4 special more than lOO practical projects multimeter generators and oscillators auto duty cycle baud rate generator pulse generator sawtooth generator Wien bridge oscillator message from your editor readership survey market appointments switchboard advertisers index missing link H.F. f M pocket radio time-signal leceiver VHFiAM air-band convener VHF converter audio, video, and sound generation amplification selector stereo doorbell audio preamp buffer digital band-pass filter guitar preamplifier infra-red headphones receiver infra-red headphones, transmitter scratch and rumble filler screen noise killer 70/90 watt amplifier small high -power amplifier stereo balance indicator stereo noise suppressor switch-on delay sync separator touch-pad potentiometer voltage-controlled audio switch hobby and car alarm clock for cars automatic reserve warning light fatigue tester guitar preamplifier ’lights on" warning pace counter remote shutter release revolution counter self switching battery charger speed regulator for disco lights stroboscope measuring and test equipment amplification selector audible ohmmeter combining 4017 counters digiLED frequency meter hi-k) pulse rate discnmmator IX meter LEO current sensoi level indicator parser three state TTL logic probe transistor polarity tester VHF dipper ‘window’ LEDs computers and microprocessors cpu dock generator elekterminai bell EPROM eraser fast analog to digital converter floppy expander loystick interface lump on reset lightpen pP infra red interface mini signal cleaner multi-channel analog to digital converter parallei/senai converter power switches for pPs 6502 bootstrap RS 232 analyser three -state indicator 2716 versus 2708 twin RS 232 2 - 2716 2732 power supplies dissipation limiter high power opamp supply linear opto-coupier low power switching regulator microcomputer power supply microcomputer power supply protection Ni(.ad charger overvoltage protection powei supply considerations power supply for computers power supply monitor transformerless mams power supply automatic clockroom light automatic reserve warning light blown-fuse indicator central-heating monitor coffee temperature indicator electionic mousetrap energy saving porch light flashing telephone light fridge alarm kilowatt dimmer musical door- bell rain indicator smgle-buiton code lock super simple bell extension miscellaneous bird imitator blown-fuse indicator economical motor-driving circuit electronic key set event counter funny bird one-armed bandit photoelectronic relay switch indicator switching delay valve simulator versatile timer voltage controlled audio switch elefctorindia Aug/Sept 1984 1 8.03 New oaae has been added to x Measurement of INDUCTANCE. CAPACITANCE. RESISTANCE are greatly simplified by X/LCR 7. No balancing, no adjustments. Connect the component to the terminals. VLCR 7 gives you directly the digital reading of value and its loss factor simultaneously. FOUR TERMINAL measurement eliminates inherent errors due to lead resistance. GUARD TERMINAL provided eliminates errors due to lead capacitance. V VLCR 7 is the only instrument in India covering the widest ranges of 0.1 pf/uH/m ohm lie. 0.0001 ohm) to 20.000 uf/200H/20 M ohm. B. W/. if you do not need. Component tester is all the while required. J f DIGITAL FREQUENCY COUNTER VDC 18 Smallest size ever made in INDIA Battery cum maina operated wide frequency range 30MHz (our model VDC 19 is 500MHz) Basic sensitivity of 1 0mv Bright LED display 7 digit yv Robot in BEL plant Bharat Electronics Limited, Banga- lore, does not stop with the prod uction of black and white TV picture tubes; it can now offer the state-of- the-art technology itself for picture tube production plant of any capa- city, on a turn-key basis. This ability of the BEL was evident when its second picture tube plant was inaugurated by the Union deputy minister for electronics, Dr M. S. Sanjeevi Rao, recently. Except for two test equipment, the conception, design and erection of the entire semi-automated plant was indigenous and an additional fea- ture of the plant was the intro- duction of a robot in the production line, according to Mr. N. L. Krishnan chairman and MD of the BEL. Controlled by a microprocessor, the robot collects screen coated TV bulbs from two continuously rotat- ing, screen coating equipment and unloads them on an indexing type, twelve-head, drying machine. It controls release of coated bulbs, rotates the lower arms to pre- determined positions, centres the neck of the TV bulb, lifts it, rotates 180 degrees and places the bulb on the drying equipment. With the commissioning of the new plant, BEL’s total installed capacity will be 300,000 TV tubes per annum. The semi-automated plant’s capa- city is 100,000 tubes per annum. The pneumatically-operated robot, while maintaining quality and imp- roving productivity, eliminates ope- rator fatigue and health hazards, points out Mr. K. R. Savoor, general manager (components) of the BEL. The BEL has also announced a reduction in the prices of TV picture tubes, thanks to the improved productivity. The wholesale, ex- factory price for 20" black and white picture tube has been reduced from Rs. 390 to Rs. 370. The wholesale price at various outstation depots will be reduced from Rs. 420 to Rs. 385. The max imum benefit, how- ever, goes to the buyer from the BEL sales depots where the retail price has been slashed to Rs. 430 from Rs. 520 Similarly, prices of other sizes of tubes have also been reduced. ELCOT Ventures Electronics Corporation of Tamil Nadu Ltd. (ELCOT), which has set up an aluminium electrolytic capa- citor plant and digital electronic watch unit with Japanese techno- logy. is implementing a two-way radio communication equipment project, in collaboration with Mar- coni, U.K. ELCOT intends to have joint venture projects with any enterpreneur. A joint venture for metal and carbon film resistors has already been taken up and two more ventures, for mini- computers and terminals and video recorders/players are under consi- deration. ELCOT also holds letters of intent for telephone instruments, inclu- ROBOT IN BEL PLANT ding push button and memory bank type instruments, medium and high power X-ray system with accesso- ries, low power X-ray system with accessories, mini-computers, termi- nals and printed circuit boards. The enterprenuers can also contact ELCOT with their own schemes Japan looks to India One of the largest firms in Japan. Software Consultant Corporation, has given an assignment to the Bangalore-based computer manu- facturers, Processor Systems (India) Pvt. Ltd., to design and develop a system for computer- based education. PSI claims this is the first time that an Indian firm has bagged an assignment of this nature from Japan. Apart from two host computers and 80 terminals, the products to be developed by the PSI include, a full set of hardware documentation, basic software for the system and a full set of related software docu- mentation, both in English and Japanese versions. The project is expected to be implemented in five phases from April, 1984 to June, 1 985 and the PSI earns a fee of about U S dollars 300.000 for the project. Software exports Computer software exports from India to USA, USSR, West Europe. South East Asia. Australia. New Zealand and the Gulf countries may touch Rs. 100 crores per annum in the next three years, forecasts the Union commerce ministry. This optimism is based on the fact that software exports registered a 48 percent increase in 1983-84 It rose toRs. 20 crores from Rs. 13.5crores Among the steps taken by the ministry for augmentation of soft- ware exports are: liberalised import of computer software, hardware and peripherals, including central pro- cessing unit, memory augmenta- tion, crossing assemblies, VD asse- mblers, D-compilers, compiler programme generator, flow chart generators, debugging and diagno- stic tools, performance evaluation and monitoring tools, testing equip- ment, consumables and supplies related to export capability. Turning to electronics in export from electronics export, the govern- ment will set up a computerised engineering data system to provide quick and up-to-date information on India’s engineering capabilities to importers abroad and potential foreign markets. 8. 1 4 elektc dta 'Vjg/Sepi 1984 Circuits Special Arguably, this is one of Elektor's most popular traditions: more than one hundred circuits packed into one issue. Some people say that it is the only issue of Elektor they buy, while others claim that it is the only one they don't buy! Be that as it may . . . Some people have expressed their doubt about our claim that all circuits are lab-tested: 'Why, sometimes the components don't even exist!' Hmm . . . This year we have included a photo of the prototype in each article. Convinced? And as to the components, well ... we can't guarantee that they are all readily available now. They exist, they are in production - but they may not be in your local shop yet. For that matter, electronic components are a problem world wide at present. Not just exotic types: even standard TTL and CMOS. Some computer manufacturers are in extreme difficulties because they can't get certain parts. This immediately affects the retail trade. Some items are impossible to obtain; others are available only in limited quantities and at incredibly inflated prices. (We heard of one case recently where a retailer was offered a batch of 1000 ICs at a price that was nearly three times his own advertised sales price!) However, let's look at the bright side. If electronic components are in such demand, electronics is still booming. So we can expect lots more interesting developments in the future: new technology, new components - and cheaper, too, once the manufacturers catch up with the demand! Meanwhile, this issue reflects the current state-of-the-art. Some projects are just good, traditional designs; some use components that have' only recently become available; finally, there are a few that may only become true 'home -construction' projects a few months from now. All-in-all, it should keep a lot of people busy for some time to come! Finally, a word of warning to new readers. Another of our traditions is to include one joker in the pack. In other words, there is one project that is rather less 'practical' than the others. If you try to build that one, we wish you luck! (or disco lights It is interesting to be able to vary the run-off speed of the disco lights featured in our March 1984 issue. Not many changes to the circuit then published are needed, and even those are small. The additional circuit illustrated is connected to that of the disco lights via A ... F shown in figures 1 and 2 (the latter is the com- ponent layout of the printed circuit for the disco lights). Preset PI in the disco lights circuit should be set to maximum resistance because it is connected in parallel with any of the four presets in the present circuit. When the disco lights are being pro- grammed, the four new channels should, of course, be included. When a logic 1 is programmed, the buffer outputs (N42 . . . N45) are low and they must therefore be in- verted in N1 . . . N4 to be able to drive the CMOS switches. It is possible to apply a logic level to more than one of the inputs A . . . D at the same time. Because of the equal values of presets PI ... P4, a four-bit digital-to-analog converter is then formed which allows up to six- teen fixed speeds. At 0000 only PI in the disco lights circuit determines the speed. M keyboard adapter with P aralle ' output when the com- puter has one or more serial, but no Parallel/serial converters have a var- more parallel, interfaces available, iety of applications in computer The circuit is based on four LSTTL technology, for instance, in interface ICs: a four-bit synchronous binary circuits for printers with a serial in- counter, IC1, a parallel-load eight-bit put, or as adapter for a keyboard shift register, IC2, a quad two-input 8.16 elektor India Aug/Sept 1 984 NAND gate, IC3, and a quad two- input NOR gate, IC4. As the circuit contains neither time-dependent nor discrete elements, you’ll have to take care of its control by the system yourself. In the quiescent state, the binary counter is set to 1010 (decimal 10; This condition is decoded by NAND 1 gates N1 . . . N3 and NOR gate N7 into a stop signal for the counter (the output of N3 = ENABLE in- put — pin 10 - of IC1 = logic 0). The output of N3 is logic 1 for all other input combinations to N1 and N7. The shift register is inhibited by the logic high level on output Qq (pinJJI of IC1 . A logic low input at input LD (XMT) actuates the conversion pro- cess. The counter is then switched to binary input 1110 (decimal 14) and at the same time the register shifts the data to output Qh (pin 9). After LD has become high again, the leading edge of the next clock pulse switches the counter to 1111 (decimal 15). The CRY (carry) output (pin 15) of the counter then goes high which causes the serial output, SO, to become logic 0 via N5 and N6. At the following clock pulse the counter proceeds to 0000 and this condition is retained during the next eight clock pulses, that is, until the counter is switched to 0111 (decimal 7); output Qp is logic 0 during this time. In this period IC2 releases the parallel-loaded data serially, that is, one bit per clock pulse. At the ninth clock pulse, the counter proceeds to 1000 (decimal 8) and output Qq becomes logic high again. The two following clock pulses cause N5 and N6 to pass the two stop bits (logic 1). The next counter position is 1010 (decimal 10) the level of XMTRDY isjogic 1 dur- ing the data transfer, LD pulses dur- ing that time have no effect whatever. The circuit works equally well with 8-bit or 7-bit plus parity bit informa- tion. If only 7-bit information is to be used, input D7 should be made per- manently logic high: a third stop bit should not affect most systems one way or another. The current consumption of the con- verter amounts to about 70 mA. M . . . counts jumps too! For a change, here is a circuit which is primarily intended for sports people: it can count steps or jumps. From now on, whenever you go through a skipping session in train- ing, this circuit can tell you precisely and at any moment how many jumps you have made. All you need is a cheap LCD (liquid crystal display) pocket calculator, a small piezo buzzer, a type 4066 CMOS 1C, and a few other com- ponents. First, the buzzer has to be prepared as it will serve as the measuring detector. Carefully cut away a strip of the plastic housing and glue a small piece of relatively heavy metal (lead or iron) onto the brass mem- brane (see figure 2). Because of the increased inertia of the modified membrane it bends at every step or jump. The consequent piezo voltage generated by the buzzer is applied to the input of the circuit. The piezo signal is amplified by dar- lington pair T1 and T2, the gain of which is preset by PI. When the signal arrives at T3, it is converted into rectangular pulses which are used to control electronic switches ESI and ES2. These switches form a monostable multivibrator whose delay is preset with P2. Any pulses arriving during the delay period have no effect whatever so that, for in- stance, noise pulses are effectively suppressed. Now comes the question, of course, how to get to the memory of the calculator. To that end, one of the keys of the calculator is connected in parallel with a third electronic switch, ES3. Which key of the calculator is taken depends on the calculator. With many calculators it suffices to input a constant by which the . counter position is increased when the + or the M+ key is pressed. With yet other calculators, first the 1 and then the + , or M + . keys are pressed. If you buy a pocket calculator specially for this purpose, make sure that it is possible to increase the memory, and thus the LCD, by means of one key, by 1 (that is, the constant). for filming and other applications It is sometimes required that one of two parallel-operating units is switched on just after, and switched off just before, the other. An example is in film or camera work where the lights must be switched on just before the camera, and switched off just after it. The present circuit provides such a delay. Because the 1C we used contains four NAND schmitt triggers, of which only two are needed for the delay circuit, we took the oppor- tunity of providing a debouncing latch. In the circuit diagram, N1/N2 form the debouncing latch, and N3/N4 the delay circuit. Suppose that switch SI is in the off position. The output of N2 is then low, capacitors Cl and C2 are discharged, and the outputs of N3 and N4 are high. The base potential of p-n-p transistors T1 and T2 is then almost equal to the emitter voltage so that the transistors are cut off and the relays, Rel and Re2, are at rest. When SI is turned on, the output of N2 becomes high, and C2 is charged instantaneously via D3. The output of N4 goes low and consequently the base of T2 becomes more negative than the emitter. This tran- sistor then conducts and Re2 is actuated. At the same time, capacitor Cl charges also, but more slowly, via D2 and R2; the output of N3 does not go low until the voltage across Cl has reached the threshold value of the gate. When the level on pin 4 of N3 is low, T1 conducts, and Ret $ actuated. When SI is switched off again. Cl discharges instantaneoulsy over Dl. so that Rel returns to rest at once. Capacitor C2 on the other hand discharges more slowly over R3 and D4 so that there is a noticeable delay before Re2 is deenergized. The delay at switch-on depends on the time constant R2/C1, and that at switch-off on R3/C2. With the values shown, both are 2 ... 3 seconds. The circuit requires an operating voltage of 6 ... 15 V; the current depends on the relays used. The maximum current through a BC 557 should not exceed 100 mA, and the relays should therefore be chosen with that in mind. It is, of course, possible to use transistors which allow a larger current. from an idea by B. Willaert M B. 1 8 etektor in i Aug/Sepi protects tomorrow's dinner provides further optimization of the volume by tuning N3 to the resonant frequency of the buzzer. Preset PI determines the sensitivity of the alarm: the smaller its value, the less sensitive the circuit is. The alarm is most conveniently con- structed on the printed-circuit board shown in figure 2. Current consumption in the quiesc- ent condition is of the order of 0.5 mA and when the alarm operates about 4 mA. H from an idea by W. Groot Nueland As we all know, it is important that doors of fridges and freezers are nor- mally closed. An alarm to tell you that it isn't is the subject of this article. It is based on a light- dependent resistor ILDR). As soon as the door of the fridge, or freezer, being guarded is opened, light falls onto the LDR: the circuit is then ac- tuated and a warning tone is sounded until the door is closed The circuit may also be used to monitor other doors (for instance, to prevent heat loss, or as a precaution against a fire spreading), but because of the ambient light it is of course impossible to use an LDR. This can therefore be replaced by a microswitch, in which case the alarm will sound when the switch is closed. Note that this requires a switch which closes when the door is opened. A delay of about 10 s between the opening of the door and the sound- ing of the alarm is provided by the time constant R3C4. If faster reac- tion of the circuit is required, the value of R3 may be reduced to 220 k. At the moment the threshold of N1 is exceeded, the gate commences to oscillate at a frequency of a few hertz. Each consequent rectangular pulse at the output (pin 3) of inverter N2 fires oscillator N3 which generates pulse trains whose rate amounts to a few kilohertz. The pulse trains are fed to inverter N4 which causes the piezo buzzer to emit a tone. Without N2, oscillator N3 would work continuously when N1 is not being triggered: the output of N1 would then be high, and the logic 1 at pin 8 of N3 would cause the oscillator to function. Inverter N4 serves to amplify the output of the buzzer. If the buzzer would simply be connected between the output of N3 and earth, the membrane would merely move from its rest position to one side. By con- necting the buzzer across an inverter, its polarity is constantly reversed and this causes a doubling of the alter- nating voltage across it. Preset P2 i Aug/Sept 1 984 8.19 Table 19229 9614 4807 2404 0.15% 0.15% I 0.15% 0.15% 0.68% I 0.15% with six switchable settings Most asynchronous receiver/transmit- ters (generally known as UARTs) operate at a clock frequency which is sixteen times the transmission rate. There are special ICs available that are dedicated to this particular timing function but they are neither freely available nor cheap. The classic switchable oscillator/divider circuit is, however, a good substitute. The clock frequencies provided in the design shown here correspond to the standard transmission rates of 1200, 600, 300, 150, 110 and 75 baud. The oscillator is based on inverters N1 and N2, in combination with the 1 MHz crystal. Its signal is fed, via N3, to the first 4024. A flip-flop, con- sisting of N9 and N10, is included in the reset line to this divider to ensure that the reset is perfectly synchron- ised with the clock signal. The out- put signal from IC1 feeds the clock input of the second 4024, which can be 'programmed' by means of SI (a double-pole six-way wafer switch). The reset pulse for this seven-stage binary counter is provided by N5, N7 and N8 and travels via a second synchronising flip-flop (N11/N12), which is, once again, clocked by the oscillator signal. The 'programming' switch, SI, can be replaced by a pair of wire jumpers in the appropriate places if one baudrate is continu- ously selected. The table here indicates the frequen- cies measured in our prototype cor- responding to the various baudrates. The exact frequencies anticipated are also given in order to show that the error is negligible in all cases. The principle used in this circuit could quite easily be extended to make it suitable for a different crystal than the one stated. It will then be possible to make use of an 'old' crystal instead of having to buy one. H for good-old-fashioned radio Nostalgia is becoming ever more of a 'big business'. This is understandable as we all have a tendency to remember 'the good old days' when life was simpler, less complicated, and everybody was happy. In reality, of course, it was quite different but we often prefer to let our minds play I tricks on us so we only remember 8.20 elektor irtdia Aug/Sept 1 984 the good things. Part of the nostalgia for many electronic hobbyists has to be the original valve radios. They were a breed apart, with the way they looked and the sounds they made, and it was impossible not to be takenjn by their magic. Many at- tics still hide these radios, which no longer work, so we thought it would be interesting to fit a transistor radio into one of these old cases and add a bit of 'magic' of our own. Then your modern radio is guaranteed to make genuine 'antique' noises and it even takes a while to warm up just like valves always do. The circuit is based on two oper- ational transconductance amplifiers (OTA), one of which, IC1, transmits the 'valve' hum while the other, IC2, transmits the audio signal that goes to the final amplifier. The outputs of the two OTAs are connected together so that the amplifier they feed receives a mixture of two signals. The volume of the hum, which is taken from the transformer's second- ary winding, can be set with preset PI and the signal level is set with P2. The gain of each OTA is deter- mined by the bias current applied to pin 5 of the 1C. The actual 'valve' se- quence of silence — loud hum — hum becoming quieter — rising sound is generated by two monostable multivibrators. When the supply is switched on MMV1 is first triggered via R1 and Cl, with the result that the Q output goes high. At the same time tran- sistor T1 prevents any bias current from being passed to IC2 so no audio signals are transmitted by this OTA. The Q output of MMV2 is still low so IC1 also receives no bias cur- rent and only silence is heard. After about seven seconds the Q output of MMV1 goes low again so MMV2 is triggered. Its Q output goes high causing the bias current fed to IC1 via the R5/C5 combination to in- crease gradually. Even though T1 is no longer conducting the 'O' level at the Q output of MMV2 prevents IC2 from transmitting the audio signal for the time being. After about five seconds the output of MMV2 changes so that Q becomes '0' and Q becomes 'V. The amplification of IC1 then drops slowly and that of IC2 rises slowly. Because of this the hum reduces gradually and the sound (music, or whatever) increases gradually until it finally drowns out the hum. The symmetrical power supply for the circuit is based on a pair of voltage regulators, IC4 and IC5. Cur- rent consumption is less than 10 inA so the circuit could be powered from the existing supply in the radio. If this is done do not forget the con- nection from the secondary winding of the transformer to provide the hum. M helps a computer to get to the point covered light dependent resistor (LDR), the exposed part of which forms a window the same size as a character on the screen. When the electron beam in the screen passes in front of this window the LDR's resistance reduces drastically. This causes transistor T1 to conduct, followed by T2, with the result that a pulse suitable for the LPEN input of the CRTC appears at the collector of the BC 559. As soon as this latter transistor saturates T3 switches off and the LED extinguishes, indicating that the lightpen is correctly pointed at the character. For correct oper- ation of the lightpen, preset PI must be 'calibrated'. This is done by placing the LDR, which is screened A light pen is a tool that allows the coordinates of a point on the screen to be entered into a computer. It is based on the principle of sending a pulse to the screen control circuit at the precise moment when it sweeps the spot just in front of the lightpen. In the case of the Elektor VDU card the screen is controlled by a 6845; when this IC's LPEN input (pin 3) goes from '0' to 'V it loads the ad- dress of the character it is writing into registers 16 and 17. We will see later what can be done with this information. The sensor in the lightpen is a partly. ) Aug/Sept 1984 8.21 as shown in the sketch, in front of a character and then trimming the preset until the LED is extinguished. Registers R16 and R17 in the 6845 CRTC store the address of the character indicated by the lightpen. This address is somewhere in the range from 0000 to 3 FFFheX- which is the 16 K of screen memory ad- dressable by the 6845. All that then remains is to convert this address into usable information. It could be considered as an index specifying the offset relative to the display start address. When these two are added the address indicated by the lightpen is obtained and could have a character 'POKEd' to it. Another possibility is to move the cursor to this point. The information provided by the CRTC must then be converted to X and Y coordinates (vertical and horizontal) which are used to modify pointers COL (ver- tical) and INLINE (horizontal). The ACURC routine (see the listing in Paperware 3) is called to move the cursor to this address. As the flow chart indicates, the information pro- vided by the CRTC must be cor- rected because in the Elektor VDU card the DF.N and CUR signals in the CRTC are delayed by flip-flops FF1 . . . FF4 in order to compensate for the delay inherent in the data handling chain.The character in- dicated by the lightpen is therefore not in the address indicated by the CRTC but in the one immediately following it. M The output of N10 also triggers monostable MMV which then switches on a simple oscillator based on N4 and the buzzer sounds, pro- vided the switch is closed. The circuit is self-setting, that is, when power is switched on, it automatically resets to the stop con- dition in preparation for the forth- coming timing cycle. To this end, the 100 Hz signal at the output of N9 is passed to the clock input of IC3 which counts down to zero very rapidly. The ZD output (low! sub- sequently resets IC1 and IC2. Pressing the stop switch to abort a timing cycle has the same effect as The timing range may be expanded or shortened by adding or omitting one or more 4518 counters. The count-down cycle may be fixed per- manently to a specific time lapse by replacing the BCD thumb-switches with hard-wiring at the J inputs of IC3. As there are a lot of spikes in this circuit, good decoupling is essential. A 100 n capacitor should be provided directly across the supply pins of each 1C. H r mdia Aug/Sept 1 984 8.2 3 dia Aug/Sept 1984 8.25 gives double protection Warning lights fulfill an undeniably important role in many technical in- stallations. However, even in the best of equipment, these lamps can fail. A glowing filament was never in- tended to have an indefinite lifespan. The circuit here cannot prevent the filament from failing, but it ensures that if the warning lamp cannot light, for whatever reason, a reserve light is automatically switched on. This secondary bulb, moreover, will only light when it is absolutely necessary, that is to indicate a fault in the equipment. Apart from the two lamps, the total component count for this circuit is just two transistors and two resistors. The principle of the circuit is very simple: assuming there is a fault in the equipment, lamp Lai lights and a small part of the lamp's current flows to the base of T1, causing this transistor to conduct. As a result of this the base of T2 is effectively shorted to earth and this transistor cannot conduct. No cur- rent flows through the reserve light (La2) in the collector line of T2, so La2 is not lit. As soon as Lai goes out, due to a bad contact, for example, or because the bulb is blown, the base current to T1 is cut off so this transistor im- mediately switches off. The current that flows through R2 then causes T2 to conduct and the reserve lamp lights. Lamps which need a higher voltage than the 12 V given in our diagram can, of course, also be used in this circuit configuration. The com- ponents must then, however, be modified to suit the new situation. IN ITT application 8.26 more, in that case Z is linked to Y. If fuse FI is used, however, Z is linked to X. A ready-etched printed-circuit board is not available for the monitor, but a pcb may be made by yourself from the track layout given on the pc board pages. Finally, please note that IC1 should be powered before the protected supply. M big brother is watching you . . . This monitor circuit is based on the MC 3424 power supply supervisory 1C. It provides two-channel over- voltage crowbar protection, which is * very useful for floppy disk systems, and overvoltage and undervoltage monitoring of the +5 V line, which is particularly important in micro- processor supplies. Each channel in the MC 3424 has an input and an output comparator. Channel 1 is the undervoitage monitor, while channel 2 provides crowbar overvoltage protection. The input comparators sense the regulated supply line (pins 3 and 151. Each of them provides a common- j mode range of 0 . . . (V cc — 1.4 V) volts. The source resistance of the inverting inputs determines the amount of hysteresis. An on-chip generated reference voltage of 2.5 V (available at pin 1) is permanently connected to the non- inverting input (pin 2) of com- parator 1 and to the inverting input (pin 14) of comparator 2. When the voltage on the supply line drops below about 4.2 V, the input t comparator of channel 1 (pins 2 and 3) changes state which causes a low logic level at pin 6 and the red LED, D1, lights. The LED could be re- placed by an interrupt routine in the computer to safeguard stored data and to switch over to the back-up battery. When the supply line rises above about 6.2 V, the input comparator in channel 2 (pins 14 and 15) changes state and pin 10 of IC1 becomes logic low.The silicon-controlled rectifier (SCR), Thl, then fires and short- circuits the supply to earth. Depen- ding on whether the stabilized 5 V or the unstabilized line (A) is connected to the anode of Thl (wire link XZ or YZ), the supply is cut off either by fuse FI in the 5 V line blowing or the short-circuit across the smoothing capacitor in the protected power supply. Note that the 1N4001 diode in the protected supply is essential to safeguard the stabilizer. If the protected power supply is already fitted with a fuse, a wire bridge should be soldered on the pc board instead of fuse FI. Further- elektor india Aug/Sepl 1 8.27 the ratio of R2 + R3 + PI to R3 + PI. The input impedance, which, at 1 M, is quite high, is de- fined by R1 as the op-amp has FET inputs. This is a suitable impedance for most guitar pick-ups. A 9 V bat- tery provides the power supply which is converted to a symmetrical + and —4.5 V for the op-amp by means of R4, R5, C3, and C4. Current con- sumption is about 5 mA. The circuit complete with battery can easily be fitted into a small case. If a socket and plug are mounted in the gives all the boost that is needed input stage of the guitar amplifier is guaranteed to clip. As an aid to this the gain can be selected between three and eleven times. The layout of the circuit is very case, as the photo shows, the preamp can simply be plugged into the guitar. If this is done, preset PI can be replaced by an ordinary potentiometer so that the amplifi- The output signal level provided by many electric guitars is not high enough to overdrive a valve amplifier. This overdriving is an essential part of the final guitar sound. The preamplifier circuit shown here simple. A single LF 356 provides the cation can be controlled by means of boosts the guitar signal so that the amplification, which is decided by a knob on the case. H an MKT type, charges via the skin resistance, which causes the output voltage of IC1 to drop linearly to zero. When the other touch-pad, Se2, is touched, the opposite hap- pens: the output potential of IC1 will then rise linearly until it reaches the level of the supply voltage. The beauty of the circuit is that when you take your finger from the pad, maintains the set value la Touch-pad keys normally use a simple digital memory, but they can be operated to give an analogue out- put voltage as is shown here in an inexpensive circuit that is easy to build. The circuit is based on IC1, an oper- ational amplifier with very high input impedance, which is connected as an integrator. When touch-pad Sel is touched with a finger, capacitor C2, 8.28 elekior indiaAug/Som 1984 1b the value of the voltage then present at the output of IC1 is maintained by the charge on C2. Owing to unavoidable leakage currents in the capacitor, the output voltage will, however, drift by about two per cent per hour towards zero or towards the supply voltage, depending on which of the key pads was touched last. To keep these leakage currents small, it is necessary to keep the circuit well away from moisture or humidity, which should be borne in mind when choosing a case. The range of applications for this cir- cuit is wide: it may be used anywhere there is a potentiometer that can be controlled by a variable voltage. If you prefer to use normal push but- ton switches instead of touch-pad types, figure 1b shows how to con- nect these in the circuit. Resistors R3 and R4 simulate the skin resistance; switches SI and S2 provide the input voltage for IC1. Pressing the switches simultaneously has no effect. Capacitors C3 and C4 obviate any tendency of the operational amplifier to oscillate. M klor India Aug/Sept 1984 8.29 relay Re is actuated until either the indicated time has lapsed or SI is pressed. Indicating seconds: close S4. Alarm: when the alarm goes off, it may be stopped with S9 or S2. Keeping S2 closed disables the alarm permanently. Pressing SI stops the alarm temporarily: after 8 ... 9 min- utes it goes off anew. If the alarm is not switched off manually, it stops automatically after 59 minutes. If the clock is used without a crystal time base, an external 1 kHz signal must be provided for N4, otherwise the alarm does not work. To reduce power consumption (par- ticularly in cars), it is possible to switch off the display with S8. This switch may be combined with the ig- nition switch. Current consumption with the display on is about 200 mA, dropping to about 20 mA when the display is switched off. M is used to actuate meter Ml. Meter Ml is connected at the diagonals of a bridge circuit, so that it indicates zero in the absence of C x or L x . The meter is set to full-scale deflec- tion (f.s.d.) by P2 when either L x = L 0 or C x = C 0 is inserted into the oscillator circuit. The tolerance of with moving-coil instrument An LC meter is undoubtedly in- dispensable to anyone involved in h.f. techniques. The present design accepts the unknown inductance, L x . or capacitance, C x , in a two- transistor oscillator circuit, of which the output voltage is kept constant between 30 and 40 mV by a regu- lator. When, in the oscillator circuit, C x is connected in parallel with capacitor C 0 , or L x in series with inductor L 0 , the frequency of the circuit diminishes. This diminution is measured by a frequency-to-voltage converter, T3/T4. The consequent output voltage of emitter follower T5 frequency-determining capacitor Cj_ in the frequency-to-voltage converter is balanced out by PI. It is therefore necessary to switch in a different 10 k preset for each measuring to help your guests This circuit should put an end to your guests fumbling for the light switch in the cloakroom. It ensures automatically that the light is switched on as soon as someone enters the cloakroom and is switched off again when that person leaves. The principle of the circuit is fairly simple. The bistables in a 4013 CMOS 1C are connected in series. One, FF2, is arranged as an R-S latch to debounce the switch. This soon as the door is opened and is therefore best located in the door- frame. When the door is opened, FF2 is set and its output (pin 13) goes therefore high. This clocks FF1 on pin 3 and this bistable toggles: its output goes high and this switches on transistor T1. The transistor cur- rent actuates a relay and the light is switched on. When the door is closed, nothing happens because FF2 is reset and its output on pin 13 goes low. It's only when the door is causes FF1 to toggle. The output at pin 1 then goes low and cuts off the transistor so that the relay is deenergized and the light goes out. The relay should operate from voltages between 5 and 15 V. Because the door may be opened and closed without anyone entering, it is possible that the light gets out of step with requirements. This can, of course, be remedied by opening and shutting the door again, but a better way is to connect a second switch, S2, as shown in figure 2. This switch reverses FF1 and brings matters back into step. The circuit diagram in figure 1 shows SI in position 'door open' and bistable FF2 is then set. Hi .... for cassette interface Compact cassettes, because of their low cost and easy availability, have been the mainstay memory in per- sonal and hobby computers for almost as long as these have been available. These cassettes convert digital computer data into audio signals and vice versa. They can, however, not prevent drop-outs caused by wrongly set signal levels. The present level indicator can help to prevent these mishaps. All that’s needed to build the in- dicator is a 3.5 mm jack plug, two LEDs, a resistor, small loudspeaker, and a jack socket. The LEDs are connected in anti-parallel. The loudspeaker serves as a monitor to indicate whether the recorder is emit- ting signals (audible as two quite distinct tones), or whether the cassette content is between two pro- grammes (when only a slight hiss is The indicator is connected to the ear-piece socket on the recorder by the jack plug and the cassette interface input via the coaxial socket. Most cassette interfaces need a signal level of 2 V p p. When the signal provided by the recorder is at about this level, the LEDs begin to flicker; if the level is too high, they light continuously. If the loudspeaker volume is too high, connect a 100 Q preset in series with it, so that the volume may be adjusted to personal requirement. M — Pin 20, incorrectly called CS (chip select) on the_2708 while its func- tion is actually OE (output enable), retains the same function. - Pin 19 (+12 Von the 2708) becomes address input A10 for the 2716. Depending on the logic level on this pin either the first or second 1 K block is selected. A switch could be used for this so if the EPROM contains a monitor, for example, two different versions of r software could be stored in the Si e 1C. The 2708 EPROM has become vir- tually obsolete, and with good reason. It needs three supply voltages for its capacity of 1024 x 8 bits whereas its immediate successor, the 2716, uses the same 24-pin package but only needs a single supply voltage for twice the memory capacity (2048 x 8 bits). Furthermore, the 2708 has become so difficult to find that it has become more expensive than the 2716, and that alone is reason enough to con- sider the modifications needed to substitute one for the other. For- tunately, few changes are required as the address decoding remains the Most of the 2716 pins are directly compatible with those on the 2708 that they replace. The following pins are, however, worthy of note: - Pin 21 (-5 V on the 2708) must be connected to +5 V for the 2716. - Pin 18, which is connected to ground for the 2708, need not be changed for the 2716 (CE, chip enable); note in passing that the 2716 will then never achieve the minimum power dissipation of 132 mW (stand by current). e several different ways of carrying out these modifications. An intermediary 1C socket could be used with the pins that are to be changed >t inserted into the socket but wired separately. If preferred, the e thing could be done without using a socket. The method we recommend, however, is to modify the printed circuit board by cutting the appropriate tracks. Be especially careful if this is done with a double- sided board. H 8.32 e., transmits serial computer information In general the normal connections between a computer and its peripherals are very effective but these cables could hardly be con- sidered decorative. A cable carrying serial information can, however, be replaced by this infra-red interface even though it only consists of a simple transmitter and receiver. As figure 1 shows, the transmitter uses a single BC 557B transistor to drive the infra-red LED. The transistor is itself controlled by the micro- processor so a short program is required to make the computer generate the transmitter signals needed. The frequencies used here are 4800 and 9600 Hz and the maxi- mum baud rate at these frequencies, is 1200 baud. The receiver, seen in figure 2, makes use of an 1C (the SL486I especially developed for infra-red applications. This contains several gain stages, a pulse-width expander, and a voltage regulator. The receiver diode (Dll is connected directly to the 1C. The stretch output, pin 11, is connected to the low-pass filter made up of R1, R2, C9. and CIO and this, in turn, feeds schmitt trigger IC2. The de- coded data is then available at the output of this 1C. When fitting the components to the printed circuit boards shown in figure 3 it is important to remember that the leads for the receiver diode should be kept as short as possible. The 5 V supply for the boards can be taken from the computer or peripheral device. The only cali- bration needed concerns preset PI which must be trimmed so that the data is received with no errors. M D1 = infra-red LED. e.g LD 271 D2 . . . D4 = 1N4148 T1 = BC 5578 C3 - 6p8 (4(471/10 V C4 = 68 (. (47 j.l/10 V C5.C6 = 33 n C7 = 10 (./10 V C8 = 150 n C9 = 18 n CIO = 6n8 D1 -- infra-red c e.g. BP 104 IC1 = SL486 IC2 = CA 3130 elektor india Aug/Sepl 1 984 8.3 3 it least 120 W into 4 Q The TDA 2030, made by SGS Ates, I is a complete amplifier contained in a 8.34 elektor india Aug/Sept 1984 single 1C with a five-pin Pentawatt package. Its class AB output stage can provide a power of 14 W into 4 S at a supply voltage of ±14 V. The amplifier has a built-in short- circuit and overload protection, and also a thermal shutdown. This means that it is not so easy to destroy the 1C as long as the supply voltage is kept below the absolute maximum of +18 V. Combining two 2030s with a few in- expensive power transistors can form an amplifier that can drive quite a lot of power into a load of 2 to 4 Q, As the diagram shows, the circuit is a standard 'bridge' amplifier so there is little to be said about it. Each half of the bridge consists of a TDA 2030 driving two complementary power transistors. The diodes, D1 . . . D4, are needed to protect the transistors from the loudspeaker coil's inductive rindia Aug/Sept 19848.35 a simple resonant frequency meter Way back when electronics was still young (shortly after the stone age) when grids, anodes and cathodes were all the rage, this device would have been called a grid dipper. Now it is more likely to be called a dip meter or a transistor dipper. No matter what it is called it is still the same instrument, and in the handy transistorised form shown here it is an indispensable aid for any HF handyman. Before we start describing the cir- cuit, we must first establish exactly what a dip meter is. A dip meter could be considered as a sort of fre- quency meter whose purpose in life is to define the resonant frequency of LC circuits. The circuits do not have to 'radiate' (in other words, they do not have to be in an oscil- lator circuit), as they can be measured, or, to be more exact, dipped, 'loose'. To see how this works we can best go straight to the circuit diagram. The parts that make up a dip meter are always the same: a tunable oscillator, a rectifier and a moving coil meter. The oscillator here is based on T1 and T2, and is tuned by means of capacitor Cl and coil Lx. This coil is fitted outside the metal case into which the circuit must be built, and must be easily exchange- able for a different coil to enable the range to be changed. When the dipper is switched on the oscillating voltage generated is rec- tified (by D1 and C2I and is then passed to the meter via PI, which adjusts the meter reading. Nothing unusual so far, but now comes the interesting bit. If Lx is inductively coupled with the coil of some other LC circuit, whose resonant frequency is the same as the oscillator fre- 8 . 36 . i Aug/Sept 1984 quency of our dipper, this other coil will draw energy from the oscillator coil. The result is that the voltage across the meter drops and the reading is reduced. What happens in practice is this: the dipper is switched on and PI is ad- justed so that the meter gives maxi- mum, or almost maximum deflec- tion. The coil of the LC circuit to be measured is now placed close to Lx and Cl is adjusted until the meter reading shows a clear 'dip'. The fre- quency can now be read off from the graduated scale on Cl. This graduation is where Murphy gets his hand in. A second, graduated, dip meter or — even bet- ter — a frequency meter is needed for this. With the layout shown here and using a free wound coil (without a former) of 2 turns of 1 mm (SWG 19) copper wire with a diameter of about 15 mm, the range of the dip meter is about 50 ... 150 MHz. A coil could be wound on a DIN plug and a DIN socket mounted in the case of the dip meter to facilitate easy changing of the inductor. A few points to note. The BF 494 transistors in the oscillator can only handle up to about 150 MHz. If higher frequencies are contemplated these transistors must be replaced by another type, such as the BFR 91, which should allow up to 250 MHz. There are various different possi- bilities that can be used for variable capacitor Cl. It could, for example, be the 50 pF capacitor from the Jackson C804 range, or a cheaper solution is to use two 100 pF mica capacitors connected in series. Another possibility is to get hold of an told) four-gang FM tuning capaci tor and link the four sections, each of which is about 10 to 14 pF, in parallel. Finally: any dip meter, including this one, can, in principle, also be used as an absorbtion meter or field strength meter. To use it in this con- figuration, leave the voltage supply of the meter off and look not for a dip but rather for the maximum reading on the moving coil meter. M r india Aug/Sept 1 984 8.37 with good accuracy at low frequencies Intersil 1C type 7226B is just the right counter for a simple but reliable fre- quency meter which covers a range of 9 MHz. The circuit of the meter divides into four functional sections: ■ input stage, T1, T2, N2, N3; ■ multiplier, FF1, FF2, IC3, IC4; ■ counter, IC5; ■ display, Ldl . . . 6 In general, the circuit is a standard design, much of which has been described in earlier issues of Elektor. The primary function of the input stage is converting the input signal into rectangular pulses that are fed to the counter either direct or via the multiplier. The stage can handle in- put voltages of up to 50 V r.m.s. which is sufficient for most measurements. Diodes D1 and D2 conduct when the input voltage is above about 600 mV so that the in- put impedance is determined primar- ily by the value of R2. that is, around 1 M. The multiplier I x 100) is particularly important for the measurement of frequencies between 5 Hz and 1 ... 2 kHz. The counter, the Intersil 7226B, con- tains a crystal oscillator, a time base, a counter, a seven-segment decoder, a multiplexer, and a number of drivers for the direct control of the LED display. In our prototype a 1 MHz crystal was used for driving the on-chip oscil- lator, but if D5 is omitted 10 MHz crystal may be used. The LED display is the popular type MAN 4640A. The function of the switches is: ■ Sla connects the input stage to the counter either direct or via the multiplier (as shown); ■ Sib ensures the correct position of the decimal point when the multiplier is in circuit; ■ S2 normally determines the pos- ition of the decimal point, that is, whether the display reads kHz or MHz; ■ S3 is the mains on/off switch; ■ S4 is the reset switch; ■ S5 serves to test the display: when it is pressed, all segments should light. Finally, note that printed circuit 84462 for the meter has no provision for the display; this may be fitted on board 80089-2 originally designed for the Junior Computer. K 8.38 elektor india Aug/Sept 1984 switches sounds electronically In a few situations it is desirable to be able to switch audio signals elec- tronically. One such situation is the muting circuit in FM receivers that substitutes silence for annoying noise when there is a very weak, or no, carrier signal present. This is the sort of switching function that electronic switch ICs, such as the 4016 or 4066, The circuit shown here works on a symmetrical ± 7.5 V supply but the 1C containing the switches simply has +15 V. This set-up allows the circuit to work with pure a.c. signals. The operation of the circuit is quite straightforward. If the control input (x) is at +7.5 V ESI will be closed and ES2 open. The effect is then that of an inverting unity gain amplifier. If, on the other hand, x is at —7.5 V ESI opens and ES2 inverting input via ES2. The voltage division, and therefore also the out- put voltage, will therefore be zero. The clever part of this design is that ESI is connected in series with A1 so it is part of the section that decides the amplification. The non-linear behaviour that could otherwise cause distortion of the signal is then to a large extent compensated by the feedback so that even hi-fi signals are not adversely affected. For this sort of application it is also a good idea to use a low-noise op-amp. closes. The circuit then acts as a voltage divider, one side of which is made up of R1 and R2 in series and the other side is the very low output impedance of the op-amp. This out- put impedance is so low (R()/(1 + Ao>) because the output is simply fed back straight to the Apart from the TL084 shown in the circuit diagram, some possible substitutes are TL074 (low-noise!), LF356, or RC4136. A single 4066 contains four elec- tronic switches so the circuit can easily be doubled to switch two channels, for stereo, for example. M based on a current source The 78XX/79XX series voltage regulators have become so typecast in their main application that we tend to forget that they can also be used as current sources. This sort of current source can be very useful in many applications, such as charging NiCad cells, as in the circuit here. We will start by taking a look at the principle of the circuit, which, as the diagram shows, is very simple. The circuit is based on a voltage regulator with a constant load be- tween the two points across which the fixed voltage is found. The result is obvious: voltage and load are both fixed so unless Mr. Ohm got it very wrong the current is also constant. The whole unit (voltage regulator and load) can then be connected in series with a (variable) load, in this case the nickel cadmium battery that is to be charged, and the current will still remain constant. This is, of course, always pre-supposing that the input voltage is high enough. The circuit has on subtle 'extra', in the form of a LED in series with the ground pin of the 1C to indicate charging. A fixed current of 8 mA ±1 mA, which is dependent on the output current selected and which must be added to this output cur- rent, flows through the LED. When the value of R1 is being decided the extra 1.5 V dropped across the LED must be taken into account. As we have already said, we decided to use this current source as a charger for NiCad batteries. Unlike lead-acid batteries these have to be charged at a constant current. Stan- dard NiCads should be charged at a current which is 1/10 of the nominal capacity for 14 hours. Batteries that are not completely discharged do not need this long. In general the bat- teries will not be damaged by charg- ing them for more than the rec- 8.40 elektor India Aug/Seoi ' ommended time. It is advisable to discharge NiCads completely from time to time and then charge them again immediately as this can help make them operate at maximum ef- ficiency for as long as possible. The table shown here indicates several different types of batteries complete with the recommended charging current and the value that should be used for R1. The charging current can be made more exact by using a number of resistors in series to get the correct resistance. The values given here are the nearest standard values. If the charging cur- rent is more than about 150 mA the half- rectification provided by D1 of the transformer and the values of should be changed to full-wave tification by substituting a rectifier here. The job of the smoothing capacitor. Cl, is thereby also eased. The maximum number of cells that can be charged at a time depends c the transformer voltage. At 15 V this M.S. Dhingra is four (depending also on the quality charging current and smoothing capacitor), at 24 V ten cells can be charged. The output current from the transformer must be 1 54 times the charging current. i mdia Aug/Sept 1 984 8.4 1 It is true, of course, that it has been possible for many years to release a camera shutter remotely. But that is normally done by a (too) short cable, which can be a nuisance, and which is invariably too expensive for what it does. It seems therefore a good idea to release the shutter optically: the only prerequisite for this is that the camera is provided with an electronic shutter-release facility. The proposed circuit (see figure 1) is built into a small case (see figure 2) which is fitted with a flash connector enabling it to be fixed to the camera instead of the flash unit. The little case may be made from the screen- ing can of an r.f. or i.f. transformer. The simple circuit is based on a type CA3140 opamp which has been con- nected as a differentiator. To drive transistor T1, the inverting input of the CA 3140 must be fed with a short negative pulse. This is here obtained from quick changes in the light in- cidence onto either of the light- dependent resistors (LDRs), R1 and R2. Two LDRs are needed to provide the difference potential to which the remote shutter release IC reacts. It does not matter whether this dif- ference is caused by a shadow falling onto R1 or a flash of light onto R2. The CA 3140 does not react to slow changes in incident light because of Cl. Ambient light, which falls equally and simultaneously onto both LDRs, has no effect. The negative pulses are inverted by the opamp and then used to switch on transistor T1 for an instant. The consequent short burst of current is sufficient to release the camera shut- ter. If you use a flash of light to operate the camera, you may be quite a distance away from it. par- ticularly in the dark. The circuit may be matched to a variety of cameras. The onset of conduction of T1 is dependent on the value of R4, while the sensitivity of the circuit may be augmented by increasing the value of R3. Current consumption depends on ambient brightness and lies between 1.5 and 5 mA. H The 'alarm extension' described in ! the October 1983 issue of Elektor, when used with a telephone cannot make a clear distinction between tne call tone and the dialling tone. If this distinction is important, you may find the present filter circuit of interest. The circuit consists basically of ac- tive filter stages A2 . . . A4, trigger stages A5 . . . A7, and digital filter stages MMV1 and MMV2. The level at the output of the circuit is CMOS compatible and is logic 1 when a call signal has been detected by L4. The output signal may be connected direct to the transmitter (pin 4 of IC1 in figure 2, page 10-53, Elektor October 1983). The supply voltage. U|j, is halved by A1 and the operating voltage for the circuit is then taken from across C12. The signal provided by pick-up coil L4 is rich in harmonics: after amplification in A2 it is applied to active low-pass filters A3 and A4 which remove virtually all harmonics. The cut-off frequency of A3 is 10 Hz and that of A4 is 25 Hz. The output of A4 is then either the 25 Hz (sinusoidal) call tone or the dialling pulses (= 10 pulses) whose rate is 10 Hz or below. When for instance '0' is dialled, the dial takes 1 second to return to rest. These frequencies can be clearly identified by the following stages. When the output signal of A4 reaches its positive peak, a positive pulse is generated by A5; when it reaches its negative peak, a negative pulse is generated by A6. Operational amplifier A7 then gives an unam- biguous control signal to IC3, the all the 'uneven' lines are grounded in order to protect against cross-talk. The second 'difficulty' with connec- ting the 8" drive is that it has four lines which the 5 '/*" unit does not. This is not strictly true, but in the smaller drive they are combined with other signals. The 'ready' is com- bined with 'index', and 'head-load' with 'select'. 'Low current', on the other hand, which is only used on the 8" drive, limits the magnetising current when the inner tracks are being written. 'Side select' is, of use 8" drives with a 5% " interface Someone once said 'Small is beautiful' and we wholeheartedly agree with him. In the case of this circuit, 'small' refers to the number of components it uses. It shows that two connectors (one 50-way and one 34-way) and a length of 50-way rib- bon cable is all that is needed to connect an eight inch floppy disk drive to an interface intended for a five and a quarter inch diskette. What we are actually doing here is making the computer 'think' that the 8" drive is in fact a 5 % " unit. In order to achieve this two things are needed. First a cable must be made to link the 8" drive's 50-way connec- tor to the 34-way one on the floppy disk interface. Note in passing that i Aug/Sept 1984 8.43 e modification needed to The re These four lines can be left uncon- the hardware of the floppy disk ir nected if they are first hardwired within the disk drive. The 8" drive can then be considered as a 5" one face to allow this circuit to operate diskette so one an at its best. The clock frequency of much information the ACIA on the interface card, any track. I except that it has 80 tracks instead which takes care of the parallel to floppy rotates faster than a 5" diskette so one and a half times as much information can be written on any track. H The circuit presented is not only a suitable replacement of an existing mechanical rev counter, but may also be fitted in cars that have not one. The counting and memorizing occur in IC2, which also arranges the multiplexing of the signals to the three-segment display. Suitable units of time are arranged by oscillator N3 whose frequency is set to 1/3 Hz by PI. NAND schmitt trigger N2 generates a short negative pulse at the leading edge of the oscillator pulses, that is once every three seconds. The content of the counter is then stored in the on-chip memory from where it will be provided to the display. Network C3/R7/D4 resets the counter at the leading edge of each LE pulse: a new counting cycle then starts. The number of revolutions is derived from the pulses generated by the contact breaker in the car. These pulses are first limited by zener diode D1 to 12 V and then used to set transistor T1 into conduction via D2 and R2. Network C2/R6 arranges a logic low at one of the inputs of NAND schmitt trigger N1 for a brief moment: the output of N1 is then 1 and this clocks IC2 so that the counter content is increased. The clock input is decoupled by D3 and R4. The display is controlled via T2 . . . Til and IC3, which arranges the BCD-to-seven-segment conver- sion (BCD is binary coded decimal). Transistors T2 . . . T4 pass the multiplexed signals to the display, while the remainder of the transistors are simply buffers. Resistors R13 . . . R19 limit the current in the display. The car battery voltage is reduced to 8 V by regulator IC4 and this then serves as the power supply for the rev counter. Current consumption amounts to about 150 mA. The rev counter may be calibrated with the aid of the mains frequency. Connect an alternating voltage of 6 ... 15 V (for instance, from the secondary of a bell transformer) via a single diode (polarity!) to the input (R1) and adjust PI until the display O O Cl Cl. = Cl IJ sex arf ® “T reads 1.50 (simulating a four-stroke, four-cylinder engine running at 1500 r.p.m.). The display reading should be multiplied by 1000 to obtain the revolutions per minute. H 8.44 elektor india Aug/Sepl ' When enough light falls on T1 it conducts and earths the base of T2 so the lamp extinguishes. The two transistors are coupled via R2 and PI to improve the switching behaviour. The light intensity at which the cir- cuit switches is set by means of preset PI . N from an idea by H.J. Hooft the gate current to Thl to drop below the holding value. When that happens, Thl cuts off and switches the charging current to the battery off: the ammeter then reads 0. When after some time the battery voltage drops below its nominal value again, the gate current to Th2 drops. SCR Th2 then cuts off, the gate current to Thl increases, that SCR conducts, and a charging current flows again to the battery. The charger is calibrated by connec- ting a fully-charged battery in the cir- cuit and adjusting PI so that just no for hibernating batteries This is a charger designed specially for batteries that are not used for long periods, e.g. those of motor- cycles that are laid up for the winter. You then take the battery from the vehicle and connect it permanently to the charger, which is switched on one or two times during a week. The battery is charged and when it has reached nominal voltage, the charger switches itself off. It remains on standby, however, and when the bat- tery voltage has dropped below the nominal value, it switches itself on Suppose that the battery in the cir- cuit diagram has a voltage below its nominal value. As soon as the charger is switched on, a current flows through D3 to the gate of silicon-controlled rectifier (SCR) Thl. The SCR conducts and a charging current flows through the battery that is indicated by ammeter M. The battery voltage increases gradually and with it the potential across R1/P1. Capacitor Cl then charges and at a given level of voltage across it, zener diode D4 starts to conduct. A current then flows to the gate of Th2: this SCR conducts and causes current flows through the ammeter. Note that the transformer must not be called upon to deliver more than 5 A charging current. This is because when Thl conducts its only load is formed by the transformer secondary and the battery. For security of operation, Thl should be able to switch currents of up to 10 A: the TIC 236A and TIC 246A, for instance, meet this requirement. The same applies to the silicon rec- tifier diodes for which the SKN26/04, SD25, BYS24-90, for in- stance, are suitable. The maximum forward current of these diodes should not be below 8 A! K General Electric, Auburn, NX, U.S.A. Notice 630.15 provides 1 x 5 V and 3 x 12 V The attraction of this power supply is based on two of its characteristics. First of all it is extremely compact and secondly it supplies three, or strictly speaking four, voltages: + 5 V/3 A, +12 V/2 A, and a sym- metrical ± 12 V/ 250 mA. The credit for the compactness is due to the fact that only one transformer is used for the three voltages. It is a toroidal transformer made by ILP (they call it a 4T344) and has three secondary windings of 9 V/7.2 A, 15 V/ 3.2 A, and 15 V/0.5 A. It could, of course, be replaced by three separate transformers but then the circuit would lose a lot of its charm. We felt no need to re-invent the wheel as regards the voltage regulating circuitry. A pair of 723s followed by TIP 142s to do the heavy work are used for the 5 V and 12 V. The symmetrical + 12 V is provided via a 7812 and a 7912 (IC3 and IC4). Thanks to the printed circuit board of figure 2, building this power supply is very straightforward. It is important to mount transistors T1 and T2 on a heatsink. This should have a temperature rise of a maximum of 1.5°C/W and can be common to both transistors. Each of the tran- sistors must, of course, be fitted with a mica washer between it and the heat sink. Voltage regulators IC3 and IC4 must each be provided with a heatsink of 15°C/W. The noise suppression in this power supply proved very good when we tested the prototypes. At full load there was barely a ripple to be seen on the oscilloscope even when it was set to 10 mV per division. The stab- ility was also shown to be excellent. Switching from full load to no load gave a voltage difference of only a few millivolts. Two final remarks. As we have already said, the toroidal transformer can be replaced by three separate ones. In this case the minimum needed is one 9 V/5 A transformer, one 15 V/ 3.2 A. and one 15 V/0.4 A. If you wish to protect this supply against overvoltages and short- circuits this can easily be done by adding the 'microcomputer power supply protection’ circuit described 8.46 elektor india Aug/Sepi 1984 8.48 elektor india Aug/Sept 1984 for variable power supplies Most mains power supplies nowadays make use of a voltage regulator 1C. These circuits un- doubtedly simplify the design and result in a more compact unit. This applies to fixed output as well as to variable output devices. The latter have, however, one unfortunate characteristic: at high input voltage, low output voltage, and low load current, the dissipation in the regulator is maximum. This loss can be minimized with little additional effort and cost as can be seen from the circuit diagram. In this, the additional components are connected between the bridge rec- tifier and capacitor Cl. Immediately after switch-on, a zener voltage, Uz, develops across D5 which causes T1 to conduct. The current through T1 then flows to the gate of silicon-controlled rectifier (SCR) Thl. The SCR then conducts and the consequent current charges Cl via R4. It is only when Cl is charged that the regulator, IC1, pro- vides an output voltage, U 0 , whose level is preset with PI. What happens next becomes clearer from figure 2. Once Cl has been charged to its maximum voltage, Uqi, the current through Thl drops 2 u, below the holding value and the SCR blocks. Power to the load is now supplied only by Cl which naturally discharges. The rate of the discharge depends on the value of direct voltage (shown in dotted lines in figure 2) at the junction D2-D4. The time when they occur is depen- dent only on the value of the load. Because of the surges, diodes D1 . . . D4 should be 10 A types, e.g. SKN 26/04. The input voltage to the regulator 1C must not exceed 80 V: the output voltage may then be preset between 5 V and 50 V. Obviously, the trans- former, D1 . . . D4, and Cl must be rated to cope with these values. M SGS, Technical Note 145 prevents multiple entries This electronic key-set, although similar to its mechanical counterpart (switching, interlocking, reciprocal release, and so on) has one import- ant advantage over it: when two or more keys are pressed simul- taneously, all contacts remain open, which is by no means guaranteed with a mechanical set. The latter may actually allow several contacts _Uj> (V . ) I l b lmA) 3 6.45 5 2.8 9 ; 10.2 15 I 23.5 to close at the same time, which can have disastrous consequences. The circuit is simplicity itself: five standard or miniature push-button switches, two ICs, four resistors, six diodes, and five LEDs. The heart of the circuit is IC2, a BCD-to-decimal decoder type 4028. This guarantees that whatever the input information only one of its outputs is logic high (= switch closed). To start at the beginning: inputs A ... D of IC2 are at low logic level via resistors R3 . . . R6 and the 0 output is therefore logic 1. This is in- verted by N1 which causes D1 to klor india Aug/Sept 19848.49 light. If then, for instance, SI is pressed, input A goes high while the other inputs remain low. The decoder then switches from output 0 to out- put 1. The high level output at pin 14 is fed back (interlocked) to input A via diode Dll so that a stable situa- tion ensues and this continues even after SI has been released. The level on pin 14 is also inverted by N2 so that D2 lights. To activate output 3, switch S3 should be closed briefly. As input A is still high when C becomes high. switches up to 1750 VA Motorola ICs 3040 and 3041 enable the construction of a simple elec- tronic relay that in spite of its simplicity has some interesting characteristics. For instance, the ICs have an internal zero crossing detec- tor that saves quite a few external components. Moreover, the ICs have an isolation voltage of not less than 7.5 kV which enables the relay to be connected to any circuit operating from mains voltage. IC1 MOC3040 MOC 3041 the input information to the decoder is briefly 0101 (decimal 5), and IC2 therefore activates output 5 which has not been connected. As output 1 then goes low, the feedback to in- put A disappears and pin 10 becomes logic 0. Only input C is then high and IC2 switches on output 3. The interlock now lies across D9 to in- put C. Simultaneously, the level on pin 1 is inverted by N4 which causes D4 to light. The operation is similar when the other push buttons are Because only decimal outputs Q0 . . . Q4 and Q8 of IC2 are ter- minated, only the corresponding binary inputs can change the state of IC2. This is, however, only so if only one switch is pressed. When more are operated simultaneously, IC2 receives input information that ac- tivates non-terminated decimal out- puls. The terminated outputs are logic low as long as the switches are pressed, and all LEDs remain off. Switch SO activates output 0: the in- put information to IC2 is then 1001 (decimal 9) As the corresponding output is not terminated, there is no interlock to the input, so that inputs A ... D go low as soon as SO is released. H.J. Probst The ICs may be compared with a normal opto-coupler in which the usual phototransistor has been replaced by a phototriac (100 mA/400 V at 25°C). The advan- tage of this is that virtually all types of silicon-controlled rectifiers (SCRs) may be used in the circuit, which would not be possible if a phototran- sistor were used. The choice of type of SCR depends on what is to be switched by the relay. If the load is resistive, a TIC 226D/400 V will do fine. If, however, inductive loads have to be switched, a 600 V SCR, for instance, a type TIC226M, is needed. Bear in mind that the operating voltage of capacitor Cl must correspond to the type of SCR used. The value of R1 depends on the in- put voltage, Ujn. and ' s calculated R1 = 1000 (U in - 1.3)/l oc where Uj n is in volts, R1 is in ohms, and l oc , the current through the LED in the opto-coupler, is in mA. Assuming Ujn = 12 V, and l oc = 30 mA (as in the MOC 30401, the calculated value of R1 = 356 Q for which a standard 330 Q should be used. In the MOC 3041 l oc is only 15 mA, so that the practical value of R1 would be 680 Q. The maximum current that the elec- tronic relay can switch is about 8 A.^ Motorola application 8.50 ele his or her luck. If D6 does not ex- tinguish (because Q6 and Q7 are not high) the same player has another go. The player who has amassed the largest number of points at the end of the game is the winner or the loser, depending on how you play this game of chance. The circuit needs a power supply of 4.5 ... 9 V. The current consump- tion is dependent mainly on the bias resistors R3 . . . R7: with values shown, the current through each of ... uses LEDs We present a new game that might be described as a sort of electronic one-armed bandit. The good news is, however, that you don't have to put money into it; the bad news is that you cannot win money from it either. So, for the stakes, you'll have to come to an agreement with your playing partners. The circuit is based on a type 4024 seven-stage binary counter/divider. At the beginning of the game it is reset by spring-loaded press-to-make switch S2. The counter outputs, Q1 . . . Q7, are then logic low, and the LEDs, D1 . . . D5, are out. The output of NAND Schmitt trigger N2 is high and switches on relaxation oscillator N3. The oscillator signal is inverted in N4 and this turns on amplifier T6 so that LED D6 lights. The game is started by pressing another spring-loaded push-to-make switch, SI, when oscillator N1 is turned on and clocks the counter. As soon as SI is released, N1 ceases to oscillate and the counter stops at a random output combination. One or more of the LEDs D1 . . . D5 will be alight at that time and each of these is awarded a point or points as you may decide. Note the number of points before SI is pressed anew. When outputs Q6 and Q7 of the counter are both high, N3 ceases to oscillate and LED D6 goes out. This is the signal for the next player to try the LEDs is about 30 mA. The size and colour of LEDs D1 . . . D5 may be chosen to your own preference. H H.J. Walter audio interrupter The stereo is on, playing loud enough to do justice to your favourite rock record. The doorbell rings but in your ecstacy you don't realize it, even if you had been able to hear it. It doesn't take long before the would-be guest gets fed up and decides to protect his (or her!) gentle ears by going someplace quieter, like a heavy metal concert. That leaves you with two options: cut the mains lead of the stereo or fit a more effec- tive doorbell. The circuit here is a Combination of the two but we guarantee it is less destructive than the first. It cuts the volume of the stereo's output drastically when the bell is operated. Then to make sure the message is received the bell gives a number of tones of different frequencies switch- ing from one channel to the other. The operation of the circuit is quite simple. When the bell button is pressed there is a voltage across the bell which is rectified by diode D1 to provide a logic 'V. This causes a number of things to happen, the first of which is that electronic switches ESI and ES2 are closed. The outputs of both left and right channels are then greatly attenuated. At the same time the rhythm generator based on N1 starts working. This controls oscillators N3 and N4 and ensures that the signals they provide are fed to the left and right channel respect- eleklor India Aug/Sepl 1 984 8. 5 ' ively. The tone in the left channel is at about 800 Hz while that in the right channel has a frequency of about 400 Hz. The current drawn by this circuit is quite small, at less than 5 mA. The supply voltage could be provided by a battery or it could be taken from the bell transformer. Note that the signal earth is taken from the centre of voltage divider R4/R5. The volumes of the bell tones are set by means of presets PI and P2. Be very cautious about setting the volume too high as the rectangular waveforms of the bell tone signals contain many high-frequency har- monics that could cause tweeters to become terminally dead. The circuit should be connected in the audio system preferably between preamplifier and power amplifier. This point is often accessible by means of a preamplifier output/power amplifier input connection. A second possi- bility is to connect the inputs of the j circuit to the tape recorder outputs and the circuit outputs to the recorder inputs. Then set the ampli- I fier to tape playing. H loudspeakers and then provides a maximum of about 90 W. The pro- tection circuit of figure 2 must also be modified if a 4 Q load is used. The values of R24 and R28 are then 3k9, R26 and R28 are 220 Q. and D5, D6 and R30 are removed altogether. The power supply (not indicated) need only consist of the usual transformer, rectifier and smoothing capacitors. The electrolytic capacitors should be about 5,000 to 10,000 pF each. The rectified voltage for the 70 W/8 Q version should be ± 40 V with the load; with no load this cor- responds to about ± 47 V. At 4 Q these values are ± 34 and ± 40 V respectively. Don't underestimate the transformer requirements! It must be able to provide 1.4 A for the 2 70 W/8 Q version (mono) and 2.2 A for the 90 W/4 Q version. It is strongly recommended that a fuse be included in both positive and negative supply lines; 2 A for 8 Q or 3 A for 4 Q. Finally, a word about the cooling re- quirements of T6 and T9. In the 8 Q maximum; for stereo this is 1°C/W. These values become 2.5 and 0.5°C/W for the 4 Q version. M can also be used in the 2 m amateur band Many enthusiasts would be in- terested in listening to what goes on in the v.h.f. air band of 108 .. . 132 MHz were it not that receivers covering those frequencies are fairly expensive. Fortunately, air communications use amplitude modulation and if you therefore have a good short-wave receiver it is pretty straightforward to connect a suitable converter to it. And that's what this article is all about . . . The converter actually covers the fre- quency range of about 106 ... 150 MHz so that apart from the air band it covers a small part of the broadcasting band (up to 108 MHz, but that's mainly f.m.) and the 144 ... 146 MHz (that is, the 2 m) amateur band. The converter consists of a v.h.f. amplifier, a mixer, and an oscillator. After it has been amplified in T1, the input signal is applied to a MOSFET mixer where it is combined with the output of crystal oscillator T3. Three tuned circuits between the aerial in- put and mixer ensure good selectivity and good image rejection. The R1 - 22 k R2 = 220 Q R3 = 1 k R4 = 100 Q R5 = 27 k R6 = 1k5 C7 = 1p8 C8.C9 = 1 p CIO = 100 p C11 = 560 p C13.C14 = 68 p C15 = 3p3 LI =7 turns with tap at 3 turns'! Crystal 100. . ,120MHz, fifth overtone TA187AR difference-frequency output of the mixer is taken from its drain and fed to a 50 ohm output via a filter con- sisting of L7, L8. L5, C13, and C14. Tuning is carried out at the short- wave receiver between 6 and 30 MHz. The bandwidth of the tuned circuits is, of course, not sufficient to cover the whole range of about 106 . . . 150 MHz. At around 106 MHz the bandwidth is some 3 MHz; at 150 MHz it is about 12 MHz. Once you have chosen the band you want to listen in, tune the crystal and circuits L1-C1, L2-C2, and L3-C3 to the centre frequency of that band. The crystal frequency, f x , is equal to the difference between the input frequency, fj, and the output frequency, f 0 : f x = f j — f 0 . where f Q should be as high as permitted by the crystal frequency which should lie between 100 and 120 MHz. For in- stance, if you want to receive the 117 .. . 119 MHz band, f x could be 100 MHz (to keep f 0 as high as possible) and the short-wave receiver would be tuned between 17 and 19 MHz. If you select the 2 m amateur band, the short-wave receiver could be tuned between 28 and 30 MHz so that the crystal fre- quency would be 116 MHz (144 ... 146 - 16 = 28 ... 30 MHz). Air coils LI . . . L4 may be wound on a pencil and L6 on a ferrite bead, while L5, L7 and L8 are available ready made. Note that the printed circuit is double-sided so that the component side is an earth plane to which the various r.f. screens shown in figure 2 should be soldered. M elektor India Aug/Sept ’ 1 8.55 an inexpensive unidimensional design Just as it is very easy to implement an 'all or nothing' control (four pos- itions identified by two bits, or eight positions identified by three bits) it is very difficult to realise an inexpensive proportional control. When a reader suggested using an analog to digital converter, such as a 3162, to convert the voltage at the wiper of a joystick potentiometer into a single binary word we recognised the potential of the idea. The 1C used is more than just a normal analog to digital con- verter as it provides a multiplexed BCD output (4 bits: pins 2, 1, 15 and 16). The information needed for multiplexing is supplied to three pins: 4, 3 and 5, in descending order of significance. The software controlling the input port must be able to inter- pret this information and the main points which should be borne in mind when writing this software can be gleaned by studying the flowchart shown here. A '0' appears on port A bits 7, 6 and 5 in turn, indicating that the BCD code on bits 0 ... 3 (which can be from 0000 to 1001) corresponds to the most significant nibble (four bits), next significant nibble and least significant nibble respectively. The position of the wiper of preset P2, which is part of the voltage divider connected to joystick potentiometer PI, determines whether the output values range from 0 to 255 (which can be transmitted as a single hexadecimal byte - FFhex* or ,rom 0 to 999 (three BCD digits). The power for the circuit could be provided either by the microcomputer to which the interface is connected, or by a voltage regulator fed a voltage of 8 ... 15 V. The voltage reference applied to voltage divider R6/P2/R7 must, however, be very stable so it iot come from the circuit's .pp\ A small 9 V battery is therefore included and this is quite sufficient for the few microamps it will have to supply. Naturally enough, there will come a time when the battery voltage is too low but we have included a circuit to indicate this condition. When the voltage drops below 8 V T1 switches off. speedily followed by T2 and T3. The LED, D1, then extinguishes. When the input port has been pro- grammed all that remains is to calibrate the interface, as follows: ■ Move the wiper of PI completely towards ground and then trim P3 to get an output of zero (000 or 001). ■ Move the wiper of PI as far as possible towards P2 and then trim this latter preset to get the maximum value (either 254 or 255, or 998 or 999). The values set during calibration may be altered somewhat if necessary in order to prevent any possibility of the upper or lower limits being passed. This is done by selecting, for example, 005 as the lower limit and 250 or 994 as the upper limit. Then there is little need to worry about the stability of the battery voltage. M P. Palisson READERSHIP SURVEY In electronics, negative feedback keeps a system stable. Positive feedback makes it oscillate. In the same way, negative feedback from readers keeps the editorial staff from swinging too far out on a limb; positive feedback, quite honestly, is the spice of life. Readers' letters are useful. Talks with 'the trade' help. But there's nothing to beat a 'Readership Survey'! you get straight answers to straight questions. There are no firm rules. Tick any box that seems appropriate (preferably not more than one per question, unless otherwise specified): add a postage stamp if you have one handy (no apologies: we are avid stamp collectors!) and send it in (the sooner the better, but there's no closing date). We are looking forward to hearing from you! Your editor FIRST FOLD Contents of Elektor -I In which of the following 1 subjects are you really audio/hi-fi □ (t) electronic music □ (2) video □ (3) radio/h.f. □ (4) computers • construction □ (5) • interface, peripherals □ (6) • software □ (7) measuring equipment □ (8) domestic applications □ (9) vehicle applications □ ( 1 0) electronics for other □(11) hobby (model railways, photography, etc.) ' find characteristic of the articles in Elektor? (more then one choice is possil readability □ good background information often too long □ often too dry □ often too witty, too chatty □ often too short with insufficient details □ clear construction details practice oriented □ easy to understand □ often too theoretical □ O How many projects do you ^ build each year? A What is your experience with ^ component availability for Elektor circuits? no problem usually fairly easy often a problem hopeless C In your experience, do the J projects usually work first time 0(1) work only after some trouble shooting □ (2) rarely work at all □ (3) build as described in Elektor □ (11 make a few modifications □ (2) make a large number of modifications □ (3) "7 How much do you spend per * year on leisure electronics? less than Rs. 300 Rs. 300 to Rs. 750 Rs. 750 to Rs. 1500 Rs. 1 500 to Rs. 3000 more than Rs. 3000 O How much do you spend or authorize per year professionally on electronic components and/or equipment. (If you are a member of a group deciding on this, what roughly is your contribution?) nil □ (1) less than Rs. 7500 □ (2) Rs. 7500 to Rs. 15000 □ (3) Rs 15000 to Rs. 30000 □ (4) more than Rs. 30000 □ (5) Q What do you look tor in advertisements? (More than one choice is possible! components "CD ( 1 ) measuring equipment □ (2) computer hardware and software □ (3) other commercial equip- ment (audio, video, domestic, etc.) □ (4) books □ (5) tools □ (6) 8 58etekior mdia Aug/Sepi 1 984 Reading habits Readership profile ■IQ On average, how thoroughly do you read Elektor? 15 Is electronics your hobby profession? qualified technician □ (3) no formal qualifications □ (4) all articles □ ( 1 ) most articles □ (2) a few articles □ (3) I only leaf through □ (4) Could you estimate how many hours, on average, you spend on this during the first week or two? .... hours On average, how thoroughly do you look at the advertisements? I check them all □ ( 1 ) I look through most of them □ (2) I study a few □ (3) I only leaf through □ (4) I never look at them □ (5) lO How do you usually obtain 1 Elektor? on subscription 0(1) from a newsagent □ (2) from a specialist electronics shop □ (3) on loan from a friend, library □ (4) hobby □ (1) profession □ (2) both □ (3) 16 A - ; 17 or under Dll) 18-21 D(2)! 22-25 □ (3) 26-30 □ (4) 31-40 □ (5) i 41-50 □ (6) 51-60 □ (7) ; Over 60 □ (8) ! 17 Occupation: student D(1) self-employed □ (2) in teaching □ (3) employed □ (4) not employed □ (5) retired □ (6) 18 Education in electronics: corporate engineer □ (1) professionally qualified □ (2) Other education: Post Graduate p (5) Technical Degree p (6) University degree □ (7) H.S.C. □ (8) Diploma p (9) s:s.c P 00) 19 If you are a subscriber or read Elektor more or less regularly: recently □ (1) about a year □ (2) about two years (UK ed) P (3) three to live years ( " ) □ (4) more than five years( " ) □ (5) 20 In what country/ state/city do you live? Country State City 13 How often do you read the following magazines, and how do they score? 14 regular reader | good occasional reader average 1 nev r read it | poor (1) (2) (3) (4) INDIAN (1) (2) (3) (4) very poor (5) □ □ □ U ELEKTOR -INDIA □ □ □ □ □ A □ □ □ □ RADIO SERVICES □ □ □ □ □ B □ □ □ □ ELECTRONICS FOR YOU □ □ □ □ □ C □ □ □ □ POPULAR ELECTRONICS INUIA □ □ □ □ □ D □ □ □ □ INSTRUMENTS & ELECTRONICS □ □ □ □ □ E □ □ □ □ TELEVISION FOR YOU □ □ □ □ □ F □ □ □ □ ELECTRONICS TODAY □ □ □ □ □ □ □ □ □ DATA QUEST □ □ □ □ □ H □ □ □ L) COMPUTERAGE □ □ □ □ □ O Q □ PLUS □ □ □ □ □ J FOREIGN □ O □ □ ELECTRONICS & COMPUTING □ □ □ □ □ , □ □ □ □ ELECTRONICS TODAY INTERNATIONAL □ □ □ □ □ □ □ □ □ EVERYDAY ELECTRONICS □ □ □ □ □ M □ □ □ □ PRACTICAL ELECTRONICS □ □ □ □ □ N □ □ □ □ TELEVISION □ □ □ □ □ O □ □ □ WIRELESS WORLD □ □ □ □ □ P e kto* mdia Aua/Sepi 198 8 59 p-4 Do you own a computer? If s< 1 please indicate type. 6502 Acorn Atom □ Apple II □ BBC model B □ Commodore 64 VIC64/CMX64 □ VIC20/VC20 □ other commercial 6502 -based system — please specify □ (8) □ (9) □ (10) □ ( 11 ) □ (12) □ (13) Elektor Junior other home-made 6502-based system Z 80 Nascom TRS80I ZX80 ZX81 ZX Spectrum other commercial Z80-based system — please specify ! □ (i4) Elektor Z 80 □ (15) other home-made Z80-based system □ (16) Other commercial systems — please specify □ (17) Elektor SC/MP □ (18) Elektor TV Games Computer 0(19) other home-made systems — please specify '.'. '. '.'. '.'. '. '. '. '.'.D (20) If you have no computer, what is your interest, if any? I don't want to kn< about them mildly interested very interested □ ( 21 ) □ ( 22 ) □ (23) 22 What do you do with your computer? (More than one choice is possibl I use existing programs I program in: machine language □ assembly language □ BASIC □ PASCAL □ FORTH □ other language □ OO For what types of application do you use your computer? (More than one cl scientific/technical educational administrative games measurements/ system control miscellaneous □ ( 1 ) □ (2) □ (3) □ (4) □ (5) □ (6) OA Concerning articles on computers, do you think that during the past 12 months Elektor has published too many 0(1) just about the right number □ (2) too few □ (3) □ ( 1 ) informative articles (e.g. mP theory) projects on universally applicable peripherals □ (2) 8-bit nP systems for home □ (3) □ (4) construction 26 Given that the subjects under question 25 would replace many pages of electronics articles, which of them would you rather not see included in Elektor? (More than one choice is possible) informative articles (e.g. nP theory) nil) projects on universally applicable peripherals □ (2) □ (3) □ (4) Please use this space for additional comments. 8 60elektor India Aug/Sept ' I enables true 'on line' testing To be able to use a computer prop- erly, you need a number of relevant peripheral units. These units are often connected to the computer via a serial interface V24 or RS 232. If anything goes wrong, or is suspec- ted of having gone wrong, the ana- lyser described here may prove to be very useful. It is simply connected in series with the relevant line. The ad- ditional load on the line is so small that true 'on line' testing is possible. The circuit consists of two transistor drivers for two LEDs: red/red or red/green. With positive levels, in the range of about 4.5 .. . 5.5 V, T1 conducts and D1 lights. With negative levels, of the order of — 5.5 . . . —7.0 V, T2 conducts and D2 lights. If R4 is reduced to about 15 k, the circuit becomes active at about —3.5 ... -5.0 V. It should be noted here that in the RS232 negative levels correspond to logic 1, and positive levels to logic 01 The printed-circuit board can house four of these circuits, so that one board can monitor, for instance, signals RxD, TxD, RTS, and CTS which are, in the majority of cases, the most important. If you want to, or must, monitor more signals, all you have to do is to build more -o* boards. The wire bridges make it possible to hold certain signals at a fixed level during testing. If the boards are fitted into a case, it is, of course, possible to replace these bridges by switches mounted at the front of the case. Current consumption in each circuit amounts to 150 pA under no-Signal conditions, and to 27 mA with signal. In most cases, the supply can therefore be taken from the +5 V line in the computer. The wiring layout of the plug is shown for the computer side of the connecting cable. H i Aug/Sept 1984 8.61 8.62 etektof India Aug/Sepl cuit, more precisely a voltage- controlled oscillator (VCO) formed by N4, D3 . . . D6, T3, and associated components. If the base of T3 is provided with a sawtooth pulse train at a rate of a few hertz, a chirping noise is produced. The sawtooth signal is generated by gates N1 . . . N3. NAND gate N1 provides a square wave to oscillator N2, which functions only when the output of N1 is logic 1 (see figure 2). When the output of N2 is logic 1, sawtooth generator N3 produces a pulse train as shown in figure 2C. The ensuing noise cannot, how- ever. be heard because the VCO is blocked by the output signal of N1 at pin 8 of N4. As soon as the output of N2 becomes logic low, N3 ceases to oscillate and its output voltage tends to rise to the positive supply level. It is because N2 and N3 oscillate at different frequencies that a totally ar- bitrary sawtooth signal ensues. That signal is then pulse-frequency modulated by N4 to drive the piezo buzzer. The frequencies of the oscillators may be varied with P2 . . . P5 as appropriate, so that a range of bird sounds can be produced. When a two-terminal instead of a three-terminal electret microphone is used, the input circuit should be altered as shown in figure 1b. M sequently the Q output, goes high. The Q output (pin 13) is applied to one of the inputs of AND gate N1. The other input of N1 receives a high signal from IC4 (pin 3) and this lasts longer than the Q output of IC2. The output of N1 is then logic 1 which causes T2 to turn on and this results in triac Tril firing: bell 1 then .rings. When the button, SI, is pressed for a longer time, it is still closed at the trailing edge of the output of I Cl. Consequently, the D input of IC2 is low, and the ^ output is high. This output is applied to one of the in- puts of AND gate N2. The other in- put of N2 is in parallel with the sec- ond input of N1. From here on the circuit action is similar to that described above, but in this case T3 conducts to turn on triac Tri2 and this causes bell 2 to ring. The width of the pulses caused by the closing of SI is preset by PI, while the duration of the signal, and therefore of the ringing, is deter- mined by P2. The two triacs make it possible for a standard bell transformer to be used. with one push-button The present circuit is particularly useful where two families share one house, and where therefore two doorbells are a godsend, but where for one reason or another two push- buttons cannot be fitted. The only solution is then to operate the two bells with one push-button. When the button, SI, is pressed briefly, bell 1 sounds, and when it is pressed for a longer time, bell 2 will ring. Pressing the button triggers monostable multivibrator (MMV) IC1. The consequent logic 1 at the output (pin 3) causes T1 to conduct and this connects the clock input (pin 11) of D-type bistable IC2 to earth. This state does not last long, however, because as soon as the output of IC1 returns to logic 0, transistor T1 cuts off, and the clock input of IC2 goes high. When SI is pressed for an instant only, that is, it is open again at the trailing edge of the output of the MMV, the D input of IC2, and con- at 25 °C) the current limiting comes into play. This occurs at a current of about 2 A with the resistor values The purpose of the 10 pF capacitors at the outputs is to prevent the cir- cuit from oscillating and so CIO and C11 should be placed as close as is physically possible to the 1C. These two capacitors should be tantalum types as should C9. The transistors must be sufficiently cooled as they can get quite warm. Finally, as regards the function of diodes D5 and D6. These are in- + and -15 volts at 2 amps There is often a need for a high- power no-nonsense supply for a cir- cuit with a large number of op-amps and other (linear) ICs. A variable supply with current limiting and other special functions would in- crease the price and complexity un- , necessarily for most applications. What is needed in many cases is a straightforward design that does what is demanded: supply-power. No The LM 325 voltage regulator that forms the heart of this circuit is unusual in that it provides sym- metrical output voltages of + and -15 V. This is made possible by the inclusion of both a positive and a negative regulator in the same chip. In normal operation the B (boost), R cl < R current limit' and sense pins for both positive and negative regulators are connected together. The current is then internally limited to 100 mA, which is far from the high power we had in mind. All is not lost yet, however, as two external transistors can be driven by the B outputs and thus compensate for the chip's low current limiting. The actual current limiting is then decided by resistors R1 and R2 which protect both transistors and the 1C. As soon as the voltage across the current limiting resistors exceeds a certain value (0.7 V for the positive regulator and 0.6 V for the negative, the output can never become much greater than at the input. This could happen, for instance, if the capacitors remain charged for a time after the supply is switched off. M for control computers The use of a microprocessor system as control computer often makes it necessary for a large number of analog signals to be monitored. If a certain degree of accuracy is ex- pected from these signals, a number of suitable A/D converters will be needed. And that is not exactly The circuit presented here offers the facility of connecting eight analog signals to a CD 4051 multiplexer. The channel select inputs. A, B, and C, select the signal being monitored at the time: at binary 000, channel AN0; at binary 001, channel AN1; at binary 010, channel AN2; and so on. The selected analog signal (which must lie between 0 and +5 V) is passed straight through to the output (pin 3) of the CD 4051. The channel select inputs may be controlled by, for ex- ample, three lines from port B of a peripheral interface adapter (PIA). The actual A/D conversion is carried out by a CMOS 1C type ADC804. This transforms the analog signal into an eight-bit data word within 100 ps. The on-chip clock needs an external RC network (pins 4 and 19). Pin 9 must have a reference voltage which lies exactly in the centre of the measuring range. The conversion process comm ences at a leading edge on pin 3 (WR) and simultaneously the voltage on pin 5 8.64, r india Aug/Sept ' becomes 5 V at the trailing edge. After 100 ps pin 5 becomes 0 V and the 8-bit data word is then available at the port A inputs (D0 . . . D7). As there are 256 possible combinations, and the measuring range is five volts, each step is 19 mV 'wide'. The circuit may be used with prac- tically any microprocessor system which has a port available (PIA 6520-21; PIA 6820-21; VIA 6520; VIA 6522; Z80-PIO; 8255; etc.). Pro- gramming is dependent upon the re- quirement. For relatively slow oper- ations, such as heating control, alarm systems, weather stations, and similar, BASIC may be used with PEEK and POKE instructions. With on-line controls, and maybe even with model railway control, it will normally be necessary to use a machine code program which is much faster. M ' india Aug/Sepi 19848.65 and the supply is taken from the cir- cuit under test. The tester indicates clearly by means of a pair of LEDs (D5 and D6) whether the logic level tested is '0' or 'V. The input is fed into the non-inverting input of a schmitt trigger comparator (IC1). The inverting input of this 3130 is fed from one of two fixed reference levels depending on whether CMOS or TTL is selected. The hysteresis in- herent in a schmitt trigger ensures that a 'dead zone' must first be passed before the circuit reacts to the change in input level. This means that when a change of level is indicated there is absolutely no doubt that the transition has taken for testing digital circuits This design is a somewhat unusual test aid for digital circuits. Most such circuits are usually tested in a static logic levels at various points by mak- ing d.c. measurements. This is often not good enough, however, consider- ing the various clock signals, resets, and trigger pulses that make a circuit 'tick'. What this really demands is a tester that can also detect single As the circuit diagram shows, either CMOS or TTL can be catered for place. The exact switching points measured in our prototype are given in the table. Pulses are detected by the section consisting of N1 . . . N4, T1, and the associated components. The mono- flop consisting of N1 and N2 (and C3 and RIO, of course) reacts to a high to low transition at the input and lights LED Dll for a certain length of time (long enough for it to be no- ticed). Pulse trains also cause Dll to light but, in this case, continuously as the monoflop time bridges the pulse interval. At the same time LEDs D5 and D6 also light and if both are the same colour a com- parison between the brightness of the two will give a rough idea of the duty cycle of the pulse train. Fre- quencies of up to about 400 kHz can be processed by the circuit. The cur- rent consumption of the tester is about 50 mA. Building this circuit is simplified by using the printed circuit board we have developed. The design is given here and also in the service pages in the centre of this issue. It is not available through the EPS service. This is not, however, a major drawback to making this simple but useful logic tester. M sistor T1 when the output of N2 is logic 1. It will be apparent that the triac will therefore only be switched on when the mains power supply is at zero potential — an ideal situation! The light La will thus be switched on until such time that the count cycle of IC1 causes its Q14 output to change to logic 1. This will then halt the clock oscillator via diode D2 and maintain the high logic level at the Q14 output. At this time, gate N2 will find a T at switches itself on (and off) This circuit forms a switch that can be triggered on (or off) at the onset of darkness and remain active for a period that is variable from thirty minutes to five hours. The main design consideration was that of keeping power consumption to a minimum while retaining a fair degree of versatility. The primary purpose was to provide an economical porch light but the circuit can obviously be used for any short- term switching application that is referenced to ambient light levels. In spite of its simplicity the circuit is relatively sophisticated. The basis of the circuit is a 14-stage ripple counter with an internal clock oscillator contained in the 4060 (IC1) which does most of the work. The frequency of the oscillator is deter- mined by C3, R6, and potentiometer PI which enables the frequency to be adjusted. The clock is started when a logic 0 reaches the reset in- put of IC1. This is achieved by the light-dependent resistor R14 and the components associated with gate N1. When the ambient light level drops, the resistance of the LDR in- creases and causes the output of N1 to revert to logic 0. The point at • which this occurs can be determined by the setting of P2. It is worth noting that, although not desirable for this particular application, if the ambient light level increases again for any reason for a period of not less than 10 ... 20 s, the counter will be reset and the clock oscillator will stop. In the normal course of events, the output of gate N2 will be at logic 1 while the 4060 continues to count up and transistor T1 will be controlled by the output of gate N4. The two gates N3 and N4 together form a simple, but effective, mains zero crossing point detector. This results in the output of N4 providing a short pulse at each zero crossing point of the mains cycle. It is this pulse that is used to trigger the triac via tran- its pin 5 input and a '0' at pin 6. The output of N2 will therefore return to a low level and switch the light off. Power levels up to 100 W can be switched by a TIC 206D unaided. However, if higher power levels (up to 500 W) are contemplated, the triac must be provided with a heat- sink, for instance, an SK13. The switched time period is ad- justable by PI. H r india Aug/Sept 19848.67 humane solution to an old problem This mousetrap is not intended to kill a mouse with the aid of electronics, but rather to imprison it in a gentle way. Afterwards, it may be given its freedom in a suitable area. The principle of the trap is the age- old trap-door up-dated by being operated electronically. Figure 1 shows the construction of the device. In this, a small wooden box is divided into two chambers: thff larger one is fitted with the trap- door, while the smaller one contains the electronic circuit and power supply. In spite of the modern construction, some bait is still required, and as of old a piece of bacon or cheese is best for this. On its way to the bait the little rodent breaks the beam of a light barrier and this causes an electro-magnet to release the trap- door which then blocks the way out. The distance between the trap-door and the light barrier must be longer than the length of a mouse other- wise the animal's tail will be caught. How the trap-door is released is shown in figure 1. The light barrier consists of a light-emitting diode, LED D1, and a light-dependent re- sistor, LDR. When the LDR is illumi- nated by the LED, it is low-ohmic, and the latch N1/N2 is not set. When the beam of light is broken, the latch changes state, and the con- sequent logic 1 at the output of N2 triggers monostable multivibrator (MMVI N3/N4 which then imparts a pulse to driver T1/T2. The width of this pulse is about one second which is sufficient to actuate the electro- magnet which then releases the trap- door. Latch N1/N2 ensures that once the circuit is triggered, it does not re- act to further breaks in the light bar- rier caused by the mouse moving in- side its prison. The electro-magnet should be home- made, preferably by using the coil of a spare relay or doorbell. Also suit- able are electro-magnets as used in cassette and tape recorders. The required power may be provided by any transformer which has a sec- ondary voltage of 8 . . . 12 V at a current of not less than 100 mA: it may be a bell transformer or of the type used in a battery eliminator. If the device is used only occasionally, a PP3 battery may be adequate. Be- fore placing the mousetrap in posit- ion, make sure that it operates satis- factorily. The trap is reset by a spring-loaded push-button, SI. The state of readiness is indicated by the lighting of LED D2. M 8.68 etektof india Aug/Sept ' react at the speed ot light The human eye has a certain 'built- in' delay. This fact is used for films, TV sets, and fluorescent lights, as above a certain flash frequency the eye does not notice any lask of con- tinuity. It has now come to light that the highest frequency flashing a per- son can detect is adversely affected by tiredness and alcohol consump- tion. A very small circuit is all that is needed to determine exactly what this frequency is at any time of the day or night. As the diagram shows, the circuit is very simple. It is based on an old favourite, the 555 timer, which is connected here as an astable multivibrator. Its output is connected to a LED that flashes at a certain frequency. This frequency can be varied between 20 and 50 Hz using potentiometer PI. The highest fre- quency that most people can detect second, but one test we conducted on a Monday morning produced a startling number of blank stares ac- companied by the question 'What LED??'. Given the nature of the circuit, it is not surprising that the current con- sumption is only about 25 mA so a 9 V battery is all the power that is needed. M dia Aug/Sept 19848.69 | motion study by flash-light WARNING! The circuit of the stroboscope is connected directly to the mains and experiments on the opened unit are therefore highly dangerous. Even after the unit has been disconnected from the mains, some capacitors may still give you a lethal shock! Preset PI must have a nylon spin- dle and be fitted so that none of its metal parts can be touched: ignoring this may be lethal. The heart of the unit is, of course, the gas-discharge tube, which is U- shaped and filled with xenon (Xe - one of the inert gases). The tube is fitted with an anode and cathode at either end, and an ignition grid. Diodes D1 and D2, together with capacitors Cl and C2, form a voltage doubler which raises the direct voltage to about 600 V. This voltage is applied to the anode and cathode of the tube. Normally, xenon (and other gases) is a poor conductor of electricity but the electric field resulting from the 600 V potential across the anode and cathode causes ionization of molecules and atoms in the im- obtained from ignition transformer Trl. To cause a high potential across the secondary, the current through the primary should be interrupted very rapidly and this is done by a silicon-controlled rectifier, Thl. Capacitor C3 charges because the voltage across C2 is 300 V and the primary of Trl is low-ohmic. As soon as the threshold of the two trigger diodes, D3 and D4, is reached, the SCR is fired. Capacitor C3 then discharges rapidly via the primary winding of Trl which induces a very high voltage in the secondary and this in turn causes the xenon tube to fire. The setting of preset PI determines the charging rate of C3 and therefore the firing rate of the discharge tube. Resistor R1 is connected in the neutral line to serve as a current mediate vicinity of these electrodes. The gaseous ions are attracted to thr charged electrodes and a small preconducting current flows. A grid potential of 5 ... 10 kV is required to fire the tube, which means caus- ing the gas to break down so that a large current flows across the tube. The relatively high grid potential is limiter, because when the discharge tube is firing it is a virtual short- circuit; without R1 the fuse FI would blow instantly. The discharge tube, which should be of the 60 W/s type, is normally pro- vided complete with ignition trans- former. The anode is usually in- dicated by a red dot. M space saving and cost effective Warning! This circuit needs to be constructed and wired with the greatest care as the full mains voltage is present at several points. The pulsating direct voltage provided by rectifier D1 . . . D4 has a peak value of 310 V. This voltage is ap- plied to the drain of power MOSFET T1 via limiting resistor R9. A control circuit ensures that the MOSFET only conducts during the short times just before and after the mains voltage goes through zero. During these times the momentary value of the pulsating direct voltage does not exceed 5 V. In the same short times smoothing capacitor C2 is charged: during the remainder of the time it provides the output cur- rent. Consequently, this capacitor has a very high value: 10000 p. The load-current pulses have a peak value, if only for a brief moment, of the order of 4 A! The stability of the output voltage is essentially dependent upon the load. The output current may be 110 mA maximum. The supply for the control circuit is provided by resistor R2, capacitor Cl, and diodes D5 and D6. The control circuit is a window com- parator constructed from three opamps. Correct calibration of the control circuit is therefore very important. Before the mains is ap- plied for the first time, set PI to the centre of its travel and turn P2 so that its wiper is at earth potential. Then connect the mains and check 8 . 7 0 eleklof india Aug/Sept 1 984 the operating voltage of the circuit. Next connect a voltmeter (10 V dc range) at the output and adjust P2 until the meter just begins to deflect. Finally, adjust PI for a meter reading of 4.8 ... 5 V. Applications of the circuit are restricted. It is evident that it cannot be used with equipment which should be electrically isolated from the mains. It is equally unsuitable for use with equipment that is allergic to mains spikes and noise. It is, however, eminently suitable where there is no space for a mains transformer. The unit should only be used for powering equipment that is contained well-insulated in a plastic case. Any equipment powered by the present unit should not be connected to other equipment by cable. Such connections, if necessary, should be by opto-coupler only. Heat dissipation in T1 and R9 amounts to only about 3 W so that even if the circuit is fitted in a small case there should be no heat prob- lems. During assembly the usual precautions relevant to mains operated circuits should be observed scrupulously. H Siemens application note The design for the circuit here was created when problems arose with the matching of some output stages with a preamp. In effect, the circuit is simply what the title suggests, a buffer between an audio preamplifier and an output stage. It does have the added facility however, of being able to drive more than one output amplifier simultaneously. The preamp load is standard at 100 pF in parallel with 47 k. To be fully versatile it was considered that the opamp used must be capable of driving similar loads at a level of 10 V without a problem. The LF 356 shown here can manage this. The amplification factor is adjustable between 1 and 5 with the aid of the preset potentiometer in the feedback loop of the opamps. These also serve to balance the output levels of the two buffer stages. If required, the balancing can be achieved very easily by means of a 50 Hz signal source and an ordinary multimeter. The 50 Hz is applied to both inputs of the buffer circuit and one preset is adjusted to provide the required gain factor. The multimeter, switched to a suitable ac range, is then placed be- tween the two outputs. The second preset is now adjusted to produce a zero reading on the meter. M 1 Aug/Sept 1984 8.71 9 ... 30 V, when wire bridge 'A' on the printed-circuit should be fitted. When a 5 ... 15 V symmetrical supply is used, wire bridge 'B' should be fitted, R2 should be replaced by a wire bridge, and R1 and C3 are omitted. NOTE: a ready-etched printed-circuit board for this project is not available: it can, however, be made from the layout (ref. 84446) given in the pc board pages in this issue. M depends on channel separation Weak VHF/FM signals, particularly stereo broadcasts, are normally received against a background of noise. When the receiver is then switched over to mono, much of the noise disappears, but so, unfor- tunately, does the stereo effect. The present circuit reduces the noise drastically, but does not eliminate the stereo effect. A potentiometer allows selection of the best compromise between noise and channel separa- tion. The circuit is simply inserted between the tuner and the amplifier. Inputs and outputs are isolated from direct voltages by coupling capacitors Cl, C2, C5, and C6. The input impedance is of the order of 100 k because of resistors R3 and R4 which also provide a direct voltage to operational amplifiers IC1 and IC2. This voltage is half the supply voltage because of voltage divider R1/R2 which is decoupled by C3. The opamps function as impedance inverters with unity gain. Their out- puts are taken to a stereo poten- tiometer, the minimum value of which is limited to 3k3 by resistors R5 and R6. The output terminals of the two sections of the poten- tiometer are shunted by capacitor C4 when switch SI is closed. This capacitor causes frequency- dependent cross-talk between the channels and the consequent decrease in channel separation pro- vides a reduction in noise. The capacitor therefore acts as a low- pass filter. The frequency response of the composite signal is not affected by the action of 04: the difference signal (the stereo component) is, however, attenuated at a rate of 6 dB/octave. The cut-off frequency of the low-pass filter may be set be- tween 1.3 kHz and 5.1 kHz with PI. The suppressor is switched on and off by SI: with this switch open, the input signal appears unchanged at the output. The output impedance of the circuit depends on the setting of PI: its maximum value is about 14 k. Current consumption amounts to about 10 mA. As is shown in the cir- cuit, the power supply may be sym- metric or non-symmetric: in the lat- ter case the supply voltage may be 8.72 elektor mdia Aug/Sept ' 67 Table lor short-wave receivers Many short-wave listeners often wish their receiver could be extended above 30 MHz. The converter described here makes it possible to add either the 30 ... 60 MHz or the 60 . . . 90 MHz band to your existing A converter transforms a certain range of radio frequencies that can- not be processed by a given receiver into one that can. Its output is nor- mally connected to the aerial input of the receiver. The present converter is suitable for the reception of either the 60 ... 90 MHz or the 30 ... 60 MHz band. It consists of an r.f. pre-amplifier, an oscillator, and a mixer. The signal picked up by the aerial is applied to a preselector con- sisting of tuned circuit L1-CI-D1, MOSFET amplifier T1, and tuned cir- cuit L2-C2-D2. Diodes D1 and D2 are varactors which enable fine tuning of the two resonant circuits. A varactor is a semiconductor diode operated with reverse bias so that it behaves as a voltage-dependent capacitor. The control voltage is derived from a multiturn potentiometer, PI, which therefore enables coarse tuning of the two circuits. The amplified signal from the preselector is applied to mixer T2 together with a 60 MHz or 30 MHz signal generated by crystal oscillator T3. Filter L5-L6-L7-C8-C9 at the out- put of the mixer only passes the dif- ference between the two frequencies, that is, about 0.1 . . . 30 MHz, which is the frequency range of most short- wave receivers. Inductors LI and L2 should be wound on a pencil according to table 1, while L4 consists of 4 + 1 turns on a ferrite bead. Wire to be used is enamelled copper, SWG21 for LI and L2B and SWG23 for L2A and L4. The remaining inductors, L3, L5, L6, and L7 are standard miniature chokes. Calibration of the converter is simple. If you have a frequency counter, ad- just trimmers C3 and C4 until the oscillator operates at exactly the crystal frequency. If you have no fre- quency counter, just set C3 and C4 at about the centre of their travel: this is usually precise enough. Next, the circuits in the preselector should be tuned to the wanted fre- quency range. Using PI, seek a signal of about 30 MHz or 60 MHz, depending on which range you have chosen, and carefully compress or elongate the turns of LI and L2 until the received signal is strongest. Then, again using PI, find a signal at around 60 MHz or 90 MHz and ad- just trimmers Cl and C2 for maxi- mum signal strength. The converter needs two supply voltages: a stabilized one of 12 V for most of the circuits, and a 24 V one for the tuning control. Current con- sumption amounts to about 40 mA from the 12 V supply, and only around 1 mA from the 24 V line. NOTE: A preselector improves the sensitivity and the selectivity of a radio receiver; it usually is a tuned radio frequency (r.f.) amplifier that amplifies the incoming signal before amplification and demodulation. H 1 8.73 o longer in the 1C sockets: | The VDU card published in Elektor | number 6, October 1983, sometimes 3ts noise appear on the screen, such s when a program is being listed. This fault can easily be remedied by means of a few gates in the VDU card that are unused in the Junior Computer/VDU card combination. The trick of the circuit consists of stopping the processor when it at- tempts to write to the video RAM during the display enable time. Only the 65C02 can be stopped during writing so this circuit operates ex- clusively with Junior Computers equipped with the CMOS Processor. I This procedure causes a slight delay they are r IC2 pins 7, 9, 11 and 13 IC4 pins 1, 8, 9 and 10 IC7 pins 1, 8, 9, 10, 11, 12, and 13 IC8 pin 8 IC17 pin 1. These pins are then connected together as indicated by the heavy lines in the circuit diagram. Pin 1 of IC17 simply remains open, while pin 2 of IC7 is already connected to ground. Note that pin 1 of IC14 and pin 12 of IC17 must remain pushed into their respective sockets even after wires are soldered onto them, program but this is Another possible 'extra' for the VDU >ractice. card is to show a frame on the the modification monitor within which all the video n the VDU a fresh look at the 723 voltage regulator In its standard application, the 723 voltage regulator provides an output of 2 ... 37 volts but in many cases it is necessary to be able to go down to 0 V. To do so, an auxiliary negative voltage is required: in the present circuit this is provided by an LM337 negative regulator (IC2). It is not sufficient just to connect an additional circuit onto the same transformer as the positive supply: to get a negative voltage, there MUST be a load on the positive supply. This is provided by R5/T2, which en- sures that a current flows at all times when the mains is switched on. The circuit provides adjustable cur- rent limiting which is effected by ap- plying a voltage of 0.6 V between pin 2 (CL = current limit) and pin 3 (CS = current sense). This voltage is the sum of the drops across R8 (proportional to the output current, l 0 ) and across P3. The latter voltage is the product of the resistance of P3 and the current through T1. Further stabilization of the base of T1 is pro- vided by T2. In spite of this double stabilization there remains a small ripple (0.3 per cent) on the current into CL. Voltage stabilization is provided by IC1: hum and noise are less than 1 mV at an output of 15 V at 150 mA. The output voltage increases linearly with the resistance of P2. Maximum output level can be preset with PI. The negative supply has a longer time constant than the positive sec- tion so that when the mains is switched off, it remains active slightly longer. If this were not ar- ranged, the output might momen- tarily rise (which could damage the equipment being powered) owing to the inability of the 723 to go down to zero without an auxiliary voltage. The 2N3055, provided it is mounted on a suitable heat sink (2 °C/W), can dissipate 30 ... 40 watts. At a transformer voltage of 22 V, this means that well in excess of 1 A can be handled. The choice of transformer is fairly critical, because strictly speaking 24 V is already slightly too high for the 723 which tolerates just about 36 V. It is therefore better to use the L146, an improved version of the 723 which can handle up to 80 V. Note, however, that even then the transformer secondary voltage should not be much higher (a few volts) because otherwise the maximum rated voltages of the electrolytic capacitors and transistors will be exceeded. Some further points worth bearing in mind: ■ The transformer secondary voltage should be about equal to the required maximum output voltage, at least, that is, if this lies 8.74, l Aug/Sept IS above 20 V. ■ Always ensure that the current rating of the transformer is at least 1 .4 times the output current. ■ The output voltage is equal to P2U ne g/R4 volts; U ne g should be set at about —5 V with PI. By ad- justing PI (and therefore U neg ) slightly, the maximum output voltage can be set precisely to 22 V. If the required maximum output voltage is quite different from this value, R4 has to be adapted so that U neg still remains about —5 V. ■ The maximum output current is determined by R8 and is equal to 0.6/0.47 = 1.28 A. ■ Do not allow the 3055 dissipating more than 40 W continuously! Finally: the earth return is inten- tionally shown as three parallel lines to give a clear point of reference where the voltage or current, in the final instance, is constant. Owing to the unavoidable voltage drops across the earth returns, regulation will always be inferior when the returns are not kept separate. M W. Vogt tually used is a normal mains transformer. The low-voltage winding is connected in series with the cur- rent that is to be defined. The '220 V' winding is now free to have the LED(s) or other measuring equip- ment connected to it. When choos- j r ,'q a transformer it is important n mind the maximum current that is v ‘ 5x P ected in ,he secondary winding a.' nd ,be maximum permitted LED current Consider this example: the current to t> deleted is 0.6 A so the low-voltage' windin 9 must be etektor indu. ' Au9/Sepn 984 ® ^ 5 able to handle at least this. Assum- ing that a current of 30 mA is the maximum for the measuring circuit we choose a 220 V/12 V transformer to give us approximately the right ratio (600/30). The voltage loss across the winding in the primary circuit is relatively small. In the ideal case the resistance and leakage of the transformer are small enough to be ignored so the voltage loss is only the LEO voltage divided by the transformer ratio. The transformer secondary must always be connected to a load, for both positive and negative half- cycles. This is the reason for adding a second LED or a diode in the cir- cuit. Without this load the primary winding would act as a normal coil which would result in a higher voltage drop across the primary and a higher voltage at the secondary. This diode or LED also protects the LED against high reverse voltages. H grabs your attention One of the great things about com- puters is that no matter what you tell them to do they never complain and always stay quiet. Sometimes, however, it is helpful if a computer can make some noise to catch your attention. In the KB-9 BASIC or Junior BASIC the ASCII character 07 is an 'end of line' indicator, and represents 'BELL' (control + G on the keyboard). The extended Junior does not use this signal, but it can if the circuit shown here is built. The circuit diagram shows just how simple the hardware for this 'bell' is. When the ASCII character 07 (0000 0111) appears on lines B0 . . . B6 (the data output lines of the UART on the Eiekterminal) and the DAV (data available, UART pin 19) line is high (indicating that the whole character has been re- ceived) the output of NAND gate N5 goe s 'low'. Th is signal then goes to the TRIGGER input of monostable multivibrator IC3. The external timing components connected to pins 1, 2 and 3 of this 4047B determine the width of its output pulse, and ad- justing preset PI varies this width. The Q output then goes high for a certain length of time and during this time T1 drives the buzzer. The cur- rent consumption of the circuit is no more than 20 mA. The printed circuit board for this cir- cuit is quite small, as could be ex- pected. This is, of course, an advan- tage when you are trying to squeeze it into what little space there is available in the case of most Elekter- minals. Construction is simply a mat- ter of fitting the components onto the board and the only point of note IC1 = 4049 IC2 = 4068 IC3 = 4047B concerns diode D2. This diode may be replaced by a wire bridge if the buzzer volume is too low. The 'software for this bell is just as simple as the hardware. In the KB-9 BASIC or Junior BASIC all it involves PRINT CHR $(7). W. Schaaij 8.76 eTektor india Aug/S*- meets the following specifications: — high clock signal minimum level: U+ - 600 mV — low clock signal maximum level: 0.45 V — rise time maximum: 10 ns at <35 pF load capacitance The circuit can be used without any problems for frequencies up to 10 MHz, in which case XI is a 20 MHz crystal. M Tekelec Airtronic for frequencies up to 10 MHz When using fast CPUs it can still be a problem to generate a good clock signal. By 'good' we mean that it must have clearly defined high and low levels and it must be very sym- metrical. What we are thinking about, in particular, is 8 MHz CPUs but this circuit can operate with crystals of up to 20 MHz and, more importantly, it provides an excellent clock signal. The actual oscillator is based on a pair of inverters (N1 and N2) and it oscillates at twice the CPU fre- quency. Its signal is buffered by N3 and then the frequency ishalved by D-type flip-flop FF1. The Q signal from this flip-flop is buffered and in- verted (by N4 and N5) and is then available for functions other than the CPU clock. The Q output, on the other hand, supplies the signal for the driver circuit for MOS levels that is based on T1 and T2. The clock signal finally output from this section CLK Z80 Z8000 ug/Sept 1 984 8.77 protects your car battery Who has not at one time or another forgotten to switch off his car's lights on a murky morning? That is not much of a problem when col- leagues or passers-by are kind enough to draw your attention to it. But if there are no such friendly souls about, you may at the end of the day find that your battery is as flat as a pancake. Some modern cars have a factory-fitted warning unit, others have their wiring arranged such that when the ignition is switched off only parking lights can be left on. The majority of cars, however, are not protected against such an oversight and it is for those that we have developed the present warning circuit. This has an advan- tage over other similar circuits you can switch on your headlights when the ignition is switched off. The circuit is based on two astable multivibrators (AMVsl of which the first is formed by NAND gates N1/N2 and associated components. It operates as a clock with a fre- quency of about 20 Hz. The second AMV, based on N3/N4, operates as a tone generator at a frequency of about 3300 Hz. The clock and tone generator are controlled by a transistor-relay logic terminal, and M is the earth. The broken lines in the circuit indicate parts already fitted in the car: L is the lights switch, Z is the ignition switch with underneath it the ignit- ion coil and contact breaker. When with the ignition switched on, the lights are turned on, transistor T1 conducts. At the same time, relay Re is actuated and short-circuits the collector-emitter junction of T1. Although this connects the +12 V line to pin 14 of IC1, the AMV does not yet operate because pin 7 is not connected to earth. There is there- fore no alarm tone from the buzzer. If now the ignition is switched off, the relay remains actuated and the that circuit which obtains its data from the car's electrical system. The warning unit is connected to the car's electrical system at terminals 15. 58 (56), and M in the circuit diagram. These are DIN designations used in the majority of cars; if yours is an exception, 15 is the ignition coil terminal, 56 is the centre contact of the dip switch, 58 is the parking light +12 V line connected to pin 14 of IC1. Pin 7 of the 1C is then con- nected to earth by the contact breaker or other load via diode D4. Both generators now function and the buzzer emits a warning note. If the lights are then switched off, the relay is no longer actuated, and the +12 V line is removed from pin 14 of I Cl, which stops the generators. If the lights are required to remain on, the lights switch can simply be turned on: the alarm will then not be actuated. *• H. Braubach with high stability The Wien bridge oscillator is a com- monly used circuit, which is not sur- prising considering that it has low distortion and its resonant frequency can quite easily be made adjustable. a pair of resistors (each = R) and a pair of capacitors (each = C) and is defined by the formula f = 1/2nRC. In the circuit shown here R consists of R1 + Pla (or R2 + Plb) and C is either a, C2 or C3 (or, C4, C5 or C6>. The oscillator proper consists of these components together with IC1, I IC2 and their associated com- ponents. Part of the output signal from IC2 is fed to the regulating attenuator con- sisting of IC3 and T1. This FET, which is used here as a variable resistor, is part of the feedback loop of IC2. The gain of this op-amp is thus made voltage dependent and can be changed by altering the control voltage of T1 using P2. This potentiometer must be set so that the circuit oscillates stably. The range of the oscillator, with the com- ponent values shown, is from about 20 Hz up to 22.5 kHz and distortion is no more than approximately 2% . M B.G. Lindsay 8 . 78 . I Aug/Sepl 1984 level, U(j, is programmed by means of resistors R2 and R5. The value of R2 is given by R2 = [KttUd - 1.31/51 kQ The integral regulation loop turns off the chip when the feedback voltage drops below 1 .3 V. Capacitor Cl is the frequency deter- mining component for the internal sawtooth oscillator in the regulation The output voltage, U Q , is given by U 0 = I1.3IR3 + R41/R41 volts for battery-operated equipment The low-power switching regulator type 4193, which is housed in an 8-pin miniature DIL package, is designed specifically for battery operated equipment. A regulated power supply can be constructed for such equipment with just eleven components: five resistors, two capacitors, one diode, a choke, a 4193, and a 2.4. . .9.0 V battery. The output of this supply will remain ■near-constant at 9 V until the battery has decayed to a terminal voltage of 2.4 V. A practical circuit is shown in The 4193 has an internal reference circuit of which the control current, l c , is set externally by resistor R1 connected between the battery and the l c pin (6). This current can vary from 0.5 p A to 100 pA without af- fecting the operation of the chip. The value of R1 is given by R1 = I(U b - 1.3)/l c ] kQ where U b is the battery voltage in volts and l c is in mA. In addition to setting bias currents throughout the chip, the reference voltage is used for the low-battery detector circuit, and to set the threshold for the input of anon-chip regulation loop for comparison with a feedback voltage, Uf (pin 7). The low-battery indicator voltage and is plotted (figure 2) against the output current, l 0 , for various values of R4 (with R3 = 82 kQ), and an input voltage of U 0 /2. H Raytheon application ROM, a ROM with 512 musical notes, tone generator, rhythm generator, timbre generator, modulator, run-off control, and pre-amplifier. Apart from the 1C, the circuit com- prises an a.c. operated power supply with voltage regulator, a push-pull amplifier for driving the loudspeaker, and a number of associated components. Resistors R1, R2, potentiometer P2, and capacitor C2 are the frequency- determining elements for the on-chip oscillator. Preset P2 is for adjusting the run-off speed, that is, the speed at which the tune is played. Resistor R7 and capacitor C4 ensure optimum performance of the internal modulator. Resistor R3, preset PI, and capacitor C3 form a volume control which controls the on-chip pre-amplifier. The circuit is operated by SI ... S3 and R4 . . . R6. Switch S2 is the normal bell-push. If you want to pre- program a given melody, an addi- tional push-button may be connected in parallel with S2. With SI closed, all melodies stored in the ROM will be sounded in se- quence; when it is open, only the one selected by S3 will be played. A particular melody is chosen by closing SI and pressing S2 con- tinuously, while S3 is pressed repeatedly until the wanted melody has been reached. Until now, four ICs in the series have become available and these differ only in the melodies stored. The UM 3481 contains eight Christmas carols and the UM 3484 the sounds of Big Ben striking one to twelve in ascending order. The UM 3482 has twelve tunes, among which "Fr6re Jacques, frere Jacques", "Happy Birthday to you", and "Cradle Song", while the UM 3483 contains melodies like "The Last Rose of Summer", "The Lorelei", and "Wedding March". At the time of going to press we understand tnat there may be dif- ficulties in obtaining the UM 3481 .3484 in some areas but we hope this problem will be re- solved soon. H change the values of R20 and R21 . In other situations the circuit based on N1 . . . N4 can be used. A clock pulse is generated every time SI is closed or S2 is opened, and, to alleviate any problems, the effects of contact bounce are suppressed by N2, N3, C3, and R25. M with LED indication An event counter, as could be ex- pected, counts events, or, to be more precise, it counts the occur- rences of a particular event. The counter here may seem a bit limited, as it can only go to 99, but in fact it can be expanded almost infinitely. The read-out consists of two lines of LEDs, one for units and the other for decades. Only one LED per line (at most) will light at a time so the cur- rent consumption is quite low, cer- tainly when compared to a set-up with 7-segment displays. The actual counter consists of two 4017 decade counters. When the reset button is pressed both 00 out- puts go high. Every clock pulse arriv- ing at pin 14 of IC1 makes the next output of the 1C go high. At every tenth clock pulse the CO output goes high and clocks IC2 and at the same time IC1 is reset to 0. After 99 pulses both IC1 and IC2 reset to zero and the sequence starts again. In principle the carry output of IC2 can be used to extend the circuit infinitely. The outputs of a 4017 cannot drive LEDs directly so it is necessary to add a simple buffer stage, consisting of a transistor and a resistor, to each output. A single common resistor (820 Q at 15 V supply) per line is all that is needed as each 1C drives only one LED at a time. All that remains now is to consider the clock or counter pulses. Sometimes these can be taken directly from another circuit and if this is the case check that the power supply is suitable and, if necessary. From now on, R-S flip-f lop N1/N2 is set on reset. The (new) BUSACK signal goes low via N5 and all bus drivers become high impedance. The NOP instruction (hex: 00) is con- nected to the data bus of the CPU via IC1. This instruction continues until page 4-K on address com- parator IC2 (set by switches SI . . . S4) is reached. The flip-flop then changes state, the outputs of IC1 become high impedance, and the bus drive rs are released by the BUSACK signal. le circuit described here gives a indication whether the rate of incoming pulse train is above or ow a predetermined value. Based a type 74LS123 dual retriggerable onostable multivibrator (MMV) with sar, it should find ready application microcomputer systems operating I program commences to run address 0000 It is, desirable for many applica- ris to access a RAM range (for in- a RAM card) at this address, e present circuit ensures that the s driver remains at high impedance the CPU reaches, for instance, t required start address of the initor program through execution hard-wire NOP instruction. As instruction has an execution time ith a clock of 4 MHz), the cannot remain in that for more than 0.06 s. is taken from its socket removed from the Z80 card, i-way wire-wrap socket together the remainder of the additional is then mounted onto a small g board. Pin 26 of the socket is because it carries the BUSACK inal of the CPU. The data and ad- bus, as well V, and earth the wire-wrap cet. The completed additional cir- into the socket for and the wire-wrap socket re ceives the CPU. The new BUSACK signal (pin 11 of IC4) is connected to pin 26 of the old CPU socket, and pin 3 of IC3 to pin 8 of the socket for (the removed!) IC5 on the Z80 card. A tip: how to set switches SI ... S4 as required is described in detail in Address Decoding' in the February 1984 issue of Elektor. H D. Paulsen 8 . 82 . signal. Input B2 of MMV2 goes high at the leading edge of the trigger pulse, so that MMV2 accepts the negative edge trigger on input A2. Output 02 then goes low, which 9 causes D2, the 'low' rate LED, to light. When fj is higher than f r , MMV1 is retriggered before its internal pulse period has lapsed. This causes out- put Q1 of MMV1 and input B2 of MMV2 to be held at low logic level. Output 02 of MMV2 then remains low and D1, the 'high' rate LED, lights. An error pulse of about 5 ms occurs on MMV2 when the circuit is switched on and this is indicated by D2 lighting. This 'reaction' pulse is necessary because the circuit needs at least one clock pulse to start up. used with a video combiner. If a PAL or SECAM (TV) signal is applied to the circuit an extra 4.43 MHz notch filter must be added at the input. The circuit can be matched to the normal 75 Q video cable by connect- ing an 82 Q resistor in parallel with R18, with the 4.43 MHz filter be- tween the two resistors. M for frequencies up to 500 kHz The design for a fast analogue to digital converter shown here clearly shows that this type of circuit does not necessarily have to be com- plicated. Instead of the usual sawtooth generator + comparator + counter + oscillator, we have used a system in which a fixed reference voltage is fed to a number of com- parators. This is known as a parallel converter. The delay normally in- troduced by the counting process is done away with so the whole pro- cess is very fast. The disadvantage of this set-up is the large number of components as each step requires a comparator but, in the three-bit example here, that is not a problem. The reference voltages for the various comparators are generated by means of a series of 1% resistors and a current source based on T1. The conversion factor is set with PI (U r ef = 1.5 ... 9 V). The analogue input voltage is fed via buffer stage IC4 to the inverting inputs of A1 . . . A8. A priority encoder is used for the conversion to binary code. It achieves this by translating the number of the highest com- parator activated into a three-bit binary code which appears (inverted) at the output of IC3. With the com- ponent values given here the circuit will operate up to about 500 kHz. Apart from the usual applications, this circuit can also be used, for in- stance, to make unusual effects in a video signal or to convert a black and white picture to colour, when J Aug/Sep! 198-1 8.83 3 bytes in two EPROMs Almost everybody who builds micro- computer projects will notice sooner ir later that he has been stockpiling certain often-used components. A i in point is the 2716 EPROM, which is so commonly used that it is wise always to have a couple on hand. In spite of the fact that the 2716 is so common, EPROMs with double this capacity 12732 = 6x8 bits) are also very popular. This doesn't mean, of course, that everybody should throw away all their 2716s. Quite the opposite, in fact, we thought it would be in- teresting to have a 4 K memory con- sisting of a pair of 2716s. All the lines intended for the 2732 e used directly by the two 2716s except for All, CS, and V p p. Every on the 2716s is common to both are taken from the outputs of the 74LS00. One of the EPROMs (refer- red to here as 2716 (1)> is addressed for the first 2 K block of the '2732', and All is then logic low. The sec- ond 2716 is enabled when the sec- ond 2 K block is being accessed (All is then logic high). Remember to ap ply th e appropriate logic levels for the"&E and V pp pins: pin 21 must be connected to +5 V and pin 20 to The method of construction and fit- ting of this circuit should be carefully considered to cause the minimum of disturbance on the printed circuit board. K modulate IC1. Pin 6 of the 567 is the trigger input so that the audio signal is superimposed on a HF (about 50 kHz) triangular signal. This causes the rectangular output signal to be pulse width modulated. The re- mainder of the 1C is used as a buffer so that the 567 can drive infra-red LED D1 directly (at a peak current of at least 100 mAI without the need also usable for other audio signals This circuit is used together with the receiver described elsewhere in this issue to form the simplest infra-red wireless headphone system im- aginable. It uses a pulse width modulation (PWM) system which, although unsuitable for critical hi-fi applications, gives a reasonable qual- ity and has an acceptable range. The transmitter is based on an LM 567 tone decoder 1C. The layout used is somewhat unusual but the chip's internal VCO and switching stage combine to give much better linearity than could be achieved with, for example, a simple circuit based a 555 ti r 1C. The operation of the circuit is quite for any external components. The straightforward. The audio signal (at transmission frequency can be set least 50 mV pp ) is amplified by tran- between about 25 and 40 kHz by sistor TI and is then used to means of preset P2. video signal + sync separator = sync of a video signal. When supplied signal with a composite video signal of at least 0.5 V pp the circuit outputs quite a respectable (9 V pp ) synchron- ization signal. This is eminently oynu o* u ™ suitable for use with the video effect circuit (video D/A) described elsewhere in this issue. The basis of the circuit is a com- parator consisting of two transistors, the inverting input (T2) of which is connected to a fixed d.c. voltage. When the input signal at the non- inverting input (the base of T1) falls below the voltage set at the base of T2 (about 3.6 V) transistor T1 switches off and T2 conducts. If a video signal is applied to the input the d.c. voltage setting of T1 will be slightly higher than that of T2. On top of this the base setting circuit of T1 contains a clamping diode which will only allow a very small change in the negative direction (roughly 0.4 V). The result of all this is that the video signal at the base of T1 will never fall below about 3.2 V. This limiting of the lower values means that only a small part of the input signal (provided it is larger than the minimum value) will affect the output signal. In the positive direc- tion T1 simply conducts all the more and T2 remains switched off (the output is then about 12 V). During the sync section, however, T1 will switch off so the sync pulses appear, amplified, at the output. The current consumption of the cir- cuit is only a few milliamps. M ... for home and garden A periodic alarm signal has many ap- plications in daily day life: "lights off' indicator in cars, water level in- dicator, alarm clock, memory aid, limit indicator, and calling signal are but a few. The circuit begins to operate as soon as its input level becomes "0"; after about 30 seconds the buzzer sounds four times at one-second intervals. This happens every thirty seconds until the input goes logic high again. The circuit is based on a 14-stage CMOS binary counter and oscillator type 4060. The oscillator frequency, f, is determined by f = 1/2.2R3C1, where f is in Hz, R3 in ohms, and Cl in farad. The oscillator is internally connected to the clock input of the counter. As soon as the reset input (pin 12) is logic low, the counter begins to operate. Because at the onset out- puts Q4, Q7, and Q10 are logic "0", pin 12 goes low when the input to N1 is "0". After about 30 seconds, Q10 becomes T. The 1 Hz signal on Q4 is then applied to the base of transistor T1. This transistor therefore conducts in rhythm with the 1 Hz signal and switches the buzzer on and off at the same frequency. After four seconds output Q7 (pin 6) also becomes logic T. As both inputs of NAND gate N3 are now logic high, its output becomes "0". This level en-. sures that the reset input (pin 12) of IC2 briefly goes high, so that the counter resets all outputs. If the in- put to the circuit is still "0", the pro- cess starts anew; otherwise the alarm stays quiet. H R. Rastetter FSK filter for computers A problem well known to personal computer users is the difficulty of swapping cassette tapes containing software. One of the main reasons for this is the setting of the read /write head in the cassette recorder. This should be at 90° with respect to the tape but in practice this is not always the case, with the result that loading a program from a 'strange' tape causes problems. When using FSK (Frequency Shift Keying) the signal cleaner here pro- vides a very marked improvement. The time spent searching for the cor rect signal level is then greatly re- duced. As the filter requires only five components there should be no problem finding a space for it within the case of any computer. The layout of the circuit is not at all complex. The signal passes first through the low-pass filter, consisting of R1 and C2, which has a cut-off frequency of about 1600 Hz. In fre- quency shift keying a '0' or '1' is recorded on the tape as a sinusoidal signal (with frequencies of 1200 and 2400 Hz respectively) so this filtering removes all the Tough edges' (figure 2al from the signal. The result is shown in figure 2b. The two diodes limit the amplitude of the out- put signal to about ±600 mV. M O&^KSJ- -T-°* An auto reset is also provided but this should be omitted if the par- ticular 6502 system has an automatic reset circuit. In the latter case, it is, of course, the very task of the pres- ent circuit to indicate that RES was the last (and first!) of the three signals. If the auto reset is fitted, it will ensure that the LEDs are switched off when the circuit is first switched on. M handy unit for 6602 users The circuit presented here was designed primarily for 6502 users. It indicates which of the following signals occ urred last RESET = RES = 0 INTERRUPT REQUEST = IRQ - 0 fJMi = 0 This information is particularly helpful in the event of failure of a 6502 microprocessor system. It is equally useful during the handling of specific software for such a system. The circuit effectively forms a three- state indicator and consists of three NAND gate latches. Each latch is set by one of the three signals men- tioned above. When that happens, the latch in question resets the other two latches via the relevant diodes. At the same time the high Q level causes the relevant transistor to con- duct and this in turn makes the ap- propriate LED light. This LED will re- main on until one of the other two latches is set. Jfll E? •? 8.86 elekior india Aug/Sepl 1984 overvoltages and short-circuits need no longer be a danger to microcomputers When the 5 V and/or 12 V part of a computer's power supply breaks down it can mean one of two things: the supply will be either too high or too low (generally zero volts). When the voltage drops the consequences are generally limited to the RAM memory being erased or corrupted. The 12 V supply is protected in much the same way. When the voltage reaches about 12.7 V thyristor Th2 conducts and shorts the 12 V line to ground. The supply to the coil of the relay is then cut off so Rel once again falls out. If there is a short-circuit within the computer itself the effect is the same, except, of course, that our circuit does not have to provide the 'short'. When the short-circuit or fault is found (and cured) the supply can be switched on again by pressing push button SI. An 'off' button has also been provided, which should be of particular interest to ZX users as these computers do not have an on/off switch. Constructing this circuit is aided by the availability of the printed circuit board shown in figure 2. It is prin- cipally intended for use with the microcomputer power supply pub- lished elsewhere in this issue and it can be connected directly to that cir- cuit's + 5 V and +12 V outputs. If you wish to use this protection with a computer with no 12 V supply line the 12 V section can simply be left out. The relay will then have to be changed so that a 5 V type is used instead of 12 V and it then has to be connected to the 5 V supply line. M R1.R2 = 100 Q R3 = 1k2 R4 = 470 Q D1 = 5V6 400 mW ze 02 = 1N4148 D3 = 12 V 400 mW z< D4 = 1N4001 T1 = BC 140 Tht,Th2 = TIC 106 Miscellane The effects of the voltage rising, due to a faulty voltage regulator, for example, are far more serious. The chances of all the 40,000 to 100,000 transistors in the microprocessor sur- viving something like that are quite slim. That is more than enough reason to find a place for this supply protection circuit in any computer. This circuit disconnects the com- puter's power supply from the mains when its output voltage becomes too high or if it detects a short circuit. If, for example, the voltage on the 5 V line increases for some reason zener diode D1 will start conducting at about 5.6 V. This causes thyristor Thl to conduct and short the offend- ing supply line to ground. (It is essential for this method of operation that the computer's power supply is current-limited). This causes tran sistor T1 to switch off with the result that the relay drops out and takes the mains supply with it. diaAug/Sep1 1984 8.8 7 i small, but useful, TTL lesi aid Regular readers of Elektor will know that we often publish various items of test gear. The design shown here is nothing really unusual but it is none the less worth considering because it is very handy. The end product is about the size of a thick felt-tip marker but this marker comes with built-in 'intelligence'. One of the three LEDs in the circuit will light depending on the voltage A1 ... A3*y.lC1 = LM 339 measured at the test point (TP). This voltage is first of all fed to two com- parators (A1 and A2). A reference voltage is fed to the other input of each comparator from voltage divider R4/R5/R6. The values chosen give thresholds at 0.8 and 2.4 V as the range between these two levels is a 'forbidden area' for TTL. If the voltage at TP is lower than 0.8 V the output of A2 goes low and causes the red LED (D6) to light. If the measured voltage is higher than 2.4 V the output of A1 will be low so the green LED (D5) will light. Sometimes, of course, the voltage will be between 0.8 and 2.4 V and then neither the output of A1 nor the output of A2 will be low. When TP is not connected to anything the same thing applies due to the action of R1 and R3. The inverting input of A3 is then pulled high via R9 so the yellow LED (D7) lights. As we have already suggested, the completed circuit can be made into a very attractive finished product. All the components can be mounted in a line on a narrow piece of Vero- board, as the photo shows, and this can then be slipped into some sort of tube. Ideally this should be transparent to enable the LEDs to shine through. H -* scales from 0 to 999 An amplification selector is an ac- curate measuring instrument that is inserted into a signal path and then allows the gain of that signal to be set precisely between 0 and 999 in I Amplifier A1 functions as a (unity gain) buffer for the test signal which is subsequently applied to a chain of resistors, R8 . . R16, and then to amplifier A4. Amplifiers A1 . . . A3 are connected in cascade. Whereas A1 has unity gain, amplifiers A2 and A3 have a gain of x10. Each of them is fol- lowed by a similar chain of resistors as A1, R17 . . . R25, and R26 . . . R34 respectively. From the chains, the signal is also applied to A4. The gain depends on the setting of switches SI . . . S3. You will see from the circuit diagram that the resistor chains, together with R35, are part of the negative-feedback loop of A4. The result is a mixing amplifier with a conversion gain be- tween 0 and 999. The total resistance of each of the resistor chains is 100 k. If then for in- stance the three switches are in pos- ition 1, the total amplification is R35/R8+ . . . +R16 = 1 10R35/R17+ . . . +R25 = 10 (gain in A2!) ?00R35/R26 + . . . +R34 = 100 (gain in A2 + A3!) = 111 8 elekloi india Aug/Sepl 1 9 amplifier, IC2, then brings the audio signal to a suitable level for the headphones. In its basic format the receiver con- tains no form of filtering so there is quite a lot of interference from other light sources such as the sun and ar- tificial lighting. This makes the final audio signal very noisy. Fortunately this effect can be greatly reduced by shielding the photodiode from sunlight, or, even better, by fitting a lens in front of it. This last idea is for unhampered listening pleasure This circuit is the receiver to match the transmitter published elsewhere in this issue. Together these two form a very simple, but none the less effective, wireless headphone, or a transmission system for any audio signal. The signal is transmitted by the infra- red LED in the transmitter. When the receiver reconverts this IR signal into electrical pulses, it will be a rec- tangular waveform in which the width of the pulses corresponds to the audio information. The signal ob- tained after amplification and filtering only has to be integrated in order to retreive the audio information. What could be simpler? The SL486 1C is a suitable single- chip receiver for our wireless head- phone system as it contains, among other things, a regulated amplifier and a filter section. The signal picked up by the infra-red photodiode (such as a BP 104) is fed to the input of IC1, while at the out- put is the integrator consisting of resistor R1 and capacitor C9. An particularly interesting as by ex- perimenting with a lens we can in- crease the range from the original 5 ... 10 metres up to 20 . . . 50 metres. That is really quite good for such a simple circuit. H 1 8.89 same sensitivity irrespective of the shape of the input signal. The ideal input to this circuit is a small signal fed to the schmitt trigger inverter via capacitor Cl . The value at the output is averaged by means of R2/C2 and fed back to the input. The d.c. setting at the input is thus 'tailored' to the amplitude of the in- put signal, with the result that any signal whose magnitude is larger than the difference between the schmitt trigger's thresholds (the hysteresis) will cause a rectangular waveform to appear at the output. The sensitivity of the circuit can be it by means of preset potentiometer It is quite a common practice to use a schmitt trigger to generate a rec- tangular waveform from some other signal. The duty cycle and frequency at the output do, of course, depend on the form of the signal at the in- put but, other than that, the shape of this latter signal is relatively unim- portant. The one essential require- ment is that the signal must exceed, or at least reach, the triggering threshold of the schmitt trigger. With the circuit shown in figure 1 that is not necessarily so. In order to be able to process smaller signals a.c. coupling may be used and the signal need then only be greater than the hysteresis. The amplitude of asymmetrical signals must, however, be larger than this or the upper or lower threshold may not be reached. The signal in figure 2a will thus produce an output whereas there will be no output if the signal of 2b is applied to the input. The cir- cuit illustrated here always retains the Unfortunately, as in so many things, there is a small cloud wrapped around this circuit's silver lining, namely that when there is no signal at the input the circuit acts as a free- running oscillator. In order to prevent this happening when there is a signal present the frequency of oscillation of R2/C2 must be at least ten times lower than the frequency at which the circuit is used (100 Hz with the values shown). It is then an ideal auto-trigger for an oscilloscope, for example. A disadvantage of the cir- cuit is that it is not very suitable for signals with a very short duty cycle as small amplitude differences then produce broken pulse trains at the output. M . . .for power amplifiers There is no doubt about the value of a switch-on delay for a power amplifier. We all know the irritating (and potentially damaging) pop heard from the loudspeakers when the power amplifier is switched on or off. The circuit described here pro- vides a technically simple, but nonetheless satisfactory, solution to this problem. A relay is used to isolate the loudspeakers until the switch-on surge has passed, as this is what causes the loudspeakers to pop. Switching off an amplifier may also cause loudspeakers to pop so this circuit prevents this happening by switching the 'speakers out of circuit just before this happens. As the diagram shows, we have kept the circuit as simple as possible. Because of this the circuit is both in- expensive and easy to build. One slight disadvantage of this design is that it can only be used with a power amplifier that has a sym- metrical power supply (with a maxi- mum of ± 60 V). This is not really such a big problem as most modern power amplifiers have a symmetrical supply. The operation of the circuit is perfectly straightforward. The a.c. voltage is tapped directly from the amplifier transformer and half-wave rectified by diode D1. The voltage divider resistors, R1 and R2 must have suitable values so that the maximum voltage on Cl is about 5 V higher than the relay voltage. The values given in the diagram are suitable for a U b of 45 V and a relay voltage of 24 V. If different specifi- cations are chosen the values of the components must, of course, be suitably adapted. The relay voltage is particularly important as this must be at least 2 V lower than U b . It should also be remembered that the relay must be able to switch a large current; something in the order of 10 amps is not unusual (depending on the power of the amplifier). When the power amplifier is switched on. Cl is charged via R1 to about 29 V (in our example). Tran- sistors T1 and T2 follow the capaci- tor voltage until the zener voltage of D1 is reached ( U zener = Urelay + 1.4 V). The voltage is now sufficient to switch the relay, and with it the loudspeakers. The value of Cl stated ensures a delay of about 5 seconds before this actually occurs, and this is time enough to allow the amplifier to stabilize so no pop is heard. This time can be made longer or shorter by changing the value of Cl. When the power amplifier is switched off the same thing happens, in prin- ciple, but in the opposite order and much more quickly. The voltage drops as Cl discharges via R2. The circuit is 'tuned' so that the voltage across Cl falls quite quickly below the relay voltage and the relay then drops out. The loudspeakers are then certain to be switched out of circuit before the pop should be heard. Finally it should be noted that even with good cooling T2 must never dissipate more than 5 W (P = l re x ( + U b - U re )>. M 0.6. . .0.8 (-4.5. . . -2.0 dB). The output impedance of the amplifier is not greater than 200 ohms. An external amplifier may be connected to pin 4 of the 1C. When the input voltages lie above 2 V RMS, a potentiometer should be used as a voltage divider as shown If the overall gain is too small, a transistor should be used instead of the FET (also with 10 mA quiescent current), but the circuit then virtually reverts to that published in Elekt> U.K. in April 19821 The MOC 5010 opto-coupler may be used to isolate a circuit from the mains, as audio interface, in medical electronics, and in many other applications. Because of its high isolation resistance (10" ohms), the MOC 5010 is eminently suitable for applications where a circuit is connected directly to the mains, as, for instance, in most TV receivers. It can therefore be used to give enhanced perform- ance to the 'TV sound interface' described in the April 1982 issue of Elektor U.K. With a bandwidth stretching from 5 Hz to well over 100 kHz. there is no need to worry about the audio response as there was in earlier opto-couplers. Basically, the MOC 5010 converts a variation in input current into a varia- tion of output voltage. Input voltages are first transformed into currents. The circuit shown in figure 1 has an amplification factor of about 0.75. Its input should not exceed 2 V RMS, while the bandwidth is 118 kHz at the -3 dB points. Field-effect transistor T1 functions as a voltage/current converter: its slope is about 3 ... 4 mA/V. The quies- cent drain-source current is about' 10 mA. Amplifier A has a transfer resistance of around 200 mV/mA so that the total gain is of the order of It is important to note that two separate power supplies are required: not only the two +12 V terminals, but also the two '0' lines must be kept isolated from one another! In many cases it should be possible to obtain the +12 V for the transmitting end of the circuit from the TV set: this is, of course, easily found out if you have the service manual or even a circuit diagram of the set. K klor India Aug/Sept 1 984 8.91 see the phone ringing In many instances it is not only the hard of hearing who are unable to hear the telephone ringing: even with normal hearing it is often impossible to detect above the noise from the vacuum cleaner or the radio. The present circuit enables the ringing of the phone to be seen with the aid of a flashing lamp. It is perfectly feas- ible to put a number of lamps in parallel and place them in different locations. Inductor LI is attached to the telephone by a suction pad: it may be necessary to try out several pos- itions on the telephone to obtain best results. A reference voltage of about 4.8 V is provided by potential divider R1/R2 and applied to the non-inverting in- put of opamp IC1 directly and to the inverting input via L1/P1. The preset is adjusted to give equal levels of | direct voltage at both inputs of the opamp: the output of IC1 is then logic low. When the telephone rings, an alter- nating voltage is induced in LI, caus- ing the potential at the non-inverting input of IC1 periodically to exceed that at the inverting input. This results in a rectangular pulse train at the output of the opamp. The trailing edges of these pulses trigger one half of IC2 via C8. This half of the 1C operates as a monostable multivibra- tor (MMV), the output of which is low during time-out. When a pulse arrives at pin 6, the timer is triggered and the output (pin 5) goes high. As long as the output is high, subse- quent pulses at pin 6 have no effect: only when the MMV has reset does the next pulse at pin 6 trigger the timer. The output pulse has a width of about five seconds, which is determined by the values of R4 and C3. The second half of IC2 functions as an astable multivibrator producing Capacitors: Cl = 1 p/16 V C2.C3.C5 = 10 p/16 V C4.C7.C8 = 10 n C6 = 4p7/16 V rectangular pulse trains when its reset input (pin 10) is high, which is as long as the MMV is triggered. The pulse repetition frequency is determined by the values of R5, R6, and C6. The output signal on pin 9 of the AMV switches relay Re on and off. As the pulse spacing is just about one second, the relay, and therefore the lamp(s) connected to it, is switched on and off five times. The quiescent current consumption of the circuit is about 10 mA at 6 V. In selecting the relay, its operating voltage as well as the power rating . of the lamps should be taken into account. The printed-circuit board for this cir- cuit is not available ready-etched; it may, however, be produced with the aid of the track layout diagram (no. 84407) given in the PC board pages at the centre of this issue. M Semiconductors: D1.D2 = 1N4148 IC1 = 741 IC2 = 556 Miscellaneous: LI = telephone pick-up coil Re = relay, see text 8 . 92 . r India Aug/Sept 1 984 with radiation counter All that's required to erase an EPROM is basically an ultraviolet (UV) lamp which radiates the EPROM window at the right distance (about 2 ... 3 cm) for a period which depends on the manufacturer (normally 10 ... 40 minutes). As we don't think you'll want to sit around Rel then ceases to operate and switches off the UV lamp. This is ab- solutely necessary as ultraviolet light is extremely harmful to your eyes. The clock frequency may be set by means of PI in two ways: (1) with the aid of an oscilloscope or frequen- cy counter to 6.85 Hz. or (2) by measuring the time taken for output Q12 of IC2 to become logic 1 after reset switch SI has been pressed: this should be exactly 10 minutes. Note that the relay contact must be rated for switching 220 V a.c.l The relay itself may be of the pc board type. The mains power supply may be any well-regulated type giving 6 V d.c. The current consumption without the relay is about 5 mA. Using the eraser is fairly easy: lay the EPROM on a flat surface and place the case over it after having set S3 to the required erase time (10, 20, or 40 minutes). Lighting of the red LED indicates that erasure is in progress. A push on SI ensures that the cor- rect erasure time will be run through: this is necessary as the counter begins to count as soon as the supply voltage is switched on. H gazing at your wristwatch while all this is going on, we have designed a timer which automatically ensures the correct radiation time and in- dicates the end of the erasure period. The counter-IC, type 4060, has an in- tegral oscillator the frequencylof which is determined by R2, R3, PI, and C3. When the supply is switched on, IC2 receives a reset pulse from C5 which makes it start counting. Outputs Q12 . . . Q14 are logic low, and T1 and T2 conduct. When S2 is closed (see below), relay Rel is actuated and the UV lamp is switched on. In addition, red LED D2 lights. The base of T3 is connected to the positive supply line via T2 so that T3 is cut off. After the time set by S3 has lapsed, the relevant out- put of IC2 goes high. Transistors T1 and T2 are then cut off, the relay switches off the UV lamp, and LED D2 extinguishes. The base of T3 is then connected to earth via the relay coil and S2: transistor T3 conducts and green LED D3 lights to indicate the completion of erasure. A tip: fit the UV lamp in a suitable case with open underside as shown in figure 2. Push-button S2 should be mounted in a way which ensures that it closes when the case is laid flat on an even surface, but opens as soon as the unit is lifted; the relay dia Aug/Sept 1984 8.93 1 PROGRAM TO CALCULATE PARALLEL AND SERIAL RESISTORS FOR 0=1 TO 36: PR I NT ■ X= 1 : Y= 1 : I NPUT " RES I STAI IF R< OR R> l64»O0eO INPUT -TOLERANCE IN X L=R-*R:U=R*(T, No, the title does not relate to a help in grammar; PARSER is not a circuit either, but rather an aid in designing one. It often happens that non- standard resistors are required. The normal solution then lies in connec- ting standard resistors in series or parallel or combinations of these. The computation on a pocket calculator to arrive at suitable stan- dard values can be quite a long one, even if the actual tolerances of the various resistors are well within their nominal values. With PARSER it becomes almost a routine operation that does not take very long, provided you have a BASIC microcomputer available. As you may have guessed by now, PARSER is a small BASIC program for the determination of innumerable combinations of resistors within a given tolerance in a very short time. When you have worked out some examples other than the one given with the program, you will see what the fuse is good (not blown) the base current for T3 is always pro- vided via R5 and D1, with the result that the LED lights continuously. When the fuse blows the base cur- rent is only provided by the AMV and. as this is not continuous, the LED flashes. The current consumption of the cir- cuit is about 30 mA, most of which is due to the LED. If the indicator is fitted to some battery-powered cir- cuit it is worth while to use a high- efficiency LED for D3 and to change the value of R6 to suit the lower LED current. M E. Neefjes ideal for those with perfect pitch It is sometimes useful to have a small instrument that can give a quick indication of the approximate value of a resistor. The present circuit enables an unknown resistor to be compared with a number of known resistors and in that way indicate between which two values the unknown resistor lies. The circuit is based on the well- known 555 which is connected as an oscillator (astable multivibrator). The output of the oscillator is used to drive a piezo electric buzzer. The frequency of the oscillator is in- versely proportional to the value of Rx (the unknown resistor) and is determined from where In2 = 0.6931, all resistors are in ohms, and C2 is in farads. By substituting one or two of the known resistors for Rx, the note emitted by the buzzer should give a fair indication of the approximate value of Rx. Of course, if you have perfect pitch, you do not need the known resistors ... In that case, we'll tell you that if Rx = 0, the fre- quency is about 4500 Hz, while when I Rx = it is 2 Hz. easy-to-make reference circuit It is often interesting (if not required) to know whether an amplifier is going into saturation, or whether certain limiting values (thermometer, power supply, etc.) are being ex- ceeded. It is, however, not always feasible to use a fully-fledged win- dow discriminator and in those cases the circuit presented may be of interest. When the level of the input voltage lies between 3.5 V and 8.5 V, tran- sistors T1 and T2 conduct (T3 and T4 are cut off) so that LED D1 lights to indicate that the input signal is 'in When the input level rises above about 8.5 V, T2 and T3 conduct (T1 and T4 are cut off) which causes LED D2 to light indicating that the input level is 'above range'. Finally, when the input signal drops below about 3.5 V, T1 and T4 con- duct so that LED D3 lights indicating that the input is 'below range'. The current consumption is for all practical purposes governed by the LED currents which are 20 mA maxi- mum: it may briefly rise above this value during switch-over. When the input (junction R1/R3) is disconnected, the input voltage equals about half the supply voltage (6 V) so that LED D1 will light. M dia Aug/Sept 19848.95 I This has got to be one of the simplest electronic bell extension cir- | cuits ever designed. In all it contains seven components and none of | them are even slightly unusual. They e the kind of parts that most elec- I tronic hobbyists will probably have 8.96 eleklor india Aug/Sepl 1984 high so a resistor is connected in series with the rectifier and a zener diode across the buzzer. The values used give a voltage of about 5 V across the buzzer, but, depending on the type selected, this can quite eas- ily be changed. Furthermore, if the input voltage is more than about 10 V a.c. the value of the series resistor will have to be increased ac- cording to Ohm's law (U = R • I). Take care not to exceed the buzzer's maximum permitted current. ► lying around somewhere, and these seven components are all that are needed to make a universal telephone bell extension circuit. The telephone bell operates on an a.c. voltage so this must be rectified by the four 1N4148 diodes in order to make it suitable for the d.c. buzzer. Obviously, the voltage across the buzzer cannot be allowed to rise too a CPU detour During its initialization procedure, the 6502 processor starts by getting the start vector which is located at ad- dresses $FFFC and $FFFD in ROM, This is a fixed instruction that cannot be changed, and it points to a memory zone in PROM, which, in most computers, is very difficult for the user to access. The circuit described here makes it easy to reroute the 6502 to a start address chosen by the user: $XFFC/$XFFD where X is any hexadecimal value. At this address the CPU will find the appropiate vector pointing to the start routine written by the user (in EPROM) instead of the standard routine written by the manufacturer. The only hardware change required to achieve this is to connect the cir- cuit shown between the 6502 and its bus. Now every time the CPU emits an address between $FFF8 and $FFFF (the address decoding is a little less precise than is necessary for only re-routing the processor when it outputs addresses SFFFC and SFFFD), the bus receives an ad- dress between $XFF8 and $XFFF, where X is determined by the user by means of four switches (or four wire links). If S4, for instance, is switched to +5 V A15' is equal to A15, but if the other position is selected A15' = A15. To use this cir- cuit, lines A3 . . . A15 on the bus must be fed to N1 and the link be- tween outputs A12 . . . A15 of the 6502 and the system bus must be broken. These lines are then con- nected to lines A12' . . . A15' of the detour circuit. Each of lines A12 ... A1 5 is connected to one of the inputs of AND gates N2 . . . N5. The second input to each of these gates is fed by the logic level set by the user with the switches. The resultant binary word constitutes the hexadecimal value of X in the desti- nation addresses SXFFC and $XFFD. In most cases this memory zone will be found in an EPROM which, apart from the RESET vector, will probably also contain the initialization routine. Remember, of course, that the change described here also implies that the IRQ and NMI vectors (XFFE/XFFF and XFFA/XFFB respectively) and the corresponding routines be modified accordingly. M _ 1 ) (") ii 'Some like it hot' You've just finished a hard day's work and are heading home, looking forward to a relaxing evening. Sitting in your favourite chair while your faithful dog brings your slippers and paper. The crowning glory is, of I course, the cup of hot coffee . . . But sometimes it doesn't work quite I like that! The dog has to be pried out of what is also his favourite chair, and it takes almost all your- effort to coax him, grumbling and growling, to reluctantly fetch your slippers and paper. You put on the slippers and your feet begin to feel decidedly damp, the rain has made the ink in the paper run, and then to top it all the coffee is too cold. Before you chuck in the towel . . . read on; we may not be much good at canine psychology, but we do have some ideas about coffee. These is little dispute that the best temperature for coffee is at least 80 degrees Celcius. That is the temperature where your tongue just begins to . . . but let's not go into that here. Because coffee has also become a common prescription for the aliment known as 'Monday- morning', we decided that it would be better to remove all traces of guesswork from this question of 'how hot is it?'. As the diagram shows, there is not very much involved in this circuit. A voltage regulator, a temperature to voltage converter, a comparator, a couple of transistors and LEDs, and a handful of resistors and capacitors, is the total component count. The operation is also straightforward. If the coffee is at less than the correct temperature the output of IC3 is low, keeping T1 switched off. The other transistor, T2, therefore conducts and r India Aug/Sept 1 984 8.97 the red LED lights to show that the coffee is too cold. As soon as the temperature is high enough (above 80°C), the green LED lights. What actually happens is this: The temperature, which is measured by IC2, is converted to a voltage. The idea is that the LM35 should hang in the coffee, so the three connections to the 1C must be isolated. This can be done by fixing the temperature sensor in an old ballpoint pen, or by sealing it with some non-poisonous two-component glue or in some heat-shrinkable tubing. The output voltage of IC2 increases by 10 mV for every degree Celsius rise in temperature. The reference voltage at the inverting input of IC3 must be set to 800 mV with PI. As soon as the voltage on the non- inverting input also reaches 800 mV, the output of the comparator- switches to high. This causes T1 to Oh, for the good old days, when if you wanted to run a motor for two minutes you switched on the power for two minutes. Now we have computer-controlled robot arms, elec- tronic mice, and all manner of technological advances. For all this, however, many people still shy away from the idea of something like a motor-driving circuit. As the drawing here shows, such a circuit is quite straightforward, especially as we have even gone so far as to design a printed circuit board for it. The circuit has two inputs and if both are T (+12 V) nothing hap- pens. As soon as the voltage on one of the inputs, A, for example, becomes zero driver transistor T5 conducts. This causes both T1 and T4 to conduct and the motor turns in a particular direction. This brings us to the stage where we must ex-, plain why the circuit is 'economical'. It will not have escaped your notice that each pair of transistors in the bridge is controlled by a single driver transistor. This not only saves com- ponents, but also saves the energy that would otherwise be used by two driver transistors. When T5 is made to conduct T1 will conduct. At the same time a current flows from T1 via T5 to the base of T4 so this tran- sistor also conducts. This means, in effect, that we are using the base current of T1 and T3 to drive T4 and T2 respectively, giving us a common driving circuit. There are two other components that merit a few lines of explanation, namely D5 and D6. These ensure that nothing untoward happens if both inputs are earthed at the same time. If, for example, input A is at zero volts both T1 and T4 conduct and the anode of D6 is connected to the +12 V line. If input B is now earthed T6 (as well as T2 and T3) cannot conduct because its base is kept positive. Input B can only be activated, therefore, after the voltage at A goes high, and vice versa. Pulse-width modulation could be used to control the speed of the motor. What this means is that the signal fed to input A or B is not continuous but a string of pulses whose width can be varied. The nar- rower the pulses are the faster the If heavier motors are to be driven T1 . . . T4 may be replaced by dar- lingtons that are rated high enough 8.98 elektof india Aug/Sepl 1984 to handle the expected current. The inputs to this circuit are inten- tionally 'active low' to enable it to be easily driven by TTL logic. The out- puts of TTL gates can switch a few miiliamps to earth but can supply very little current themselves, cer- tainly not enough to drive a tran- sistor. If the supply for the motor is greater than 5 V the TTL gates must have open-collector outputs. The maximum current that the motor can draw is about 1 amp and the quiesc- ent current consumption is almost from rectangle to triangle I Sawtooth generators are required in electronics for many purposes. Typical examples are found in music I electronics where the rectangular j output of a single octave divider must be converted into a sawtooth- shaped signal, or in measurement j technology to provide the control i signal for an analogue-to-digital | converter. "Ln JZL n_n_ cessive Q output of IC1. For in- stance, the amplification of output 02 is two times that of Q1. Since the frequency is halved at each suc- cessive Q output, this means that the higher the pulse rate, the lower the amplification as is shown in figure 1. If high stability resistors are used, the resulting steps in the out- put will be symmetrical; with the values shown small deviations from linearity in the step will occur. The time/voltage characteristics show clearly how the stepped waveform is built up. For con- venience's sake, the inversion in the opamp has been ignored: what is im- portant here is the mathematical relation between the various waveforms. In reality, the stepped waveform would be a descending rather than an ascending one. Where an ascending waveform is required, a second opamp with unity gain should be connected to the output. The output waveform has 256 steps; this number may be halved by omit- ting R8. halved again by omitting R7, and so on. Resistor R9 must be made about half the value of the last resistor used, as otherwise the height of the output signal will be halved. The fundamental frequency of the In spite of its modest configuration, . the circuit provides a perfectly usable sawtooth signal is the output signal. The (external) clock pulses are applied to a 7-stage binary counter, a CMOS type 4024 1C. The output signals of the 1C, Q0 . . . Q6, are together with the clock signal ap- plied to an opamp which has been connected as a summing integrator. Resistors R1 . . . R8 are so arranged that the value of each is half that of the preceding one. In other words, R2 = J4R1, R5 = !4R4, and so on. The effect of this is that the gain of the opamp doubles for each suc- s that of the finally used output of the 1C. The clock signal should have a fre- quency 256 times the required output frequency. If fewer divider stages are used, the clock frequency may be halved (compound!) for each omitted stage. The height of the clock pulses at 00 ... Q6 should preferably be the same to prevent asymmetry of the stepped waveform. Power requirements are 15 ... 18 V with a current consumption of about 12 mA. h l Aug/Sept 1984 8.99 the same way as N2 (so with 8 in- puts an extra EXOR is placed be- tween N2 and the MMV, the free in- put to the gate is then connected to switch position 7 and position 8 only gets a LED and resistor). Either TTL or CMOS ICs can be used for the EXORs, such as 74LS86, 74HC86, 4030, or 4070. If TTL is used the output pulse will not always have the same width as the MMV reacts at different levels of the waveform. The supply voltage for a TTL or HCMOS version is 5 V, in one pulse per switch position It is sometimes necessary to generate a pulse as soon as a mechanical switch is operated. The circuit proposed here does just that and only needs a small number of components. Furthermore it can easily be expanded for more con- tacts. The circuit diagram shows the layout for a six-way switch. An im- portant feature of the circuit is that it works with both make-before-break and break-before-make switches. A couple of EXOR gates (N1 and N2) are used to detect when switch- ing takes place; N1 handles positions 1 ... 4 and N2 takes care of pos- itions 5 and 6. A number of LEDs (D1 . . . D6) indicate the switch position selected. Every time the position is switched the level at the output of N2 changes, thereby trig- gering the monostable multivibrator consisting of N3, R1 and Cl. With the values stated, this causes a 200 ps pulse to be output at pin 8 of N3. When building this circuit it must be remembered that the inputs of N1 and N2 have pull-up resistors . . N3 = * IC1 = 74LS86 (R2 . . . R4) so there is always a defined level present. The value of the resistor is not at all critical. The LEDs with their resistors may be left out if a visual indication of the switch position in not desired. Extra EXOR gates will have to be included if the number of switch positions is more than six. These are added in other cases 3 ... 15 V is permiss- ible. The length of the output pulse may be changed by using different values for R1 and/or Cl. If CMOS ICs are used the value of R1 can be as high as a few megaohms. The current consumption with CMOS is 10 mA, with TTL this rises to 20 mA. * n-p-n or p-n-p? Transistor testers are nothing new in Elektor: almost every year sees at least one new one. There are, however, not too many which can in- dependently differentiate between n-p-n and p-n-p types. True, the making of such a distinction is not often called for, even though in many a component drawer the two types are thoroughly mixed up. Nor- mally, the data sheet or a list of comparative types quickly gives the answer. If these, however, are not available or there are other reasons why this method cannot be used, the tester described here will prove very useful. Operation is very simple: the test transistor is placed in the socket and push-button SI is pressed. If the transistor pins correspond to pins B-C-E of the socket, it is an n-p-n transistor which is optically indicated by LED D1. If LED D2 lights, it in- dicates that the transistor pins cor- respond to pins (Bl-(C)-(E) and that therefore the transistor is a p-n-p type. How does it work? Transistors T1 and T2, together with associated resistors and capacitors, form an astable multivibrator (AMV), the frequency of which can be set with potentio- meter PI. The test transistor is connected to one of the outputs (collector of T2) of the AMV via pro- tection resistor R6. If the test tran- sistor is an n-p-n type, it conducts when T2 is cut off. At the same time, T3 conducts so that D1 lights. If, however, the test transistor is a p-n-p type, it conducts when T2 does, and this cuts off T3. As the 9.00, collector potential of T1 is then high, T4 conducts and D2 lights. Terminals 1 and 2 have been added as an aside and may, for instance, be used to test the continuity’ of con- ductors. This is possible, because when the terminals are short- circuited both LEDs light. The termi- nals may also be used to deter- mine the anode and cathode of a diode: the LEDs remain extinguished when the cathode is connected to 1, but light with the anode at this terminal. Meter M indicates the current flow- ing through the test transistor: capacitor C3 smoothes the rec- tangular pulses from the AMV. If you do not want this metering circuit, simply connect the anodes of the LEDs to the positive supply line via Rs = 330 Q. The supply voltage should be not higher than 6 V to prevent the emitter-base reverse potential ex- ceeding the maximum permissible level of 6 V should the emitter and base connections be accidentally input of the comparator (pin 5). The trigger level may be set between 4.5 ... 17 V with PI. Points B, C, and D are all connected to the unregulated power supply line. Note that the voltage at pin 12 of the 723 should not be less than 9.5 V. If the unregulated line is lower than this value, pin 12 (point B) must be connected to an auxiliary voltage of not less than 9.5 V. When the voltage at point A exceeds a value predetermined by PI, pins 9 and 10 of the 723 become logic high and the SCR (a type TIC 106 or equivalent) fires. This creates a vir- tual short-circuit between the positive terminal of Cl and earth which causes fuse FI to blow. The time lapse between the overvoltage occurring and the trip action is 1 . . . 2 ps. H for u ;t power supplies Although the circuit described uses I an SCR (silicon-controlled rectifier) i as protection device, it does not depend on direct crowbar action: instead the SCR causes a fuse to blow. A 723 voltage regulator, used as locked comparator and SCR driver, provides an internally generated reference voltage of 7.15 V at pin 6. This voltage is divided by two (R4/R5) and applied to the inverting input (pin 4) of the comparator. The voltage to be protected (at point A) is divided in R1, PI, and R2 and then applied to the non-inverting i Aug/Sept 1984 9.01 nidi iv.fcisi TERMINAL STRIPS Instrument control devices have single station terminal strips which can be mounted side by side or in front of each other with top covers to prevent acci- dental touch and to enable markings The strips, mounted on insulated panesl can take conductors upto 1 8 swg and are rated for S amps. MICA PARTS Mica, one of the best known electrical insulating materials, is dished out in various shapes, sizes and formats. Semiconductor mounting mica washer, mica insulators for trimmers, flashers, regulators, starters and transmission equipment, mica gasket for boiler gauge glass picking, mica powder and flakes and micanite products are some of them. For further information contact: Instrument Control Devices. 14. Manor am a Niwas, Datar Colony. Bhandup, Bombay— 400 078. TV COMPONENTS For colour television sets. MG Electro- nics are manufacturing EHT transfor- mers, width coil, linearity coil and driver transformer. The company claims leadership in the manufacture of deflec- tion components too. 9 i For further information, write to: MG Electronics Pvt. Ltd., Twiga House, 3-Community Centre. East of Kail ash. New Delhi— 110 06 5. MULTIFUNCTION OSCILLATOR VFO 13, a function generator with sine, square and triangle wave forms, incor- porating 1C circuitry, is a product of Vasavi Electronics. Its features include pure sine wave'output. large frequency range of 1 Hz to 1 00 KHz and oscillations without motor boating. * * |j ***>&£> i) For details, contact: Vasavi Electronics, 162. Vasavi Nagar. Secunderabad— 500 003. © ©0 0© O 8 .Ft. Enterprises. 91. Netaji Subhas Road. Calcutta— 700 001. 1C POWER SUPPLIES Omega Electronics offer a wide range of 1C regulated power supplies having variable, dual, dual tracking, triple and fixed outputs. Functioning either in constant current or constant voltage mode, the output voltages range from zero to 60V. An additional feature is automatic crossover for overload and short circuit protection. WELDING RECTIFIER A fully-electronically controlled, thyri- storised. welding rectifier with dual function has been developed by Advani- Oerlikon. Known as Ador Thyroarc. the rectifier functions on three phase power source and the current is infinitely variable from 20 to 500 A. Rigged and housed on sheet steel, the rectifier is protected against water drips, splashes and overheating. It is ideally suited for manual welding using stick electrodes, claim the manufacturers. More details will be available fr Advani-Oerlikon Ltd., Post Box no. 1546. Bombay— 400 001. For more details, contact: Omega Electronics. 36. Hathi Babu ka Bagh, Jaipur — 302 006. CABLE STRAPS Novoflex strapping system offers an effective cable binding method with perforated strapping and small studs. Made from a specially formulated PVC for prolonging the elastic properties and continuous firmness of the strapping, the system offers the facility of opening and closing of cable looms for modifica- tions in all original wire binding purpose. Standard colour of the strapping is black but different colours are made available for bulk orders. The strapping is supplied in 100 metre and 10 metre rolls and the studs in 1.000s. CRYSTAL OSCILLATOR Statek Corporation has announced a series of surface-mountable oscillators using miniature quartz crystals, from 2 MHz to 10 MHz and the package is in a standard 24-pin leadless, ceramic chip carriers. The manufacturers claim small size, low current and high shock resistance as the virtues of their oscil- lators. For information, contact: Electronic Devices, 14. Hanuman Terrace. Tara Temple Lane. Lamington Road, Bombay— 400 007. More details can be had from: Novoflex Cable Industries, Installation material division, P. Box. 9159. Calcutta-700 016. 9.02 elektof India Aug/Sept 1 984 f A AVAILABLE EX-STOCK- LOW PRICES • MICROPROCESSOR IC’S • CMOS/TTL/LS IC'S • LINEAR IC S • TANTALUM CAPACITORS • CRYSTALS/TRIMPOTS • 1C SOCKETS/ZIF 1C SOCKETS • FLAT CABLES/CONNECTORS • RECTIFIER DIODES • SWITCHING DIODES • ZENER DIODES • TBA 810/1044/7204 • MAGNETIC HEADS • TRANSISTORS/SCR • TRIACS 00Q0QOB GENERAL SALES AGENCY, No 4, YAMUNA BUILDING, TARA TEMPLE LANE, OFF. LAMINGTON ROAD BOMBAY-400007. TEL: 369250/386 i 78 V / now you can buy packs at mini rates r \ IS RED AND BLUE [PC© SEPERATION YOUR PROBLEM? Here is a studio equipped with computerised Graphic art equipment Approved by Renowned companies like Larsen & Toubro, SUM Electronics, Tata Burroughs, etc MOJ TYPE FOUNDRY Unit No. 5, 7, & 8, Municipal Ind. Estate, s Above Market Vile Parle (W), £ Bombay 400 056. s — Teh612 34 79. 3^ IZUMIYA IC INC High Precision Products Are Now Available At Very Economical Prices. (JAPAN) VERSATILE 4 in 1 AC ADAPTER USING MODERN TECHNOLOGY REMOTE««=, SUMIC0 ' 1 NOW THE ULTIMATE A R F A IN TV BOOSTERS OR LOW POWER TV TRANSMISSION -nj — VIEWERS YOU CAN EXTEND r^fFll THE RANGE OF ANY L=^ TV TRANSMITTER Y|:| i.e. LOCAL/FOREIGN l L_ j By installing specially f ■ designed and matched high TV-1 " gain SUMICO Booster and i. Antenna, accurately tuned h 1 l to your requi red channels. M I : Outdoor/Indoor Multichannel |_ FT, _2L models also available. _ THE ONLY REGULATED POWER AC ADAPTER OF ITS KIND Convertible tor 2 or 3 Pin Mains Socket. Selectable D C. Outputs 1.6. 3. 4.5. & 6V Polarity Reversible. Regulated upto 300 mA Short Circuit protected with Indicator Spare cords with single, universal plug are available Contact for dealership giving full particulars of experience in the line. IDEAL RADIONICS p. o. Box 4100, Bombay-400007. Manufacturers WE HAVE MOVED TO elcom 1 CTR precious ELECTRONICS corporation eIeIctor fIectronics pvt LtcJ CHHOTANI BUILDING. 52 C. PROCTOR ROAD, GRANT ROAD STATION, BOMBAY-400007 PHONES: 367459. 369478 R.N. No. 39881/83 MH/BYW-228 UC. No 91. combined 30 cms tweeter. •20-20,000 Hz • ‘200-600 Watts. COVOX 1500 Components: 2 full range woofers. 16 cms. 1 tweeter. •40-18,000 Hz. * *30-40 Watts. COVOX 3500 Components: Enclosure- infinite baffle, sealed, 1 acoustic - suspension woofer 20.32 cms., 1 acoustic-suspension mid- range, 1 tweeter, with divided network. COVOX 4500 Components: Enclosure- infinite baffle, sealed. 1 acoustic - suspension woofer 25 cms, 1 acoustic- suspension mid-range 16 cms., 1 tweeter, with divided network. •30-20,000 Hz. ••60-200 Watts. COVOX 5000 Components: Enclosure- infinite baffle, sealed, 1 full range woofer and mid-range combined 25.4 cms., 1 tweeter, with divided network. •30-20,000 Hz. "100-200 Watts. ' Frequency Response Range ** Matching Amplifier Nominal Impedance 8 ohms. International quality created for India by CD5miC iter & Publisher - C. R Chandsrana. 2. Koumari. 1 4th A Road. Khar. Bombay-400 052. & Printed at Trupti Offset. 1 03. Vasan Udyog Bhava Off Tutsi Pipe Road. Lower Parei.' feombay-400 01 3 MODEL JBL Components: 1 full range wooier and mid-range combined, 20.5 cms, 1 tweeter, with divided network. •30-20,000 Hz. "60-200 Watts. COVOX 7000 rnmnruianls; 1 full range COVOX 6000 Components: Enclosure- ' infinite baffle, sealed, 1 acoustic-suspension woofer 30.5 cms., 2 mid-range. 1 tweeter, with divided network. *20-20,000 Hz. "200-300 Watts. COVOX 2500 Components: Enclosure- infinite baffle, sealed, 1 acoustic-suspension woofer 16 cms., 1 acoustic- suspension midrange, 1 cone-type tweeter, with divided network. *30-18,000 Hz. *•40-60 Watts. □□smic super hi-fi stereo speaker systems for balanced sound and true reproduction from 30 watts to 600 watts total power output.