 formant - the elektor music synthesiser (8) .......... 2-10 formant - the elektor music synthesiser (9) ......... .3-26 formant - the elektor music synthesiser (10) ........ .4-33 master tone generator ...................... 9-09 one chip does not make a piano 9-08 piano . 9-1 2 resonance filter module 10-12 singing SC/MP 12D8 24 dB VCF 9-34 Power supplies improved 723 supply .7-19 modulatable power supply ................... 7-73 SC/MP power supply 3-08 symmetrical ± 15 V/50 mA supply ............... 7-1 1 tern pe r atu re -co m pe n sa ted ref eren ce v ol tage , . 7 -89 voltage mirror 7-23 0-30 V regulated supply 7-72 78 L voltage regulators 7-75 Microprocessors ASCI! keyboard 11-06 cassette interface .4-20 CMOS FSK modulator . , 7-69 data bus buffer ’ 7-78 data bus buffer 10-18 data multiplexer 7_7g debouncer 7-29 digital clock using the SC/MP . .6-24 elekterminal 1246 elektor software service 6-20 experimenting with the SC/MP 13) , 1-24 experimenting with the SC/MP (4) , , ,2-28 experimenting with the SC/MP (5) ...... .3-03 fun with a RAM 3-22 hexadecimal display . 7-91 programmable address decoder 7-88 RAM diagnostic program 1 2-25 reaction timer program 6-32 SC/MP Mastermind program . . , , . 11-35 SC/MP power supply ............ . , , , ...... 308 singing SC/MP 12-08 software Kojak siren . . 7-08 stable -start-stop -square wave 7-31 supply failure indicator , ’7-02 word trigger .......... , , , . , , . , . . 7_g1 4k RAM card 3-20 Miscellaneous analdgue-dtgital converter .......... P B ... . 7-21 electronic soldering iron 7-70 infra-red receiver .7-14 infra-red transmitter . . . 7-20 Test and measuring equipment AM/FM alignment generator . analogue-digital converter CMOS function generator CMOS noise generator . . digtscope . ditigal spot sinewave generator DIN cable tester .... FET milli voltmeter . . , frequency synthesiser . HF current gain tester , 1C Counter timebase . . LED X-Y plotter .... mini counter , , peak programme meter real load resistors .... sawtooth oscillator , , , * signal injector ....... simple function generator sine-cosine oscillator * . squarewave oscillator , timebase scaler , . . . . . Tone-burst generator . , TV scope, introduction TV scope, basic version TV scope, extending the TV scope, extended version UAA 1 80 LE D voltmeter , universal logic tester .... voltage comparison on a 'scope word trigger zener tester ....... 1/4 GHz counter .... 1 5 duty-cycles at the turn of a switch . . 7-80 . .7-21 . . 2-20 . ,1-0 5 . 1147 . .7-03 . . 7-77 . . 7-60 . , 7-93 . . 7-20 * , 7-56 . ,7-94 . .6-21 . , 1-17 , .4-13 . .7-12 . .. 7-27 . , 140 . . 7-08 , .7-88 . . 7-76 . 1240 , 10-03 , 10-30 . 11-25 . 12-34 . , 1 -20 . . 5 -3 8 , .3-33 . . 7-61 , . 2-22 . .6-01 . 11-30 12-30 — elektor decern ber 1978 tup-tun -dug-d us TUPTUNDUGDUS Wherever possible in Elektor circuits, transistors and diodes are simply marked 'TUP' {Tran- sistors, Universal PNP), TUN' (Transistor, Universe! NPN), 'DUG 1 (Diode, Universal Ger- manium) or 'DUS' {Diode, Universal Silicon). This indicates that a large group of similar devices can be used, provided they meet the minimum specifications listed in tables la and 1b* 1 type Uce 0 max lc max hfe min. Ptot max fr min. TUN TUP NPN PNP 20 V 20 V 100 m A 100 mA 100 100 100 mW 100 mW 100 MHz 100 MHz | Table la. Min TUN. timum specifics tions for TUP and Table 1b. Minimum specifications for DUS and DUG. type Ur max IF max (R max Ptot max CD 1 max DUS DUG Si Ge 25 V 20 V 100 mA 35 mA 1 JUA 1 00 [lA 250 mW 260 mW 5 pF 10 pF 1 Table 2. Various transistor types that meet the TUN specifications* TUN BC 107 BC 208 BC 384 BC 108 BC 209 BC 407 BC 109 BC 237 BC 408 BC 147 BC 238 BC 409 BC 148 BC 239 3C413 BC 149 BC 31 7 BC 414 BC 171 BC 318 BC 547 BC 172 BC 319 BC 548 BC 173 BC 347 BC 549 BC 182 BC 348 BC 582 BC 183 BC 349 BC 583 BC 184 BC 382 BC 584 BC 207 BC 383 1 Table 3. Various transistor types that meet the TUP specifications. TUP BC 157 BC 253 BC 352 BC 158 BC 261 BC 415 BC 177 BC 262 BC 416 1 BC 178 BC 263 BC 417 BC 204 BC 307 BC 4T8 BC 205 BC 308 BC 419 BC 206 BC 309 BC 512 BC 212 BC 320 BC 513 BC 213 BC 321 BC 514 BC 214 BC 322 BC 557 BC 251 BC 350 BC 558 BC 252 BC 351 BC 559 The letters after the type number denote the current gain: A: a' (J3,h fe } = 125-260 B: a' = 240-500 C: a’ = 450-900. Table 4. Various diodes that meet the DUS or DUG specifications* OUS DUG BA 127 BA 217 BA 218 BA 221 BA 222 BA 317 L - — BA 318 BAX 13 BAY 61 1N914 1N4148 OA 85 OA 91 OA 95 A A 116 Table 5, Minimum specifications for the BC1Q7, -108, -109 and BC177, -T7S, -179 families (according to the Pro-Electron standard). Note that the BG178 does not necessarily meet the TUP specification (ic,max = SO mA), NPN PNP BC 107 BC 177 BC 108 BC 178 8C 109 BC 179 u ce 0 45 V 45 V max 20 V 25 V 20 V 20 V 6 V 5 V max 5 V 5 V 5 V 5 V lc 100 m A 100 mA max 100 mA 100 mA 100 mA 50 mA Ptot. 300 mW 300 mW max 300 mW 300 mW 300 mW 300 mW f T 150 MHz 130 MHz min. 150 MHz 130 MHz 150 MHz 130 MHz F 10 dB 10dB max 10 dB 10 dB 4 dB 4 dB Table 6. Various equivalents for the BC107, -108, . . families. The date ere those given by the Pro-Electron standard; individual manu- facturers will sometimes give better specifi- cations for their own products. NPN PNP Case Remarks BC 107 BC 108 BC 109 BC 177 BC 178 BC 179 •Q 1 E BC 147 BC 148 BC 149 BC 157 BC 158 BC 159 ■Q E Pmax “ 250 mW j BC 207 BC 208 BC 209 BC 204 BC 205 BC 206 ‘O £ BC 237 BC 238 BC 239 BC 307 BC 308 BC 309 ■Q £ BC 317 BC 318 BC 319 BC 320 BC 321 BC 322 & lemax = 150 mA BC 347 BC 348 BC 349 BC 350 BC 351 BC 352 <3 BC 407 BC 408 BC 409 BC 417 BC 418 BC 419 Pmax = 250 mW BC 547 BC 548 BC 549 BC 557 BC 558 BC 559 <3 p max - 500 mW BC 167 BC 168 BC 169 BC 257 BC 258 BC 259 GHj 169/259 ! cmax - 50 mA BC 171 BC 172 BC 173 BC 251 BC 252 BC 253 "Q 251 . . ,253 low noise BC 182 BC 183 BC 184 BC 212 BC 213 BC 214 •<3 lc max = 200 mA BC 582 BC 583 BC 584 BC 512 BC 513 BC 514 •<3 ! cmax = 200 m A BC 414 BC 414 BC 414 BC 416 BC 416 BC 416 9 © low noise BC 413 BC 413 BC 415 BC 415 •O iow noise BC 382 BC 383 BC 384 •<3 BC 437 BC 438 8C 439 III; Pmax = 220 mW BC 467 BC 468 BC 469 01 Pmax - 220 mW BC 261 BC 262 BC 263 ■Q low noise variable fuzz-box eloktor decern ber 1973 — 12-31 variable fuzz -box A simple circuit for musical sound effects Particularly in modern pop music, electronically-produced sound effects are an extremely common device. One only has to think, for example, of the widespread use of the 'wa-wa' pedal and the fuzz- box. With this in mind, Elektor have designed a super simple circuit which, by employing signal clipping techniques, can produce a large variety of manually controlled sound effects. Figure 1. The response of an amplifier which IS driven into 'hard' clipping (a) and the corre- sponding input and output waveforms. Figure Z There are five parameters of the clipping response of an amplifier which can be independently varied. Using only a handful of components it is possible to construct a highly effec- tive variant of the well-known fuzz box. This type of circuit commonly employs a pair of anti-parallel connected diodes, which are inserted in the amplifier (be it IC — or transistorised) so as to clamp the output signal above certain values of the input signal. This process is illus- trated in figure 1 , where for the sake of clarity the amplifier is assumed to have unity gain over the linear portion of its transfer characteristic. As can be seen, above an input voltage Uj, the output voltage shows no further increase; simi- larly, the output voltage will not fall below the input value -U 2 . If U* is equal to U 2 (which is typically the case in fuzz-box circuits) and the input signal is sufficiently large, then the input and output signals will differ from one another as shown in figure lb. When, as is the case in figure lb, the output signal clips symmetrically (i.e . U x = U 2 ), it contains only even harmonics, and it is this which gives the la ■Jout ft) 12-32 — elektor decern b* r 1978 variable fuzz-box 10 ... 15 V© #see text TUP 10*. 15 V Q 10 ... 15 V 10... 15 V 9984 3 Figure 3, The circuit diagram of a variable fuzz-box for a symmetrical power supply. Figure 4. The circuit of a variable fuzz-box for an asymmetrical power supply. Figure 5. Example of a suitable power supply for the circuit of figure 3 (a) and of fig- ure 4 (b). resultant musical signal its characteristi- cally Tough and "fuzzy 1 edge. However the tonal character of the music signal can be considerably enriched by arrang- ing that the output signal clips asym- metrically (i.e. Ui # U 2 ). In this way it is possible to influence the sound of the fuzz-box to produce more varied effects. The circuit described here is designed to offer the best of both worlds by al- lowing the clipping levels Ui and U 2 to be altered independently of one another, thus permitting the type of effect to be varied as desired. Varying the clipping levels is not, how- ever, the only way of influencing the sound of the (clipped) output signal A further factor determining the type of effect produced is whether the amplifier starts clipping almost immediately it reaches a particular level (hard clipping), or whether the transition between non- limiting and limiting is more gradual (soft clipping). In the circuit described here it is possible to continuously vary the response of the amplifier between these two extremes. The various control facilities offered by the "variable fuzz- box" are illustrated in figure 2, Circuit diagram The complete circuit diagram of the fuzz-box is shown in figures 3 and 4; figure 3 gives the circuit for a symmetri- cal power supply (plus, minus and earth), whilst figure 4 is designed for an 1 variable fuzz -box SoiS/COv 11 ' s a u ' 5V FI i _r 250 mA % © t 2 k TZ.„tB V 50 mA r % r ^ 4 x 1N40CT1 ISoop X ^ ^OOn 40V C=l A- - EU —I ^400mw 12 15 V 400mW BD 136 BD 137 ^ __ r 1 BD 139 \ N 40V (60 V>© ♦ — rT^5H — r— © 21 5 v <32,5 V) lOmAl V' H *? 5 fill ^ 2 20u 40V 22V (33 V) 400mW loon SJ9EH - 5b asymmetrical supply (plus and minus/ earth). The current consumption of the circuit is 1 0 to 1 S mA, whilst in both figures 3 and 4 the input impedance is 100 k. The operation of the circuit is fairly straightforward. The input signal is first amplified by IC1, the gain of which Is R2 1 + The gain can be altered by varying Rl; with the value given the gain is 11. The output signal of IC1 is fed via R4 (and 06 in figure 4) to the volume control P5, the wiper voltage of which forms the output signal of the circuit. The clipping is controlled as follows: as soon as the voltage at the right hand end of R4 exceeds the wiper voltage of P3 (or falls below the wiper voltage of P4) the output signal is attenuated. The de- gree of attenuation is determined by the elektor dec amber 1978 — 12-33 ratio of R4 to PI (P2). With PI (P2) set to its minimum resistance value the out- put signal is completely attenuated, i.e. is ‘hard' clipped. With PI set to its other end stop (minimum resistance) ‘soft 5 clipping is obtained. Thus by adjusting the four potentiometers P i , . . F4 which control both the levels at which the amplifier starts to clip and the degree of clipping, it is possible to vary the tonal character of the resultant sound as desired. As far as a power supply is concerned, various possibilities exist. Figures 5a and 5b show a suitable arrangement for the circuits of figures 3 and 4 respectively. Other applications Apart from its use as a variable special effects generator the circuit also has other possible applications. For example, it can be used to limit the input signal of a power amplifier to the value which just produces maximum output. In this way one obviates the danger of current limiting or clipping occurring in the power amp and its undesirable conse- quences for the listener. At the same time the circuit thus represents a type of overload protection for the power amp. Another interesting idea would be to use the circuit in conjunction with P.A. systems. There are various theories which claim that it is possible to increase the volume of the signal by clipping it in a certain way. The increase in the harmonic content of the signal is said not to impair the intelligibility of the address. It has also been suggested that different clipping characteristics may explain some of the oft-debated differ- ences between ‘transistor-sound' and Valve-sound'. The circuit described here seems ideally suited to test the truth of these ideas. However readers wishing to experiment in this field would do well to spare a thought for their neighbours! 12-34 — eleklof decern ber 1978 TV scope extended The TV scope, introduced in the recent October issue, can now be completed. Two months ago the basic version was described in detail and the necessary circuits were given for converting a standard domestic television receiver into an oscilloscope — be it with limited capabilities. Last month the possibilities for extending the TV scope were discussed; in particular, the principles and advantages of a variable timebase and trigger facilities were explained. This third and final article in the series gives the necessary extension circuits. Although the basic version is an extremely useful tool when testing low-f requency circuits, its 'big brother' offers vastly superior display possibilities. Any signal within the frequency range of the scope (DC to 100 kHzJ can be displayed as a clear and stable trace on the screen of any domestic TV receiver. version Several block diagrams can be given for the same circuit, as shown in figure 1: the third (and last!) block diagram of the ‘TV scope - extended version’. The sections required for extending the TV scope are shaded in; the unshaded portions are the basic version described in the October issue. Furthermore, the block diagram clearly shows the separ- ate portions of the circuit which will be mounted on separate printed circuit boards. In the following, these extension circuits will be discussed; the trigger circuit, the input timebase and the ‘bucket-brigade memory’ timebase expander with its associated drive circuits. The portions of the diagram shown in dotted lines are only required for a two-channel version of the "scope’. The timebase expanders (bucket -brigade memories) of the A- and B -chan nets (assuming that the latter is also included) ‘slow down’ the analogue input signals Uya and u y b as required. For correct operation, they require three control signals; two clock signals, pi and 02, and a "mode 5 signal {ujq) that deter- mines which of the memories is in the ‘read’ cycle and which is in the ‘display’ mode. The basic principles involved were described last month. The three control signals are derived from a fairly complicated logical circuit, which is designated quite simply in the block diagram: the ‘control circuit’. This circuit derives the necessary output signals from several input signals. Four inputs are provided by the main board of the basic version of the TV scope. Two further inputs are provided by the trigger circuit, which forms part of the extension circuitry. One cl the latter signals, Ug^mples corresponds to twice the clock frequency required during the ‘read’ cycle of the memories. The frequency of u sam pi e therefore deter- mines the timebase expansion obtained; in other words, it determines the time axis of the final TV scope display. The second signal provided by the input timebase board is u x . This signal is derived from the Ureset output of the basic version; a pulse train at the 50 Hz frame frequency. The u x signal is also a pulse train -in other words, it is a digital signal and it goes ‘high’ (+1 5 V) in the same rhythm, but with a delay TV scope — extended version T 1 : I i ' I J i i ( with respect to u rese t that is determined i by the setting of a potentiometer,‘x-posl The u x signal determines the start of the display cycle with respect to the start of each frame of the TV picture. In effect, it therefore determines the position of the display along the time axis or I ‘X-axis’, which is why the corresponding control is labelled ‘x-po$(ition)\ ( The last input signal to the control circuit is ut r . This signal is derived from one of the y -input signals or from a third input signal (‘external trigger’). 1 When the selected input signal to the trigger circuit exceeds a certain level, as set by the ‘trigger level’ control, the U| r output goes ‘high’ - see figure 2. Since the positive-going edge is detected by the control circuit, triggering occurs on the leading edge of the selected input signal. Basically, this circuit is equiv- alent to similar circuits in any ‘normal 5 oscilloscope. Figure 3 shows the relationships be- tween the various in- and output signals of the control circuit. At each reset pulse, the mode signal (u m ) changes state, selecting first one memory and then the other for each channel. Simul- taneously the u x signal goes low for a fixed period; at the end of this period the display cycle starts. Two things happen at this point. The signal Ug a te goes high; this signal is fed to one input of N15 in the circuit which generates the white-level pulses for the trace on the TV screen (see the description of the basic version, figure 4c), so that a trace only appears on the TV screen when one of the memories is actually in its display cycle. Between display cycles, the trace is blanked. The second signal initiated by the positive-going edge of u x is the display clock pulse train. As explained last mon month, this signal must be fed to the two memories alternately — each in turn being used as display memory — and for this reason these (relatively low fre- quency) clock pulses appear in alternate bursts on the 01 and 02 lines. When one memory (for each channel) is in the display mode, the other is ‘storing the input signal. This ‘read’ cycle is initiated by the first trigger pulse, ut r , after the reset pulse. A control signal to I the input timebase, u e t ( ‘enable time- TV scope — extended version elektor decern ber 1978 — 12-35 base')* goes positive. The input time base clock generator starts on the positive- going edge of u e t and produces a rapid burst of clock pulses, alternately on the 01 and 02 lines. As in a conventional scope, this facility is included in order to obtain a stable display - independent of input frequency. Having surveyed the basic principles involved, it is now time to discuss the necessary circuits in greater detail. Trigger circuit The trigger circuit is shown in figure 4 . A three-way switch selects the desired trigger input: either of the two input signals to the TV scope, or an indepen- dent trigger input. If either of the two Y-inputs to the scope is selected, a signal at "standard level 1 is required. This signal, ut> is derived from the input amplifiers) of the TV scope - see "TV scope - basic version’, figure 2 on page 10-31. Depending on the type of signal selected, either AC or DC coupling may give better results. Switch S2, bridging Cl, is used to select the desired option. IC1 boosts the input signal to a level suitable for the comparator stage, IC2. This comparator refers the input signal level to a DC voltage set by the "trigger level 1 potentiometer. If the (amplified) input signal exceeds the preset level, the output of IC2 goes high (see figure 2) — note that IC1 inverts the input signal! As long as trigger pulses are appearing at regular intervals, LED D2 lights (Trig’d). The output signal, ut r? is fed to the control circuit on the memory board. The printed circuit board for the trigger circuit contains one further control: potentiometer P2, "trace distance 1 . This control has nothing to do with the trigger circuit, but it happened to fit neatly at this point on the front panel. The function of this control will be described further on. The trigger circuit and the trace distance control are both mounted on the board shown in figure 5, The 4 AC/ DC* switch S2 and capacitor Cl are mounted off-board. Basically, of course, any LED can be used for D2, However, the neatest memory board trigger | extern ' u ya, y sha “03 Ya QO — Tv scope basic version Co— j Ushb u ob u gate y tb control circuit y sample u* | i'- r 'i, • '■ ^ i ',! L i T . ^ Q3 r Q4. Q1 1 trigger circuit trace ! distance free trigger run u reset X position input timebese 7902 S 1 Figure 1. Block diagram of the TV scope - extended version. The shaded sections are the extension circuits. Figure 2. Trigger pulses {u{ r ) occur when the 'trigger source' signal exceeds the 'trigger level'. 7902$ 2 12-36 — elektor december 1973 TV scape — extended version 1 5 V/25 m A result is obtained if a 3 mm 0 green LED is used; if it is mounted on the copper side of the board it will fit in the corresponding hole in the front panel. Input timebase The input timebase is the generator for the input clock signal, u sam pi e , The frequency of this signal determines the sampling rate of the input signal{s) to the TV scope. As described earlier, u sample a square- wave signal at twice the frequency required for sampling and storing the input signal(s). The circuit of the input timebase is given in figure 6, The upper portion of this diagram is the clock generator proper. This is a CMOS multivibrator. the frequency of which can be selected by means of the six-way switch S2; fine control of the output frequency is provided by the "time/div’ poten- tiometer, PL The clock frequency can be varied between approximately 32 kHz (S2 in position 6) and 1.6 MHz (S2 in position !). The exact values will be given in the calibration procedure. It was mentioned earlier that the input timebase is started and stopped under control of the u e t signal. In practice, two options are provided; timebase synchronised to the trigger pulses (SI in position ‘trigger’) or timebase running continuously (SI in position ‘free run’)* As in the case of the trigger circuit, the Figure 3. The relationships between the various in- and output signals of the control circuit. Figure 4, The trigger circuit, which derives the Uf r pulses from the analogue 'trigger source" input. The corresponding printed circuit board also contains the "trace distance" control, P2, since it happens to fit neatly at this point on the front panel. Figure 5. The Hrigger board' {EPS 9969-2L All components shown in figure 4, with the exception of S2 and Cl, are mounted on this board; the miniature LED, D2, is mounted on the copper side of the board. Figure 6. The input timebase is a CMOS square-wave generator; the frequency is determined by a selector switch and a fine frequency control. The remaining gate in the same 1C, N4, is used to derive the u x pulses from the u rese t signal. "input timebase’ printed circuit board actually contains a further circuit that has nothing to do with the timebase proper. N4, with its associated com- ponents, derives the u x signal from the u rese t signal from the main board of the basic version. As illustrated in figure 3, u x determines the beginning of each display cycle. Effectively, therefore, it determines the position of the display along the ‘time* axis. Reasonably enough, the potentiometer that deter- mines the length of the u x pulse is labelled *x-pos ? for "x-position’ (P8). Both output signals from the input timebase p.c. board are fed to the "memory board’. A printed circuit board and component TV scope — extended version elektor decern foer 1978 — 12-37 1 Parts list trigger circuit Resistors: R1 r R3 = 100 k R2 = 390 k R4 = 4k7 R5,R6 -10 k R7 = 680 k R8 - 47 O R 9 * 330 n PI = 2k2 (2k5) lin potentiometer P2 - 22 k (25 k) lin potentiometer Capacitors: Cl = 100 n C2,C3 = 1 ju/25 V tantalum C4 = 3p3 C5 = lOOp C6 - 1 g Semiconducto rs: IC1 = 741 IC2 = 709 D1 - DUS D2 = LED (3 mm, green I Sundries: 51 = 1-pole, 3- way switch 52 = single-pole, single-throw switch CA © 15 V/20 mA -© IC1 lOOn 0 0 layout for the input time base and x-position circuit are given in figure 7 . Switch SI is mounted off-board. The adjustment procedure for the six preset potentiometers will be described later. Meanwhile, as with almost any other circuit, it is advisable to set all presets to the mid-position for the time being. Memory board It was already fairly clear from the block diagram shown in figure 1: most of the extension circuits for the TV scope are contained on one board. The "memory board’. Two circuits are mounted on this p.c, board: the control circuit (figure 8a) for the memories, and :he bucket-brigade memories with their ass oc ia te d c ircuitry ( f igu re 8b), The control circuit is a fairly straight- forward logic circuit using CMOS ICs. Like most logic circuits (and in contrast to most analogue circuits) it receives a multitude of input signals and produces a similar plethora of outputs. From the main hoard {basic version) it receives three timing signals, Q 3s Q 4 and Q n , and u rese t (both in normal and in inverted format); the input time base provides both u x and u sam p|e, deter- mining the start of the display and the sample-rate of the store cycles, respect- ively; finally, the trigger circuit provides a signal, ut r > which determines the start of the store cycle. From these input signals, the control circuit derives several output signals. The clock signals for the bucket-brigade memories, pi and 02, both n on-inverted and inverted (01 and 02) and the mode signal, u m , plus its inverted form (u^). In addition to these six control signals for the memories, two further outputs must be obtained: Ug a t e , which controls the white -level pulse gate on the main board, as described earlier; and u e j, which enables the clock generator on the input time base board. Operation of this part of the extension circuit can be considerably simplified by referring to the pulse diagram shown in figure 3. This diagram shows the re- quired relationships between half the input signals to the control circuit and 12-38 — alektor decamber 1978 TV scope — extended version Figure 7, The "input timebase board' (EPS 9969-3), The components shown in figure 6, with the exception of SI , are mounted on this board. Parts list input timebase Resistors: R 1 - 22 k R2 ~ 100 k PI = 47 k (50 k] tin potentiometer P2 - - . P7 = 47 k (50 k) preset PS - 100 k I in potentiometer Capacitors: Cl = I8p C2 = 100p C3 = 560 p C4 = 100 o C5 - 22 n Semiconducto rs: IC1 = CD 4093 Sundries: 51 = 1-pole, 2-way switch 52 = 2-pole, 6-way switch Table 1 DC test voltages at the points indicated in figure 8b Note that these voltages apply after initial adjustment has been completed - see text. test point DC voltage set by means of 5.5 V |S1, 5.5 V 3 V 3 V P4 on the corresponding input amplifier board —3 V it Nil = IC6 =4081 N12. . . N15 = IC7 =4011 M l 6. . . N2Q = ICB = 4040 m uet -c# 79025 33 © © 1 <3 — II II •^-®s|©— 1 ^11^ CM II n U IN IN 1 § 0 — the output of N1 1 (the u e t signal) to go high, starting the input time base oscil- lator (figure 6). The output from the latter, ^sample? is passed through a flip/flop (FF2) to obtain the input dock pulses. Depending on the mode signal, this pulse train is either fed to the 01 or 02 output to clock the mem- ory which is in the head' mode. Simul- taneously, the pulses are counted by IC3; when this reaches its maximum count of 256, its output goes high. This causes the u e t output to go low, stop- ping the timebase oscillator and, with that, the head 1 cycle. In the absence of trigger pulses, no head 1 cycle would normally occur. How- ever, approximately halfway through the total period Q n goes high. This has the same effect as a trigger pulse, enabling the input timebase to initiate a read cycle. Effectively, the scope switches over to a Tree run' mode in the absence of trigger pulses. So much for the control circuit. Not surprisingly, the second part of the ‘memory board 1 contains the memories (figure 8b). The portions enclosed within dotted lines axe identical: the upper section is for the A-channel and the lower for the yb input. The latter section can, of course, be omitted if only a single -channel scope is required. The heart of the circuit is IC9. This contains two independent bucket- brigade delay lines, each with_512 ‘buckets’. Clock signals 01 and 01 are fed to one of these delay lines; 02 and 02 are used to clock the other. In Figure 8. Two closely related circuits are mounted on the 'memory board 1 : the control circuit tfigureBal and the memory circuit proper (figure SbK For a two-channel version of the scope, two memory circuits are re- quired; in figure Bb both of these circuits are shown in full. Principle, both delay lines could receive the same input signal; in practice, however, u s b (for ‘shifted) is not quite identical to Uy. As is apparent from the circuit of the input amplifier (figure 2 in the article dealing with the basic version - see page 10-31 in the October issue), both the signal amplitude and the superimposed DC level of the u s h signal can be varied slightly with respect to u y . This is done to compensate for any minor differences between the two delay lines, as will be described for the calibration procedure. The two delay lines in 1C9 each have tw r o outputs. Mixing these two outputs in a resistor network helps to suppress the clock components, providing a total analogue output (at point 3 or 4, respectively) that is relatively ‘clean 1 . Under control from the ‘mode 1 signals (u m and n^), the analogue switches contained in IC10 select one of the two memory outputs for display purposes. The ‘unwanted* output is connected unceremoniously to ground. The selected output is fed through a sixth-order low-pass filter, A1 . . . A3, with a turnover frequency of 2.5 kHz. This frequency corresponds to the highest frequency the basic version of the TV scope can be expected to display. At the end of the chain (at point 5) a clean, time- expanded replica of the original analogue input signal is now available for display purposes. Or rather, not quite. The A-channel output is amplified slightly (approximately Limes three) and a DC component is added to R 29 TV scope - extended version elektor decern bar 1970 — 12-41 10 0 © x CM U> © 40 *■* o o > g £ O (O o ® flH=M I CM V eo G* ‘in r- u 1! c o o ► n> <> <> O ■— “ Si u?i CM CO CM CO i#£>< ^ - i J i r\ c~ E[! 1 3 JL -A a \_r V. i i A p V K 2 x LJ ° s 1 yfi *“ Q O- tft 0> 7 l( 3 3 © to oS * > u> ? il c O o O XI HI •£ bn.' : S3 n " i E| 53] -’■I ^ (SJ ^ I ^ r & 2 £ n S 25 ; ; : o a ! ! ' „ * _ J P 2 O | 9 f* 5 s| - * i < < a D t ■ f • i i^ju _c. jl! 3 3 o to X -0- ,J CM a tO 10 lO CM CE CM CO m O il o 0 °8 H § -6-6- _ [ h- * o II c o o 05 9 Si hS° O isO^OtS'O A A A A [fl ® ¥ st£ t 2 fi’SL a < as Z z « z j- I- O - 3 2 > s> © fi 79025 Sb 12-42 — elektor december 1978 TV scope — extended version TV scope — extended version elektor decern be r 1978 — 12-43 Table 4 Supply currents (mAl circuit board no. + 15 V -15 V input amplifier* 9968-1 5 ... 9 5 . . .7 main board 9968-2 45 20 video mixer 9968 3 9 memory board 9969-1 22 5 j trigger circuit 9969-2 25 20 ... 25 input timebase VHF/UHF 9969-3 20 — TV modulator 9967 15 sync circuit** 9968-4 4 * two required for two-channel version of TV scope ** only required for basic version of TV scope Figure 9. The 'memory board' {EPS 9969’ 1}, Insulated wire should be used for the wire links wherever there is any danger of inadver- tent short circuits. Furthermore, the resistors are mounted 'on end' in order to keep the &i*e of the board within reasonable limits. Parts list memory board Resistors : R1 = 33 k R2 - 22 k R3,R3\R4 ( R4' h R7 # R7', R8 f R 8' - 470 H R5 r R5',R6,R6 r ,R9,R9 r , R1G.R1Q\R29 h R29', RSOjRSO' = 1 k R11 l Rtr,R!2,Rt2' # R14 . R 1 6 ,R 1 4 ' . . . R16', R18 , . * R20 J R18' , , * R2Q\ R22,R22' - 10 k R13,R13',R17,R17', R21 r R21' = 15k R23,R23',R25,R25' = 100 k R24,R26 = 270 k R24' = 330 k R27 = 1M5 R28 f R28' = 10O PI' = 10 k preset P2,P2' = 47 k (50 k) preset Capacitors: C 1 ,C 1 2;C 1 2 ' = 10 n C2 - 47 p C3,C4,C6 = 22 n C5 ( C7 P C14,C14\ Cl5 f Cl 5 J = 100 n C8,C8',C1 0,C1 0' = 6n8 C9,C9',C11,C11' = 2n2 C13,C13' = 1 n C16 f C1 7 - 100 m/16 V Semiconductors: D1,D1',D2,D2'= DUS IC1 JC2JC7 = CD 4011 IC3JC5 = CD 4020 IC4 = CD4Q13 IC6 = CD4081 ICS = CD 4049 IC9,1C9' - SAD 1024 (Reticon) IC10JC10' = CD 4066 IC1 1 JC1 1' » LM 324 Sundries: SI = single-pole, single-throw switch Table 5 Calibration data for the input timebase (see text} time/div. preset calibration signal IC5 (main board) pin no. calibration signal number of divisions * sample frequency period time 2 ms P7 Q8 13 480 Hz 2.08 ms 1.0 31.98 kHz 1 ms P6 Q7 4 960 Hz 1 ,04 ms 1.0 63.96 kHz 500 jis P5 Q6 2 1.92 kHz 520 ms 1.0 127.9 kHz 250 jus P4 05 3 3.84 kHz 260 m s TO 255,8 kHz 100 ms P3 Q5 3 3.84 kHz 260 2.6 639.6 kHz 40 ms P2 Q3 6 i 15.38 kHz 65 ms 1.6 1599 kHz ] produce the final output, u oa , which is fed to the input of the main board. The output of the B-channel undergoes similar treatment, with one minor difference. The fixed gain is slightly greater (R24 1 is slightly larger than R24), but this output level can be reduced by means of PIT In effect, therefore, the total gain at the output of the B-channel can be adjusted to compensate for any minor differences in the attenuation caused by all the preceding stages. The adjustable DC shift set by P2 brings the output level into line with the sawtooth reference voltage on the main board . One last point remains to be discussed: the Trace distance 1 control, P2 in figure 4, which was mentioned briefly in connection with the board for the trigger circuit. The DC output from this potentiometer, is connected via R26 and R27 to the non-inverting input of the A-channel output amplifier and the inverting input of its partner in the B-channel. The term Trace distance 1 aptly describes the function of this control: it shifts the two traces in opposite directions on the screen. At one end of the control range, the two traces will overlap (useful for com- parison of minor differences between the two), whereas at the other end of the range the two traces are each approximately centered on their own half of the screen . All the circuits shown in figures 8a and 8b are mounted on a single p.c. board, the memory board, as given in figure 9. In order to pack such a large number of components on a relatively small, single- sided board, the ease of construction has had to be sacrificed to some extent. Insulated wire will be required for several of the wire links, to avoid the possibility of highly undesirable short circuits. Furthermore, in marked con- trast to most Elektor boards, the resistors are mounted ‘on end 1 and care should be taken to avoid shorts between the long connections to adjacent re- sistors. Switch SI Cx magn 1 ) is mounted off board, as shown in figure 1 0, The components for the B-channel normally have the same values as their opposite numbers in the A-channel (e.g. R3 = R3 1 = 470 £2). There are two ex- ceptions to this rule; R24 (A-channel) - 270 k, whereas R24' (B-channel) ” 330 k; furthermore, preset PI 1 (B-channel) is not included in the A-channel circuit. If a single -channel version of the TV scope is required, all components for the B-channel may be omitted. For test purposes, some voltages at the various test points are listed in Table 1 , Further details are given in the descrip- tion of the calibration procedure. AH the bits , . . The three boards described in this article, together with those already described for the basic version, are all that is needed for the extended version of the TV scope. As an aid in procuring the necessary components, all the relevant parts lists have been com- pressed into one complete "bulk parts list 1 : 146 resistors, 10 potentiometers, 20 preset potentiometers, 82 capacitors, 34 ICs, 24 other semiconductors, 1 1 switches and a few "sundries'. For each component value or type, the quantity required for the complete extended version of the scope (basic version plus extension circuits) is listed 12-44 — elektor decern ber 1978 TV scope — extended version - . r 1 p f : : : ■ ' ■ ■ ' *. 1 : — ^ A ^ B Rb I TOO Ra ioo n :*M*r*w ,/Wjw: iMtectntt i^^r> T*:.- • VOpau o Ca 220w/ie V 220 m/1© V MKm^hTa qwgoou X; X"* ... .£ nl: 1 » MH 9 TV scope — extended version eiektor december 1978 — 12-45 Figure 10. Wiring diagram for the complete TV scope, extended version. For those connections which require screened cable, it should be noted that the screening is only connected to ground at one end. A complete list of all interconnections is also given in Table 8. 12-46 — elektor decern br 1978 TV scope — extended version TV scope — extended version eiektor decern ber 1978 — 12-47 in one column; a second column gives the quantities required for extending the basic scope. Both columns are valid for the complete two-channel TV scope. To avoid misunderstandings: in the first column ('complete*}, the components required for the UHF/VHF TV modu- lator are also listed (see the recent October issue, page 1 0-27). This was not the case in the bulk parts list for the basic version of the TV scope given in the October issue. Back to the basic version During extensive testing of the extended version of the TV scope, it was dis- covered that some samples of the analogue delay -line ICs exhibited notice- able anaomalies in their signal-storage characteristics. This effect is less appar- ent if larger signal levels are used. For this reason, one or two minor modifications of the basic version are required if it is to be used in conjunc- tion with the extension circuits. Note that there is no advantage in incor- porating these modifications in the basic scope if it is not to be extended - although they won't do any harm, either. As shown in Table 2, three resistor values are changed: two in each y-input amplifier (four resistors in total for a two-channel scope) and one resistor on the main board. These components are included in the bulk parts list. The effect of the modifications in the input amplifier is to significantly increase its output level, making it more suitable for processing in the analogue delay lines, A minor disadvantage is a slight reduction in bandwidth: - 6 dB at 100 kHz, instead of the original — 3 dB. Since the input signal to the main board is now at a higher level, the amplitude of the reference sawtooth must be increased in step. This is accomplished by modifying the value of one resistor on the main board. Construction The connections between the various boards are shown in figure 10. The position of the boards corresponds to the lay-out of the front panel shown in figure 1 1 . The front-panel controls are listed in Table 6, with a brief descrip- tion of their function. Construction of the complete unit is basically similar to that of the basic version, as described in the October issue. Suitable cases are available from West Hyde (type DES 306} and GSA i type V7G04). As with the basic version, it is advisable to screen the more sensi- tive circuits {input amplifiers, trigger circuit, input timebase and video mixer) with strips of copper laminate board which can be soldered to the earth plane. The interconnections between the vari- ous boards are also listed in Table 3, For those connections which require screened cable, it should be noted that the screening is only connected to apply common at one end, to avoid ground loops. As shown in figure 10, the supplies to the input amplifiers are decoupled by means of two resistors, R a and Rt> (100 £2), and two electrolytic capacitors (C a and Cfo, 220 /i/16 V). These com- ponents are mounted 'off board', as shown. When comparing the basic version with the extended version, it may be noticed that the 'reset 1 output is omitted from the latter. This provision is not required in the extended version, since it has adequate triggering facilities. However, if the 'reset 1 output is already mounted there is no point in removing it. The two IC voltage regulators in the main supply should be adequately cooled, for instance by mounting them (with mica insulation!) on the back of the case. The main reason for this is to keep the heat out of the case: exper- iments have shown that the perform- ance of the bucket-brigade delay lines, in particular, deteriorates rapidly as the ambient temperature rises. There is no need to cool the regulator on the UHF/VHF modulator board. A further aid when constructing and testing the TV scope is the list of supply currents for the various circuits, given in Table 4. The current consumption of the complete TV scope (extended version, two -channel, including the VHF/UHF modulator) is approximately 150mA from the positive supply (+15 V) and 55 mA from the — 15 V supply. Before applying power for the first time, it is advisable to first give the wiring a final 'once-overi, checking for inadvertent short circuits. Some photos of the completed unit are given in figures 13 , , . 15. To play things safe, the supply can be connected to the various boards one at a time: first disconnect the output from the main supply, apply power and check the + 15 V and —15 V outputs. The other boards are then connected to the supply one at a time, checking the supply current against Table 4. Differences of up to 1 0% are no cause for alarm. Once all the boards have been connec- ted up and checked in this way, it is time for the final step: the calibration procedure. Initial calibration In spite of the 20 presets involved, calibration of the TV scope is not too difficult. The only measuring equipment required is a standard universal meter, with a sensitivity of at least 1 0 k£2/V. The first step is to calibrate the com- ponents associated with the basic version of the TV scope. The relevant adjustment procedures were discussed in the article in the October issue. If one has already built and calibrated the basic version and is now adding the extension circuits, the only adjustment affected by the component modifi- cations according to Table 2 is the setting of PI on the main board, How- Figure 11. Front panel layout for the ex- tended version of the TV scope. Table 6 lists the front-panel controls, with a brief descrip- tion of their function. Figure 12. A word of warning! The bucket- brigade delay-lines have a limited dynamic range, and excessively large input signals will be clipped, as shown here. Figure 13, Detail of the interior of the proto- type. To the left of the picture, the main supply regulators are mounted on the back panel. 12-48 — elektc December 1978 TV scope — extended version -f'v *!> 1 i '* 1 • SW’'^ Wm -i: II; # i # 11: IP P iNpyilPPil- ever, final adjustment of this preset is part of the calibration procedure about to be described. The setting of all other presets in the basic version remains unaltered, with the exception of P3 and P4 on the input amplifier board(s). These two potentiometers have no influence on the operation of the basic version. Having completed the adjustment pro- cedure outlined for the basic version, the next step is to adjust P4 in the input amplifier. For this initial (approximate) adjustment, no input signal is required. The controls of the y-amplifier are set as follows; ‘volts/div’ selector switch a 10 V and corresponding potentiometer fully clockwise ('cal*); ‘AC/DC 1 switch in position 4 ACT Initially, set P3 and P4 in their mid -positions. Measure the voltage at test point 1 (or I") on the memory board, and adjust the l y-^os" control to obtain 5,5 V at this point. Now measure the voltage at point 2 (2 T ) and adjust P4 until 5,5 V is obtained at this point also. This adjustment is sufficiently accurate for the time being. Operation of the bucket -brigade mem- ories can now be checked. To this end, the Trace distance" control is turned fully anti-clockwise; the Time/div" switch is set to 500 pts, and the corre- sponding potentiometer fully clockwise ("cal’); the Trigger /free run" switch is set to ‘free run 1 ; all presets on the input time base board are set to their mid- position. The voltages at test points 3 and 4 (3 1 and 4") should now be ap- proximately 3 V; furthermore, it should be possible to vary these voltages over a total range of about 1 V (e.g. from 2.5 V to 3.5 V) by means of the "y-pos" control. If this is the case, the memories are in working order . The next point to check is the voltage at point 5 (5*). This should be approxi- mately equal in value to the voltage at points 3 and 4 (3* and 4 1 ), but negative with respect to supply common. In other words, if the voltages at points 3 and 4 are indeed 3 V, the voltage at point 5 should be approximately —3 V. This being the case, the next step is to measure the voltage at point 6 (6 5 ) and set it to approximately +4 V by means of P2 (P2 1 ). For the time being, PI 1 is set to the mid-position. The output of the TV scope should now be connected to a TV receiver. PI on the main board (basic version) is ad- justed until vertical lines appear that are not part of the calibration graticule. In principle, there should be four of these lines, but it is possible that one or more of them are actually off the screen, if this is the case, the missing lines can be brought into the picture by adjusting P2 (P2 1 ) on the memory board. One pair of lines is generated by the A-channel. This pair can be found by operating the corresponding 4 y-pos" control: the lines can be shifted over a total range of approximately 4 scale divisions, or just less than half the width of the screen (the range is limited by the maximum output swing of the bucket- brigade memories). The y-position controls for both channels are now set so that the corresponding vertical lines are mid-way between their two extreme positions. Finally, P4 on the input amplifier board(s) is adjusted until the corresponding pair of lines become one: the zero-volt line for that particular channel. This completes the initial setting-up Figure 14, Complete unit, with all wiring and screening panels in place. Figure 15. A 10 kHz square-wave on an ancient TV set and a 20 kHz sine-wave on a more recent model . . . procedure; the next step is the final calibration. Final calibration If a dual-channel version of the scope is being calibrated, it is a good idea at this point to adjust P2" on the memory board so that the zero-volts line for the B-channel is moved to the edge of the screen. For the present, the A-channel display is the only one which is of interest. The first step is to calibrate the input timebase. For this, a stable and accurate AC test signal is required. Fortunately, no expensive test equipment is required: a suitable signal can be derived from the TV scope itself. The input of the y a input amplifier is connected to pin 2 ( i Q 6 *) of ICS on the main board (basic version). Since this is a digital IC and the relevant oscillator is crystal- controlled, the signal at this point is known: it is a square wave with a peak- to-peak amplitude of 15V and a fre- quency of 1.92 kHz, corresponding to a period time of 520 juts. With the Trigger source" selector switch set to Va\ should now be possible to adjust the Trigger level" control so that the Trig"d" LED lights. This adjustment is fairly critical. The Trigger/ free run" switch is now set in position Trigger", whereupon a stable picture should appear on the screen. The square-wave trace can be moved horizontally and vertically across the screen by means of the V-pos" and Tt-pos 1 controls, respect- ively Note that the y-position control should be adjusted so that the trace is centered in its range, as otherwise the signal will be clipped by the delay-lines. At this stage, the display will almost TV scope — extended version elektor decern ber 1978 — 12-49 Table 6 Front-panel controls function part no. figure no.* remarks power SI B14 mains switch signal intensity PI B7 intensity of trace grid intensity P2 B7 intensity of calibration graticule x-pos PS E6 trace position on x-axis time/div** S2 h P1 E6 x-axis scale x-magn SI E8a "lx J ; scale corresponds to that selected by 'tirne/div' switch; f 2x': x-axis scale multiplied by 2 trigger/free run SI E6 trigger circuit enable; select 'free run' in absence of trigger signal trigger source SI E4 selector switch for trigger signal trig'd 02 (LEDIE4 indicator LED; lights if trigger signal is present trigger level PI E4 reference level for signal selected as 'trigger source' trace distance P2 E4 sets distance between the two traces for A- and B-channef y-pos P2 B2 trace position on y-axis volts/div** SI, PI B2 y-axis scale AC/DC S2 B2/E4 AC or DC coupling of input signal at respective Input * figure numbers preceded by the letter 'B' refer to figures in the article 'TV scope, basic version' in the October issue; figure numbers preceded by letter *E' refer to figures in this article , ** The values given for the various switch positions are only valid if the corresponding potentiometer is in position 'cal'. certainly consist of two square- waves, one corresponding to each of the delay- lines in the A-channeh The differences between the two signal paths can now be compensated, by means of presets P3 and P4 in the A- input amplifier. The slight difference in gain between the two delay-lines can be compensated for by adjusting P3 until the amplitude of the two square-wave signals is identical; the varying DC shifts are brought into line by setting P4 so that both traces are symmetrical around the same centre-line - in other words, the y-position of both traces should be identical. Since these adjustments interact , this adjustment procedure must be repeated until no further improvement is obtained. Having reduced the display to a single square- wave signal, the next step is to calibrate the Input time base. With the Time/div 1 switch in position 500 /is and the corresponding potentiometer in position ‘cal 1 , one complete period of the square-wave should correspond to a fraction over one division — 520 fjs is 1 .04 divisions. The preset potentiometer P5 on the input timebase board is adjusted until the correct period length is obtained. The ‘voits/div 1 selector switch should still be set at TO V’. This means that the 1 5 V peak- 1 o-peak square- wave should correspond to an amplitude of 1.5 divisions. This result can be obtained by (re-) adjusting PI on the main board of the basic version. It will be noticed that this adjustment also influences the y-position, but this shift can be compen- sated for by readjusting P2 on the memory board. The same adjustment procedure can now be carried out for the second (B-) channel, after first bringing the corre- sponding ‘zero-volts line* back into the centre of the picture by re-adjusting P2 1 on the memory board. There is, how- ever, one difference in the calibration procedure: the amplitude setting (for 1.5 horizontal divisions) is carried out by means of PF on the memory board, instead of using PI on the main board. For the B-channel, the procedure can therefore be summed up briefly as follows: P3 and P4 on the B-channel input amplifier are adjusted until the two square -wave signals overlap; the timebase, obviously, needs no re- adjustment, since if is common to both channels; P V on the memory board is so that the amplitude of the square-wave signal corresponds to L5 horizontal divisions; finally, P2 : on the memory board is adjusted to compensate for DC shift (y-position). Final adjustment of P2 and P2’ is now in order. The ‘trace distance’ control should still be set fully anti-clockwise (‘overlap’) and the ‘y-position 1 controls are both set so that the corresponding traces (with no input signal applied) are midway between their two extreme positions. P2 and P2 1 can now be adjusted so that the corresponding traces overlap in the exact centre of the screen. The remaining presets in the input time- base now require attention - P5 has already been adjusted. The other presets are calibrated in the same way — the Trigger level 1 control being manipulated to obtain a stationary display. Suitable calibration signals are derived from the TV scope itself, and the relevant preset on the input timebase board is adjusted until the correct period length is ob- tained. All timing signals are derived from ICS on the main board. All rel- evant details are given in Table 5: the position of the ‘time/div* selector switch; the corresponding preset poten- tiometer; the necessary calibration sig- nal, and the pin number (of ICS) from which this signal can be obtained; the frequency and period time of this signal; the number of divisions that corre- sponds with one complete period of this signal when the preset is correctly adjusted. The last column in this table gives the correct frequency (f sample) of the u sam pje signal, for the benefit of those who have access to a frequency counter: setting the ‘trigger/ free run 1 switch in position ‘free run 1 causes the timebase oscillator to run continuously, so that it can be measured and cali- brated accurately without the need for recourse to calibration signals. For completeness’ sake, the calibration data for P5 are also listed in Table 5, even though this adjustment has already been completed at an earlier stage, A possible complication in the ‘2 ms/div 1 position is that it may prove impossible to obtain a stationary display. The reason for this problem is that in this case the ‘read 1 cycle may take so long that it overlaps the ‘display 1 cycle - the memories are read out before they have been completely read in! In this case, triggering is not possible and the ‘trigger/ free run 1 switch may as well be set in the ‘free run 1 position. The adjustment of the input timebase was the last step in the calibration procedure. It may, however, prove useful to give all adjustments a final one e-over. In particular, presets F3 and P4 in the input amplifiers may require 12-50 — elaktor december 1978 TV scope — extended version Bulk parts list for extended version of TV scope {two input amplifiers, main board, video mixer, input timebase, trigger circuit, memory board, VHF/UHF TV modulator and supply — note that the sync circuit is not required for the extend version) ‘complete" {second column) gives the total number of each component type required for the TV scope in extended version; "extension only" (third column) lists the components required to extend the basic version Linear potentiometers: value complete 2k2 (2k5) 1 10 k 2 22 k {25 k) 1 47 k (60 k) 1 100 k 3 220 k (250 k) 1 470 k (500 k) 1 Preset potentiometers: value complete Resistors: 2k2 (2k5) value com pi ate extension 4k7 (5 k) only 10k 10 a 3 2 47 k (50 k) 47 a 4 1 68 n 1 ioo a 150 a 4 1 2 Capacitors: 220 n 2 1 value 270 n 1 330 n 1 1 3p3 470 a 10 8 8p2 820 a 2 15 p 1 k 21 12 18 p 1k5 1 22 p 2k2 1 33 p 3k3 1 47 p 4k7 1 1 82 p 5k6 4 2 100 p 6k8 1 120 p 3k2 4 220 p 10 k 26 20 560 p 12 k 2 1 n 15k 12 6 2n2 18 k 2 6nS 22 k 3 2 10 n 33 k 3 1 22 n 47 k 2 100 n 82 k 6 1 M 100 k 15 7 1 ju/25 V 220 k 2 tantalum 270 k 2 2 2p2/25 V 330 k 1 1 tantalum 390 k 1 1 4^7/35 V 6S0k 1 1 tantalum 820 k 2 100 ju/16 V IMS 1 1 220 W1 6 V 2M2 2 2 470 W35 V 2 2 2 4 10 complete 1 3 2 1 1 2 1 1 5 1 1 1 2 4 4 5 4 24 1 10 1 2 2 2 extension only 1 1 1 1 extension only 1 8 extension only 1 1 1 2 1 2 4 4 3 4 8 1 2 2 2 Semiconductors: type complete CD 4011 6 CD 401 2 1 CD 401 3 2 CD 401 7 1 CD 4020 2 CD 4040 1 CD4049 1 CD 4066 2 CD 4068 1 CD 4071 1 CD 4081 1 CD 4093 1 709 3 741 2 LM 324 2 TL 084 2 SAD 1024 2 78L05 1 7815 1 7915 1 TUN 3 TUP 1 BF 494 3 BFY90 1 DUS 9 1 N 400 1 4 1N4148 1 LED 2 extension only 2 {3}* 1 2 1 2 1 1 1 1 2 2 5 1 switches; type complete extension only single-pole, single-throw 4 double-pole, singi e-throw 1 double-pole mains switch 1 single-pole, 3-way 2 single-pole, 4-way 2 2-pole, 6-way 1 2 2 1 sundries (complete); 4,433 MHz (colour TV) crystal 1 27 MHz (approx,) crystal 1 1 /uH mini-choke 1 100mA fuse 1 2x1 8 V/250 mA mains transformer 1 1 1 1 1 1 4 2 off if the CD 401 1 originally mounted on the sync board can be re-used on the memory board, otherwise 3 off. minor re-adjustment after the ‘bucket- brigade" ICs have warmed up. These chips have proved to be rather sensitive to changes in ambient temperature, with shifts in output level that vary from one sample to the next. This may result in trace flicker’ after the unit has warmed up (actually, two separate traces are displayed very close together, giving a flickering effect) but this can be elim- inated by slight re-adjustment of P3 and P4. A good test signal for this final adjustment is a sine- wave with an amplitude of about three divisions on the screen. 'There's a hole in the bucket' The Reticon SAD 1024 bucket-brigade delay -line is a useful 10, but it is in- tended primarily for audio reverberation systems. Tests have shown that a small percentage of these ICs contain one or more leaky buckets". Since, in the TV scope, the input signal is stored for a short time without being shifted, one or more of the input samples may remain in a leaky bucket for a relatively long period. This results in noticeable 'dips 1 in the trace. For this reason, the signal level has been boosted (by means of the modifications listed in Table 2). The dips are now reduced, relative to the signal, to the point were they are no longer a nuisance. For the same reason it is advisable, if triggering is not required, to select the 'free run" position (e + g. when measuring DC voltages). Since the time base then runs continuously, the contents of the memories are refreshed right up to the point where the display cycle starts, giving the best results. Littemture: Elektor ; May 1978 , P.5-20: TV scope using bucket-brigade memory V Elektor, October 1978: - page 10-03: l an introduction to the TV scope 3 ; - page 10-30: TV scope, basic version r ; - page 10-27: * VHF/UHF TV modu- lator’; - page 10-44: ‘ analogue reverberation unit 3 ; Elektor ; November 1978, page 11-25: 'extending the TV scope*. H extending the elektor piano elektor decern bar 1978 — 12-51 || i> f . | ||p ft wm i i ' . . : ■ j ! Although the compass of the Elektor piano which was published in September of this year (Elektor No. 41) originally covered only five octaves, it was stated then that thanks to its modular design, it could be easily extended to span 6, 7 or even 8 octaves. Judging by the reaction from our readers, it does in fact appear that there is a considerable interest in an electronic piano with a full 8-octave compass, i.e. ninety-six keys and a fundamental frequency range extending from 17.4 to 4148 Hz. For this reason the following short article provides readers with details of the changes in component values and of the additions to the filter circuits which will be required to extend the compass of the Elektor piano. A13,A14,A15 = 741 of 1/4TL074, 1/4TL0B4 1/4 R4212 octave 7 octave € junction of R 6/fl 12/937 junction of R3Q/R56/R5B TL074,TL0S4,XR4212:©=rf4 (U 3 ) 0«rfii {Ui> 741 ; (Ua) Os #4 ) 79027-1 12-52 — elektor december 1978 extending the elektor piano Firstly there is the question of the values of the discharge resistors R1 through R12 on the extra octave boards (EPS 9981). For the highest octave, i.e, octave 8, R1 through R6 = 100 k, and R7 through R12 = 120 k. For octave 6, the second /owejf octave, Rl through R6 - 1 M, and R7 through R12= 1M2, whilst for octave 7, the lowest octave, Rl through R6 = 1M5, and R7 through R12 = IMS. The original board layout for the filter circuits (EPS 9981) was designed for the 5 -octave version. Thus, depending on however many extra octaves are re- quired, it will be necessary to incorpor- ate one or more of the additional filter circuits shown in figure 1 . These will, of course, have to be mounted on, e.g, a small piece of Vero board, and housed in the vicinity of the filter board proper. This seems a useful opportunity to mention an interesting possibility for those readers who, without wishing to redesign the voicing circuits, might like to experiment with altering the sound of the original piano. The following tip has been received from a Dutch reader, a Mr. Hulshoff from Rotterdam, who recommends introducing even harmonics (the second, fourth, sixth harmonic etc. of the fundamental) into the squarewave outputs of the master tone generator, which being symmetrical, are presently composed only of the fundamental and a number of odd harmonics. With the aid of the resistor network shown in figure 2, the master tone generator outputs are not fed straight to the corresponding inputs of the keying circuits; rather they are summed with one, or in most cases, two squarewave signals which successively have twice the frequency but half the amplitude of the •'**»***■»*** t— Gi ® !*■ ip N i- octave 8 octave 1 octave 7 EPS 9915 octave 2 octave 6 octave 3 **•***•*••** octave 4 octave 5 79022 4 3 iCiAtn Lf)iniA M • . mm m llllfill If #• #• j wmm 5 m m m i .... f a ' / :U. .,.. ;• ^ / ' m ' Am: •••••••• " w> ; - •• ^ •••*•••• ••• ,5SS '' vhw Mi . # fe m. ff s® ip/iM p-. f -Ji m 4 - ... ■ ~ * 51 39 27 4 52 40 28 :: RTf A 1 ' ■" 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 38 89 90 91 92 93 94 95 96 9 to fi a i 1 2 3 4 6 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 8 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 n wm ! 59 w 1 # 5 ; 1 " Af * ' m % I.!.!.:. 48 73 61 49 74 62 50 75 63 51 76 64 52 77 65 53 78 66 54 79 67 55 30 68 56 81 69 57 82 70 58 83 71 59 84 72 60 85 73 61 86 74 62 87 75 63 88 76 64 89 77 65 90 78 66 91 79 67 92 80 68 93 81 69 94 82 70 95 83 71 96 84 72 N.S. 1 A standard upright piano keyboard extends from key 4 up to and including key 88. N.B. 2 The shaded section of the table covers those keys included in the 5-octave version of the piano. WC*¥9 1 ociro: 1 OrfB¥E 9 original squarewave, As figure 3 clearly shows, the result is a staircase waveform which has a high proportion of even harmonics. {Obviously, in the case of octave 1, i.e. the second highest octave, it is only possible to sum the corre- sponding master tone generator outputs with one signal of twice the frequency, i.e. that of octave 8 - the highest octave, whilst in the case of octave 8 itself, the above procedure is impossible). The connection points X, Y and Z, for the resistor network of figure 2 are given in the accompanying table. As is apparent, each note is accorded both a key number (a procedure which was necessary to simplify the figure on page 9-10, Elektor 41), an octave num- ber (I . . ,8) and a number which indi- cates its position in the octave. Finally, to prevent any lingering con- fusion figure 4 once more shows the component layout of the board for the master tone generator. As most readers soon spotted the original layout on page 9-1 1 of Elektor 41 transposed the indications for notes 1 to 1 2 on all hut octave 4. However this fault does not appear on the actual boards. M 12-54 — elektor december 1978 marks! missing link Modifications to Additions to Improvements on Corrections in Circuits published in Elektor CMOS FSK modulator ? Mini push button Digi tat raver be r at i a n tin i t Elektor 37, May 1978, p. 5-08. Three possible pin-compatible equivalents are listed for the 1024-bit shift register. However, when using the MM5Q58 the dissi- pation may in some cases be on the high side. The reason for this is the clock pulse level: 12 V top- top, whereas this particular 1C is quite happy with only 5 V, A simple circuit modification (only required if the MM5058 is used!) FET audio amplifier Summer Circuits 1978, circuit no. 39, Components C5, C7, R14 and D9 are shown in the circuit as connected to supply common. However, they should be connec- ted to the R7/R25 junction, as shown in the diagram below. A SC 1 I keyboard November 1978, page 11-06. The parts list should read: R1 - 680 k; R2 - 100 k. The keyboard switches are listed as type MM9; the full type number is MM9-2, In Table 3, the last few Summer Circuits 1978, circuit no, 72. The reset connection of the 4020 (IC1) should be connec- ted to supply common, instead of to positive supply. Master tone generator September 1978, page 9-09. The M 087 is an SGS/ATES type; not Motorola, as stated in the article. Our apologies for any incon- venience this may have caused. Data bus buffer October 1978, page 10-18. On the component layout (figure 3) the and "0 1 supply connections are transposed: the *+’ connection runs along the outer edge of the hoard to pins 20 of 1C l and IC2, Tag! November 1978, page 11-12. P3 should be a 2k2 (or 2k5) linear potentiometer. Ring the ball and win a prize! November 1978, page 11-22, The output voltage tip, referred to at several points in the text, is present at the emitt er of T 2 (point *M*). Furthermore, the component numbers given in the parts list are incorrect; the cor- rected list reads: Resistors: R 1 f R 4,FH 9 r R20,R22 = 10 k R2,R3,R9,R1Q,R12 r R 1 7,R21 = 100 k R3 - 33 k R5,R6 = 2k2 R7 - 220 k R 1 1 ,R 1 4,R 15= 1 k R13 = 180 k R 1 6,R 1 8 = 15 k R23 = 220 H Pi ,P3 r P4 = 100 k preset P2 = 10 k bn Capacitors: Cl r C7 r C8 - 100 n C2 P C5 = 4n7 C3 = 1p5/16 V C4 = IOOu/16 V C6 r C9 = 220 n = CR + erasure to end of line =■ cursor <- = scroll up = CR (carriage return) without erasure = erasure of current line = home cursor switches A range of miniature snap-action push button switches, designed for printed circuit board mounting on a 0.1 inch (2.54 mm) hole grid, is now available from Impectron Ltd. Switches in the Petrick 320 range feature attractive, ergonomically designed faces and a positive snap or rocker action. Single pole press-to-break and changeover configurations axe all available as standard, and an LED indicator can be incorporated to provide visual warning of function selected. The basic body size of all types is 12.4 x 12,4 x 7,5 mm (approx. 0.5 x 0.5 x 0.3 inches) with 4 mm (0.16 in) long connection tags. The selected push button or rocker action will add a further 5 mm to body depth. This space saving design allows very compact keyboards and control panels to be constructed economically. Switches may also be mounted in line or block format to form key sets with mechanical interlocking and release mechanisms. Despite their small size the switches are capable of handling up to 25 mA at 50 V DC. Contact resistance is less than 20 milli- ohms and operational life is in excess of 10 s operations at full load. Operating temperatures range is -25" to+75°C. Impectron Limited, Impectron House, 23-31 King Street, London TV5 9LH (984 M) C-MOS 64-bit memory RCA Solid State has introduced a new compliment ary MOS 64-bit random-access memory, the CD40114B, which is equivalent to, and pin-compatible with, industry-standard TTL types such as the 7489, but offers the additional C-MOS advantages of low power consumption, high noise immunity and wide supply voltage range (20 V rating). The CD401 14B is equivalent to the C-MOS type 74C89, The CD40114B is a 16-word x 4 -bit random -access memory w ith four address inputs, four data inputs a Sv rite-enable* input, a ‘memory-enable 1 input, and four 3 -state data outputs. The memory -enable input and the 3- state data outputs allow memory expansion, and the four address inputs are decoded internally to select one of the 16 possible w r ord locations. An input address latch facility is included; the address information is latched on the negative edge of the memory-enable input by an internal address register. The selected output assumes a high- impedance condition when the device is waiting or disabled. RCA Solid State , Sunbury -on- Thames, Middlesex , TW16 7 HW (981 M) DC/DC converters Gresham Lion Limited have added a further series of miniature encapsulated DC/ DC converter units to its popular GEMINI range. Designated the Gemini 800 Series, the new additions will extend the existing range by 24 units, making it one of the most comprehensive currently available. The 800 series consists of 16 single output and eight dual output Converters operating from nominal DC inputs of 5, 12, 24 or 28 volts. Designed to operate over wide input voltage ranges, these 10 and 1 2 watt units provide excellent line/load regulation and high efficiency (65% typical) even with output loads as low as 10% of rated output. All models employ Pi-type input filters to minimize reflected input ripple current. EMI/RFI protec- tion is provided through complete six-sided continuous shielding incorporated into the design of each unit. In addition, all units in the range feature output current limiting and output short-circuit protection for up to eight hours, with automatic restart wiien the short circuit is removed. The units will also start with reverse polarity current injected into the output. The units are designed for PCB mounting, they measure 89x63x22 mm (3.5 x 2 ,5 x 0.88 inches) (L x W x H), and will operate in the temperature range -25 to +71°C. Gresham Lion Ltd., Gresham House, Twickenham Road, Feltham, Middlesex. (927 M) lines are incorrect. The corrected section of the table is given below. CTL+ M = CR (CARRIAGE RETURN) CTL+ H = BS (BACKSPACE) CTL + [ = ESC (ESCAPE) CTL+ ] CTL + Z = SUB K = FS (FILE SEPERATOR) Subminiature Relay One of the world’s smallest relays, the G2E, is available with coil ratings from 1 .5 V to 24 V at currents of 18.8 mA to 300 mA, This subminiature component measures only 16 x 11 x 10,5 mm and weighs just 3.5 g. Features include long service life. in excess of 5 x 10 & operations, and a switching frequency of 1800/hr max. Coil resistances range from 5 ft to 1280 O and this highly sensitive relay requires only 290 mW of pull-in power. Both operate and release times are > 5 ms, and the G2E has an insulation resistance of <100 MH (measured with a DC 500 V megger). Designed specifically fotp.c.b. use, the relay has its pins set on the international 2.5 mm grid terminal arrangement. A transpaxant case protects the component from dust, while the operating ambient temperature range is -25 Q C to +55°C. Relay contacts are rated at 0.5 A (100 V ACFF- 1) or 1 A (24 V DC), while the carry current is 2 A for both versions. IMO Precision Controls Limited, 349 Bdgware Road , London W2 IBS (973 m 23 mm potentiometers with switch The new Iskra 23 mm carbon potentiometers are spindle- operated and incorporate a switch rated at 250V AC, 4 A continuous current, 80 A surge current. Known as the P8SL series, they are available from Iskra Limited of Coulsdon, Surrey, the British trading subsidiary of the multinational Iskra organisation. Each switch is approved to BS415 and SEMKO standards, and the series is produced in both linear and log values. The linear values extend from 100 ohms to 10 megohms, and the log values from 2.5 kilohms to 5 megohms. Maximum voltages which can be applied to the carbon track are; linear units, 500 V DC; log units, 350 V DC. All potentiometers in the P8 range can be supplied with any one of three tappings - 40%, 50% or 60%. Each unit has a 295-degree angle of rotation and a smooth rotary action. End-stop torque is 8 kgcm. Low-noise carbon-brush wipers are standard throughout the range. Rated dissipations are 0.4 watts (linear) and 0.2 watts (log). Mechanical fixing is by a standard M10 x 0.75 mm screw T tin cad. Two mounting styles are available, one standard and one special. The standard is designed for a 10.1 mm diameter centre hole, with a 2.1 mm (+0.1 mm, -0 mm) locating hole. On special request, Iskra can supply units for mounting in a 10 mm diameter hole with a rectangular section extending from the circumference. This section measures 2.5 mm x 2.6 mm (+0,1 mm, -0 mm). Dual-gauged versions for stereo- phonic applications are also available. Iskra Limited, Redlands, Coulsdon, Surrey, CR3 2HT (986 Ml 16.8 mm carbon potentiometers There are four types in this new 16.8 mm spindle-operated range, and together they cover many applications. Type P 10 units are available with linear resistance values from 100 ohms to 5 megohms, and log values from 2.5 kilohms to 5 megohms. Rated dissipations are 0.15 watts (linear) and 0.08 w r atts (log). Designed for mounting with a M7 x 0.75 mm screw , all P10 potentiometers incorporate a metal contact wiper. Angle of rotation is approximately 295 degrees, and ail units have a smooth rotary action. End-step torque is equal to or greater than 4 kgcm. All potentiometers in the P10 range are available with an integral switch rated at 12 V, 2 A. When a switch is incorporated, the type number is designated PIGS. For stereo hi-fi and similar applications, Iskra has produced a tandem version of the PI 0. Called the P10T, it has the same specification as other units m the P10 range. Matching can be achieved down to 26 dB, 16.8 mm series with plastic cases Many applications today demand a higher resistance to the ingress of dust and other harmful particles. To cater for this demand, Iskra Limited has introduced a series of plastic- cased potentiometers. Known as the PI IP series, these 16.8 mm spindle-operated carbon units are specifically designed for printed- circuit applications, but are also available with solder terminations. As well as featuring a plastic case, all Pi IP units have a 4 mm plastic spindle and a plastic threaded sleeve (M10 x 0.75 mm). Resist- ance values arc 1 00 ohm to 5 megohms linear, and 2,5 kilohms to 5 megohms log. The angle of rotation is approximately 270 degrees. Each PI IP potentiometer has a smooth rotary action, and an end-stop torque of 4 kgcm. Rated dissipations are 0.15 W linear, and O. 08 W log. Iskra Limited, Redlands, Coulsdon, Surrey, CR3 2HT (987 M) P, C.B. mounted illuminated push-button switches Highland announce the release of Series 99 illuminated push-button switches and indicators for printed -circuit-board mounting. Designed for applications on both industrial and domestic equip- ment, this new range has low-level switching elements rated 100 mA, 50 V max. Push buttons and indicator assemblies are square - screen size 1 8.6 mm square with both flat and concave illuminated screens. The switch and indicator terminations are by ] mm pins for direct p,c. board mounting. They may be flow soldered. Illumi- nation is provided by miniature T1 size bi-pin lamps available in 6, 12, 18, 24, 28 and 36 Volt versions. The range comprises Indicators (signal lamps) and both momentary and maintained- action switches with low-level switching elements having a 1 make/1 break (non-shorting) contact configuration. Film legends, designed and produced to customer’s requirements by Highland, are intended for use on Series 99 devices. These arc sandwiched between the clear outer screen and the inner white diffusor with a colour backing film if required. High la nd Elec tronks L td . , Highland House, 8 Old Steine, Brighton, East Sussex, BN I 1EJ (988 M) The plug which thinks for itself This ingenious three pin 13 amp plug by means of neons and indicator on the front of the plug shows immediately if; 1. Socket is Ai OK 2. DANGER reverse polarity 3. DANGER no earth 4. DANGER live fault 5. DANGER neutral fault This tester is universally needed by DIY buffs, craftsmen, tradesmen, householders, installers of microwave and solid state circuitry and throughout industry and commerce. It was initially designed to protect electronic machinery installed by C. & A. Modes in all their stores throughout the U.K. Gala trek Engineering, Scotland Street, Llanrwst, Gwynedd, LL26 OAL, North Wales, Great Britain, (975 market 12-56 — elektor decern bet 1978 Rectifiers for X-Ray power supplies Semtech have introduced the 4 X-Way Stic \ a new series of open rectifier sticks specifically designed for X-Ray power supplies* Each X-Way utilizes hermetically sealed ‘Metoxilite’ multi-chip ^avalanche" rectifiers mounted on a PC B* These ‘Metoxilite 1 multi-chip rectifiers (initially developed for high reliability aerospace programmes) are now available at commercial prices. In addition to X-Ray power supplies, these rectifiers can be efficiently used in most standard, single and polyphase circuits. Type: X 100KS, X 125KS, Y 1 C f\Y O P1V : 100, 125, and 150 KV. Average Rectified Current (at 55*0 oil): 220 mA* Size : (overall) 6.22’"L x .690"W x (end cap) *5 00 ”D (anode) *39fTL* Bourns (Trimpot) Li mi red, Hod ford House, 17/27 High Street , Hounslow , Middlesex. (926 M) Desk console Recently introduced by BOSS Industrial Mouldings Limited is the new r BIM 8005 Small Desk Console, the main feature of which is its three piece construction which allows rapid and easy removal of both top and bottom panels. The main body of the console is manufactured in ABS, as is also the removable bottom panel which incorporates 9 mounting bosses for supporting either a full 100 x 160 mm Eurocard or 2 half size boards* The flush mounted 1 mm thick grey aluminium top panel tones with the orange, black or blue console body, the latter — as with the majority of BOSS enclosures - also incorporating slots for vertically mounting 1.5 mm thick pcb’s as well as the recently announced BIMDAPTORS which allow pcb’s to be mounted sandw ich fashion* Measuring internally 165 x 122 x 60 mm high and incorporating small, self adhesive, rubber stand-off feet on the base, this new small desk B1MCONSOLE is equally suited to initial development, pre-production and OEM type applications. BOSS Industrial Mouldings Ltd f Higgs Industrial Estate, 2 Heme Hill Road, London, SE24 OA V, England (921 M) Voltage-to-frequency converters A new family of Voltage to Frequency (V-To-F) converters, designated LMI 31/231/331 is now available from National Semiconductor Corporation, with better specifications than any previous device. It is claimed that the new devices ail function with guaranteed 0,01% maximum non-linearity as 10 KHz V-toF or F-to-V converters, and can be operated from a single +4.0 to +40 volt power supply* The standard device is available with a gain temperature coefficient (T*C.) of 150 parts per million/ 0 C typical (150 ppm/ 0 maximum). These devices are particularly suited to high- resolution data acquisition applications. The premium-grade versions of the LMI 31 family (LMI 31 A/2 31 A/331 A) also feature a guaranteed gain T*C. of ± 50 ppm/ Q C maximum. The LM 1 3 1 A is the first voltage to frequency converter in the industry guaranteed to have a 50 ppm/ 0 C T.C. spec over the entire — 55°C to +125°C tem- perature range. Designed to function as an improved, pin for pin replacement for the 4151 V-to-F converter, the LMI 31 is specified to operate over its 4 to 40 volt supply range with improved accuracy. The LMI 31 also features low pow er dissipation (14 milliwatts) when operated from +5 volt supplies, and this is highly suitable for battery powered applications. National Semiconductor , 301 Harpur Centre , Horne Lane , Bedford MK40 1TR. (909 M) High-voltage video transistors Micro Electronics Ltd. has introduced a wide range of high-voltage n-p-n silicon planar transistors designed for video-amplifier applications in colour and monochrome television receivers* Available in TO-39 and TQ-92 cases, the devices feature collector-emitter breakdown voltages up to 300 V and have good frequency characteristics, with current-gain bandwidth products of 50 MHz* Devices in the TO-39 package include the BF 257/258/259, designed for high-voltage video output stages, and the BF 336/337/338, designed for R4T-B and coulour-difference output circuits. TG-92 packaged devices include the BF 297/298/ 299 and BF 391/392/393 transis- tors. The devices are designed for maximum emitter-base voltage of 5 - 8 V, and cover maximum collector currents from 1 00 mA to 500 mA. Total maximum power dissipation at a case temperature of 25°C ranges from 1*5 W to 5 W. Micro Electronics Limited, York House ; Empire Way, Wembley r Middlesex * (918 M) Low-noise audio transistors A new range of small-signal transistors from Micro Electronics Ltd* are designed for iow-noise preamplifier, audio-frequency amplifier driver stages and sjgnal- processing circuits in television receivers. Designated the BC 413/414/415/416 Series, the transistors are low -cost silicon planar epitaxial devices featuring high breakdown voltage, high gain and !ou r flicker noise* The BC 413 and BC 414 are n-p-n transistors, while the BC 415 and BC 416 are the complementary p-n-p devices* Maximum collector-base voltage is 45-50 V, maximum emitter-base voltage 5 V. Maximum continu- ous collector current is 100 mA, and total power dissipation is 300 mW. Current gains ranging from 125 up to 900 are available, typical noise figure is 1*2 dB, and flicker-noise voltage referred to base is only 0*11 mV for the BC415 and BC416 and 0*135 mV for the BC 41 3 and BC 415. Micro Electronics Limited, York House, Empire Way , Wembley, Middlesex. Temperature range -55° to 125°C LMI 31 H - LMI 31 AH —25° to +85°C LM231 H -LM231 AH 0° to +70°C LM331N and LM 331 B-LM331 AH (930 Ml market elektor decern ber 1978 — UK 15 fj.P appliance timer A single package MOS-LSI microcircuit appliance timer, consisting of a central processor with on-chip memory, has been introduced by General Instrument Microelectronics. The circuit - basically a 4-bit micro- processor - is available as a basic 28-lead or a more sophisticated 40-lead version, and will considerably widen the scope of domestic appliance designers. The device is essentially a versatile, low cost timer, providing designers with the type of facilities necessary for controlling cookers, driers, central heating, etc. The 28-lead version designated AY-3-1250, accepts instructions from ‘hours up 1 , ‘hours down’, ‘minutes up’, or ‘minutes down* keys* where momentary depression of keys cause single increments or decrements, and continuous depression causes the displayed digits to cycle. In use the circuit is linked to a 4-digit LED display indicating any function selected. It has 3 separate outputs for which on and off times may be programmed in. In a cooker application, all normal instructions may be programmed in, including ON time* OFF time, cook duration, etc. Other facilities include a ‘minute minder 1 with audible alarm output, indication of mains failure, 12/24 hour operation, temperature setting or an optional fluorescent display driver output interface. The 40-lead version - designated AY-3-1251 - is designed for more- sophisticated systems where 10x 4 keyboard or touchpad entry and 14-digit permanent display are required. It has four controlled outputs, each w ith a variable mark-space ratio for control of hotplate duty cycles, etc. The 14 -digit display facility could be used for a minute minder (3 digits), oven temperatures (3 digits), time on/off (4 digits), and hotplate temperature (4 digits). When used for fully automatic cooker control, the time programme would be entered and cooking temperature selected using a key pad. When the start time is reached the appropriate output would be activated and an ‘ON’ indicator lamp energised. When the stop time is reached the output is deactivated and the minute minder audible alarm activated for 10 seconds. All 3 programmable outputs may be separately controlled in this way, but the device can also be used in a semi-automatic or manual mode. A further facility allows a set programme to be repeated at 24 hour intervals by the simple depression of a ‘repeat 1 key. Both 28-lead and 40-lead versions include a built-in standby frequency source, which allows the devices to function normally during mains failure. In this event the circuit detects the abscence of 50/60 Hz input, and a 200 kHz oscillator takes over timing under external battery power and lights a ‘mains failure’ warning lamp. On the re-application of the mains supply the circuit returns to mains power, but leaves the lamp on (until it is manually reset) to indicate that mains failure has occurs d. Although the tw o versions described here are standard devices, the microcircuit chip is mask programmable to include a wide range of options and output requirements. These include steady or pulsed temperature display outputs, error ‘E’ display, programmable time delays, touchpad inputs. Both versions are extremely suitable for general control applications requiring simple arithmetic, display of up to 43 digits and interfacing with analogue systems. General Instrument Microelectro nics Ltd., Regency House, 1-4 Warwick Street, London WlR 5WB f England (982 M} Video monitoring oscilloscope Gould Instruments Division has won a substantial order from the British Broadcasting Corporation for the supply of an oscilloscope designed for video monitoring applications. The oscilloscope is a modified brighter version of the Gould Advance OS 3 300 B with a BBC designed timebase module incorporating comprehensive video triggering facilities, which is being made by Gould under a manufacturing licence agreement from the BBC, The new r timebase generator allows the oscilloscope to be used for detailed line-by-line examination of 625-line television waveforms or to display a television picture. It accepts a standard level video signal, which may contain ‘Sound -in-Sync 1 signals and provides six different triggering modes: field 1; field 2; field 1 and 2 alternating; line repetitive; single line selectable by front panel switches (with the line number indicated on a 3 digit light emitting diode display) and line pairs in the range 16/329 to 22/335. The triggering can be delayed continuously by up to 90 jus via a multiturn potentiometer, which allows the signal to be examined in detail. The displayed video signal may be clamped or not, as required. When the unit is used to display a television picture, the triggering point selected may be observed as a ‘bright up line 1 on the picture, enabling the waveforms to be rapidly related to the picture. The changeover from waveform to picture is effected by a single front panel switch. The modified timebase retains its normal triggering facilities, so that the instrument may be used as a general purpose single timebase oscilloscope, Gould Instruments Division, Roebuck Road, Hainault, Essex (974 M) 1 MHz to 1600 MHz in one sweep Gould Instruments Division has entered the sweep-generator market with the introduction of the Gould Advance SW100, a high-performance instrument which can cover the frequency range from 1 MHz to 1600 MHz in a single, flat sweep. Other key features of the instrument include a comprehensive automatic marker system for precise frequency determination and an output level unit that can be varied in 0,01 dB steps. The Gould Advance SWIOO incorporates electronic switching between the three basic ranges (1-500 MHz, 450-1000 MHz and 950-1600 MHz) to give the effect of a single sweep. Deviation from flatness is t 0.25 dB using a power meter, or ± 0,35 dB measured on a full-wave detector. The three operating modes of the SW100 are: start/stop or F1/F2 mode* in w hich a sweep occurs from one set frequency to another; FC/ f (delta F) mode* in which a centre frequency is set in the middle of a spectrum of interest with one control and a deviation about the centre frequency is set with another control; and signal generator mode, in which the sweep drive is sw itched off and the sweeper is used as a straight signal generator. An extremely accurate output attenuation system offers ten steps of 10 dB, 1 dB, 0.1 dB or 0,01 dB, and the output level range can be varied from +6,66 dBm (471 mV or 4.6 mW) to -103.33 dBm (3 mV or 0.18 pW) in 0,01 dB steps. The SW100 incorporates a unique Mgh-accuiacy automatic harmonic marker system with frequencies of progressively increasing resolution being cut in as they are needed. Markers are provided wdth 100 MHz, 1 MHz and 100 kHz separations, and the markers can be oriented in horizontal, vertical or 45 y directions to prevent them being obscured by the waveform under examination. A unidirectional marker facility is also provided for cases where the instrument is used wdth low-speed output devices such as X-Y plotters. Internal and external amplitude and frequency modulation (both fully swept) are available, and comprehensive internal and external sweep-rate controls are incorporated. The sweep may be varied between 70 Hz (14 ms sweep) and 120 seconds per sweep, A manual sweep control is also incorporated. Harmonic-related spurious outputs are typically -35 dB, and nonharmonic spurious outputs are less than -40 dB, Linearity is better than ± 2%. Construction of the SW100 is based on a modular approach, w ith easy accessibility for servicing and maintenance, leading to low ? cost of ownership. Three blank panels are provided on the front panel to accommo- date future options; a hard-copy output option is already available, The Gould Advance SW100 measures 320 x 145 x 365 mm, and weighs 8.6 kg, Gould Instruments Division, Roebuck Road, Hainault T Essex (980 M) ELEKTOR BACK ISSUES is this the first issue of Elektor you have seen? if it is, you wifi be pleased to know that you can obtain back issues direct from Elektor. Unfortunately our rapid growth is creating heavy demand and we regret some issues are now completely sold out. You may take some consola- tion however in knowing that we have 1 20,000* plus readers every month. We suggest you follow their example and order your back issues now using the reply paid order card in this issue, and piace a regular order with your newsagent or Component shop for future issues. 1976 issue no's W f 1$ 8t 7 9. .. . issue no. 15/76 . 1977 issue no's 21-26 f 29 & 32 issue no, 27/28 1978 issue no 's 33-38, 41 — 44 issue no. 39/40 ....... . £ 0 1 55 8 f. 50 each. .£0.95 S3- 00 each. . £0.60 $1.50 each. . £1.05 S3. 00 each. £0.65 £ 1.30 $ 1, 50 each. S3. 00 each. Based qti 2.62 readers per spld copy. Elector is a member nf the Audit Bureau of circufations. Elektor December 1978 — 16 advertisement More and more people are reading Elektor for your copies of Elektor 3* .o' a v0 & «■ ✓ It is evident that in your profession and/or hobby the design ideas published in Elektor are referred to time and time again. We are therefore now introducing this new cassette style binder to keep your copies of Elektor clean and in order. The chamfered corner of the cassette allows instant recognition of each months issue without the need to thumb through pages of previous months issues. No wires or fastenings are used so copies are easily removed and replaced and each cassette will hold one year's volume of Elektor. Their smart appearance will look good on any laboratory shelf. apfe*?* /;* / . ; Ml ii eiliiili Pm «Hii W:- ■ nr i - i r .-w:-:-. j&V ‘ m t 7400 IOp 7401 IOp 7402 IOp 7403 IOp 7404 12p 7405 12p 7406 25p 7407 2bp 7408 12p 7409 12p 7410 12p 7411 15p 7412 15p 7413 25p 7414 4&p 7416 25p 7417 25p 7420 I2p 7421 20p 7422 15p 74?3 20p 7425 20p 7426 22p 7427 22p 7428 25p 7430 12p 7432 20p 7433 28p 7437 20p 7438 20p 7440 12p 7441 45p 7442 40p 7443 60p 7444 6 Op 7445 bbp 7446 50p 7447 50p 7448 50p 7450 1 2p 7451 12p 7453 12p 7454 12p 7460 12p 7470 25p 7472 20p 7473 25p 7474 2bp 7475 25p 7476 25p 7480 40p 7481 86p 7482 75p 7483 75p 7484 70p 7485 60p 7486 25p 7489 130p 7490 25p 7491 40p 7492 35p 7493 30p 7494 70p 7495 46p 7496 45p 7497 120p 74100 80p 74104 40p 74105 40p 74107 25p 74108 lOOp 74166 75p 74109 74118 74120 74121 74122 74123 74125 74126 74128 74130 74131 74132 74135 74136 74137 74138 74141 74142 74143 74144 74145 74147 74148 74150 74151 74153 74154 74155 74156 74157 74160 74161 74162 74163 74164 74165 74166 74167 74170 74173 74174 74175 74176 74177 74178 74179 74180 74181 74182 74184 74185 74188 74190 74191 74192 74193 74194 74195 74196 74197 74198 74199 74293 744500 745112 40p 35p 35p 60p 120p 90p 45p 90p 80p 90p lOOp 50p 180p 270p 270p 55p lOOp 90p 65p 45p 45p 70p 45p 45p 45p 55p 55p 55p 55p 6 Op 80p 75p 160p lOOp 80p 60p 60p 50p 50p 75p 120p 90p 130p 50p 120p lOOp 3 2 Op 70p 70® 60p 60p 55p 50p 50p 50p lOOp lOOp 90p 18p 80p 7805 78'2 . 7815 7818 7824 140p LINEAR AY 38500 CA3039 CA3046 CA3060 CA3065 CA3076 CA 3080 CA3084 CA3085 CA3086 CA3088 C A 3089 70p 60p 22&p 20 Op 25 Op 75p 250p asp 60p 190p l&Op lOOp 60p aop 80p 2S0p CA3090AG 36 Op CA3123E 130p CA3130 CA3140 LF356 LF357 LM211H LM3002R5 170® LM301AN 30p LM304 200p LM307N LM317K LM324 LM339 LM348N LM380 LM381N LM382 LM391 LM555 LM709C LM710T05 LM710DIL LM723T05 LM723DIL LM 733 LM741 LM748 LM1303N LM 1 458 LM3080 LM3900 LM3909N MC1310P MC1312P MC1314P MC1315P MK50398 MM5314 MM5316 NE529K NE555 NE556 NES62B SAD 1024 SL917B SN760O3N SN76013N SN76013ND 125p SN76023N H0p SN76023ND 12Sp SN76033N 150p SN76227N SN76228N SN7666DN TAA300 TAA350 TAA550 TAA570 TAA6610 T A A 700 TAA790 TAD100 TAD110 90p 400p 150p 650p 150p 110p 160p 3S0p 350p 150p 1»p 6 Bp 40p 4 Op 120p 20p 40p lOOp lOOp 76p 66p 6Sp 1 40p ISOp 190® 230p 150p 25p XR4739 ZN414 95M90 150p lOOp 700p CMOS 4000 12p 4040 60p 4001 12p 4043 60® 4002 12p 4046 90® 4006 SOp 4047 80p 4007 14p 4048 50® 4009 SOp 4049 25® 4011 12p 4050 25® 4012 12p 4054 100® 4013 30p 4055 130® 4015 50p 4056 120p 4016 30p 4060 lOOp 4017 50p 4066 35p 4018 55p 4069 12p 4019 40p 4070 12® 4020 50® 4071 12p 4022 50® 4072 12® 4023 12p 4081 12® 4024 40® 4062 12® 4025 12® 4093 70® 4026 80p 4516 60p 4027 30p 4510 70® 4028 45® 4516 65® 4029 50® 4518 65® 4030 30p 4520 65® 4032 80p 4528 80® 4033 lOOp 4583 70® POWER SUPPLY CAPS 2200 » 6 35p 2700 40 50p 4 700 63 ’20a 2200 63 §0p 4 700 70 1 35p 7700 '00 ISOp 10000 '0 'OOp 3300 30 50p 10000 25 ISOp 3300 63 »p 15000 '5 ISOp 4700 40 66p 27000 75 ?00p 4700 25 50p ENQUIRES FOR ANv OTHER fvPES ELEC CAPACITORS 1 25 1 50 2 2 25 2 2 35 3 3 25 4 7 10 4 7 16 4 7 25 4 7 50 68 25 10 10 10 <6 10 25 10 50 22 6v3 22 10 72 16 2? 25 22 35 22 50 33 6V3 33 16 33 25 33 40 33 50 7p 7p 7p 7p 7p 7p ft 7p Tp 7p Tp Tp Tp 7p Tp Tp 7p Tp Tp »® *> Tp >P ®P Bp «P 4 7 10 47 16 47 25 47 35 47 50 100 10 100 16 100 25 100 50 100 63 220 16 220 25 770 50 330 25 330 35 330 50 4 70 10 470 25 4 70 35 470 50 1000 16 1000 25 1000 35 1000 40 1000 63 1 200 63 7200 10 22p I7p IBp 20p 14p IBp 20p 24p 27p **> 50p 60p AA 113 AA71 7 AC121 AC126 AL 1 77 A (.17/ 01 AC 128 AC15I AC 153 AC 153* AC 1 54 AC 1 8/ AC 1 88 acvi; A< YI9 ACV70 AC T 7? ACY40 AC V 4 1 ACV47 ADI 30 ADI43 AD 1 49 ADI6I AD 167 ADI61 7MP AC 1 14 AMIS AC 116 AMU AC I 18 AC 1 78 AC 1 3M AC 719 AC 779 Aul 10 HAI 14 MAI?1 HAI54 HA157 HAITI HAJM l BA* Ip HAW7I 88105 hAi in BC 10/ 8C108 8C108C BC 109 BC.109C BCH3 8CH4 8C 1 1 5 BO 16 BC 117 BCU8 BC 1 19 00 25 80 258 80 26 80 34 80 36 8037 80 38 8040 8041 BC142 8043 8047 8048 BC148C 80 49 8053 80 54 8057 8058 8059 BC16/A BC 16 B BO 69 8071 IOp 30p 30p TOP 20p 25p 70p 25p 30p- 40p 30p 70p 20p 35p 35p 35® 40p 50p 50p 50p I50p ISOp 80p 30p 30p 70o 75p 25p 25p 25p 35® 45® 50® 180® 17p 12p 17p 5p TOp 35® 35o IOp 10® 15® IOp 15® I2p TOp TOp 15p 16® 16® 30® 30p 30® 30p 30p 10® 10 ® 1*® 10® 16® 15® 12® BO 72 BC 1 7 7 BCI78 BOB? B087I 8083 BO 831 BO 84 BC 1841 HO 86 HC704 EtC 705 BL707 Bl.21? BC7171 fM.713 Bi ?13l BC 714 BC7141 HC7J7 BC 737H HI 768 HC 794 AC 300 BC301 BC.303 H<. 307 He XW •C3II H< . 3 1 H BC 373 BC328 BC337 He 338 HC348 Hi 461 BC516 BC;51 7 He 54/ Be 54 7H BL548 BC 54BC Be 54 9H B( 549c HC557 BC55/6 BCV34 BCY38 8CY47 BCV43 BCV58 BCY59 BCV70 8CYT1 BC v 7? BD115 80171 B0131 80 1 37 80 133 BOI35 80136 00137 8DI39 BDI40 8DI44 B0181 B018? 80188 8070 7 80220 B0722 80733 BD738 80242 BD75? 80263 80607 BO 6 O 6 8D609 BD6'0 80679 B0680 MURATA ULTRASONIC TRANSDUCERS 40k Hz Typp MA40 LIS Transmit Type MA40 LIR Receive £2.00 each £3.50 pa.r Data IOp ROTARY SWITCHES BY LORLIN 1 POLE 1 ? WAY ? POLE 6 WAY 3 POLE 4 WAY 4 POLE 3 WAY All at 40p Eech OPTO ELECTRONIC CORNER SPECIAL SCOOP OFFER 0 125 or 0? inch RED LEDS ISp ••eh. 10 lor £1.00, 100 for £7.50. 1000 for £80.00 0.125 or 0.2” Y*llow and Green LEDS 15®. 10 for £1.40. 100 for £12.00. HP 5082 - 7750 16 Op each 0L747 Seven Segment Common Anode Displays. Character Height 0.6” £2.00 each FN0500 Seven Segment Common Cathode Displays. Character Height 0.5" £1.30 each 4 for £5.00. 2N5777 Photo Cerl.ngton 60p each. 0RP1 2 Japanese 75® each Muderd £1.25 eech. DECODER BOARD CONTAINING 18 * 74156 2 * 74155 2 * 7409 1 * 74180 1 * 74150 1 * TIP32 2 « 60 WAY EDGE CONNECTORS Few only left of this unreap unrepeatable bargain £3 50 each • • • • • XTAl MIC Inserts 75p eech 5” Scopetubes SE5J (for callers onlyl £15 00 each 8>as R elector c®i*s 50 100KHZ 65p “ c 4 • • • • 1 MHz CRYSTALS £3.00 Push to Make Switches 2 Op each Chokes lOuH 35® each IOOuH 65p each Futaba 5LT02 Non Muitipie*ed 4 Dig-l Phosphor Oiode Display With A M. P M Colon £5.00 66® T8A1205 60p 100® T8A480Q 2 OOp lOOp ISOp TBA520Q TBAS30Q 200p 200® POLY CAPS lbOp TBA540 200p 32Sp TBA550Q 250p '000 Pf ip 0 1 uF 6c 70p TBAS60C 250® 2200 So 0 2? uF 7® 60p TBA641A12 250p J300 4® 0 33 uF *> 90p TBA700 IBOp 4 700 0 47 uF 12p 60p T8A720Q 22Sp 6800 ip 1 0 uF 20p 90p TBA750O 200p 0 01 uF in 77 vf ?6p 90® TBA800 SOp 0 02? uF ip 4 7 uF »P IBOp TBA810 'OOp 0 033 uF 5p 6 8 uF 40p *>P TBA820 'OOp 0047 uF 5p SOP 40® TBA920Q 280® 60p TCA270Q 22 Op ★ ★ 1N4148 DIOOES BY ITT TEXAS 100 for £150 Please note, these are full spec devices Te.as TIS88A V H F F E T 10 lor £2 30 100 for £20 00 555 TIMER 10 for £2 50 741 O P Amp 1 10 for £2 00 TCA270S 220p P TCA760 JOOp TC A 4500 A TOA 1008 460p 360p TANT. BEADS TO A 1034 460p 3.3/16V TDA2002 300® 0 1 '35V 14. 14a TDA2020 300p 0 15:35 V 14a i4a TL084 1 20p 0 22/35 V 14a i4a XR320 2SOp 0 33/35 V 14® 14a XR2206 4bOp 0 47/10V 14® 68. 6 V 3 14c XR2207 450p 0.47:36V 14a 68'35V 14a XR2208 600p 0 68:35V 14a 1 0/35V 14a XR22T6 650p 1 00: 10V i4a 22/ 15V 21c XR2567 250p 1 00 /35V i4a 33/16 25P XR4136 150p 1 5:36V 14c 47/3V 20p XR4202 150p 2 ?:76v 14a 47/16V 26p XR4212 150p 7 7/35V 14a I00/3V 26a SPECIAL OFFERS ASSORTED GIANT SCREW PACKS INCLUDES 100 off 741 1 500® SELF TAPPING. SELF CUTTING 100 Off 555 s 1900® NUTS. BOLTS. WASHEHS. 100 off IN4148 150® EYELETS ETC. ETC. 100 off AD161 2500® WEIGHT 1 kg 2.2ib* 100 off AD162 2500® Appro* 1400 item* 100 off .185 Rad Lads 750p ONLY £1.80 (U K. only) 100 off .2 Rad Lads 750® 100 off Graen/ Yellow Led* 1200® .. MULLARD MODULES LP1152 LP1153 LP1165 LP1166 LP1168 LP1169 LP1173 LP 1181 LP1400 EP9000 EP9001 EP9002 ELC1043/05 lOOp 400p 400p 400p 400p 400p 400p 280p 280p 280p 450p ^POTENTIOMETERS IK LIN 5K LOG 5K LIN 1CK LOG 10K l IN 25* lOG 25* LIN 50* LOG 50* LIN '00* LOG 100* LIN 250* LOG ★ ★ LIMITED OFFER 8C237 100 for £5.00 >* 500* LIN IM LOG 1MLIN ?M LOG 2M LIN AH at 30a Each ELEKTORNADO AMPLIFIER KIT. All parts £1100 P.C. Board £2.96 GANGED POTS AM at 80a Each 5* • 5* UN o» LOG 10* • 10* LIN or LOG 25* • 25* LIN o* LOG 50* • 50* LIN or LOG 100* • 100* L IN or LOG 250* • 250* LiN or LOG 500* • 500* UN or LOG 1M • IM LIN or LOG ( 2M ♦ 2M LIN or LOG A Power Supply part* P.O.A. OIL SOCKETS SPIN 13a 14 PIN 14a 16 PIN Ife 28 pin 45p X OH* COUNTER Tima Data and control board part* . £22.44 Counter and Otlpley board parts £31.76 P C. Board e*tra £8.96 L.F. Input board part* . £4.01 P.C. Board a*tr# H.F. Input board . £13.25 P.C. Board extra Price of Total k.t las* mechan-cel part* £01.86 A.. P.C. 8oerdi extra . . £19.50 Total Price* £101.36 PRE SET POTS 100m* Hon/ Vertical 50R 1M Ohm 8® Each MULL AHO POT CORES LA3 100500* HZ 75® LA4 I030KMZ 100® LA5 30100KHZ LA13 200® 82 YM ZENER DIODES 400m* OVT 33V IOp EacF BRIOGE RECTIFIERS I AMP 50V 1 AMP 100V 1 AMP 200V 1 AMP 400V 1 AMP 600V 1 AMP 1000V 2 AMP 50y 2 AMP 100V 2 AMP 200V R C. A. TRI ACS 400V 8 AMP £1 20 400V 15 AMP £1 60 MICRO BLOCK 2102 250 Nano Sec Static RAM 11024 * 1 BITI £2 20 each 4 for £8 40 8 for £16 00 ft 2102 450 Nano Sec Static RAM (1024 * 1 BIT) £1.00 each a 2112 450 Nano Sec Static RAM (256 * 4 BIT) £2.50 each ft 251 3 Character Generator Upper Case £7 00 e®ch ft 2513 Character Generator lower case £7.00 aech ft MM5204 E Rom £8.00 each ft 8212 8 Bit in/out Port £3.00 each 1 8060 an MPU(CPU) £1 2 .00 eech . 8831 Tri State Lme^ Driver £2 00 eech 8833 Tn State Trans Transceiver £2.00 eech 8835 Tri State T ranseiver £2. 00 eech AYS- 1013 U/ART £8.00 CONSONANT KIT PRE CONSONANT KIT LUMINANT KIT Prices on application High quality Trimmer Caps Min-Max 2 5pF-6pF All one 3 5pF-13pF pr.ee 20p 7 OpF 35pF FERRITE BEADS 6MM long OD 3MM ID 1MM 3p eech 100 for £2.00 Paper 0.5uF 400V AC Caps ideal for Car ignition Systems etc SOp eech Assorted Japanese I.F. Transformers 20 for £1.25 FERRITE RINGS FX1593 O.D. 12 m.m. I.D- 6 m.m 10 for 70® T. POWELL 306 ST PAULS ROAD HIGHBURY CORNER LONDON N1 01-226 1489 shotmoors BARCLAYCARO r ^ i I a'ds'*” . hA>qe MON - FRI 9 - 6 30 PM SAT 9 - 4.30 PM Christmas holidays dosed:— Dec. 23-Jan. 2nd. PRICES valid it the time of going to press ALL PRICES INCLUDE POST AND VAT