with more than 100 circuits contents □7 AUDIO, MUSIC clipping indicator common-base virtual earth mixer complementary twin-T selective filter . . . . fuzz box guitar preamp LED peak indicator loudspeaker delay circuit microphone preamp noise cancelling preamp Phaser signal powered dynamic compressor stereo noise filter stereo pan pot symmetrical constant-voltage crossover filter tremolo rr-qualiser 4 W car radio amplifier 1 5 W bridge amplifier for car audio MISCELLANEOUS . 84 . 46 . 72 . 67 . 88 . 98 . 93 . 81 . 78 . 75 . 70 . 21 programmable frequency divider . 37 . 18 . 35 . 47 threshold box . 27 . 19 . 60 . 77 . 79 ultrasonic transmitter for remote control . . 59 DESIGN IDEAS I AC touchswitch 51 ' —rrtJto trigger level control 66 capacitance-frequency converter 3 CMOS CCO 92 common-base virtual earth mixer 46 compensated light sensor 26 complementary emitter follower 53 frequency divider using one TUN 30 frequency doubler using 4069 41 frequency-voltage converter 45 LC resonance meter 103 level shifter 57 noise cancelling preamp 75 non-inverting integrator 63 positive-triggered set-reset flip-flop using inverters 65 sawtooth CCO 28 self-shifting register 95 super zener 7 transistor solar cell 100 using LEDs as reference diodes 86 voltage controlled LED brightness 83 voltage controlled monostable 82 voltage-frequency converter 2 DOMESTIC add-on timer for snooze-alarm-radio-clock .. 43 CMOS alarm circuits 69 drill speed control 104 electronic scarecrow 4 electronic weathercock 48 fade asleep 89 multi-purpose time switch 6 one-shot for immersion heater 38 room thermometer 76 scarecrow, electronic 4 telephone 'tell-tale' 55 temperature-to-voltage converter 5 thermostat 8 triac relay 10 weathercock, electronic 48 POWER SUPPLIES automatic charger 71 automatic NiCad charger 13 inverter 23 minireg power supply 62 multiple voltage supply 101 negative supply from positive supply 74 TTL voltage doubler 15 two-TUN voltage doubler 14 0 V voltage reference 40 0 ... 10 V supply 24 +/— 1 5 V regulator 49 TESTING AND MEASURING A/D converter 12 audible logic probe 52 auto trigger level control 66 capacitance-frequency converter 3 complementary twin-T selective filter 72 digital capacitance meter with 555 timer ... 42 distortion meter 102 distortion suppressor 31 eight-channel multiplexer 73 frequency-voltage converter 45 LC resonance meter 103 ohmmeter 97 phase meter 50 simple transistor tester 11 spot-frequency sinewave generator 25 timebase input circuit 22 tri-state voltage comparator 56 voltage-frequency converter 2 3V> digit DVM 105 FUN, GAMES, MODEL BUILDING model speed control 17 osculometer 91 reaction speed tester 54 TV games 44 wheel of fortune 34 MdtV ’ van* i>. HF, RECEIVERS CMOS PLL high-level mixer LED tuner low-noise VHF aerial amplifier MOSFET SSB-adapter RF amplifier with 100 dB dynamic range . . . short-wave converter funing indicator 90 U ' 39 96 16 99 36 64 80 summer circuits : 100+ ! More than 100 circuits in one issue. Complete units, extension circuits, design ideas, tips. The recipe is quite simple: Dream up wild ideas — some of them really wild! — and incorporate them in practical circuits. Cull some novel ideas from stacks of manufacturer’s application notes - and then test them to make sure they really work . . . Evaluate hundreds of reader’s contributions, select the most interesting g ones - and test and (where necessary) improve them. Squeeze ‘more than 100’ of these circuit ideas into a manageable num- ber of pages. The result is our traditional July/August double issue. m For those readers of this issue who don’t (yet) know Elektor, it should be stressed that this is not a ‘normal’ issue. We only do W this sort of thing once a year! Nor is this issue a review of cir- a cuits already published, nor is it a preview of circuits that wf will be discussed in greater detail in the coming year. | J And now its ‘over to you’. We hope you will enjoy studying this collection WT and building* some (if not all!) of the circuits. Maybe even dreaming up new ones for next year’s ‘Summer Circuits’ issue? Happy soldering! W * Availability of components shouldn't be much of a problem: see page 78! ess 0 m+i K+l K+i E+ E+ m J \ 1 long interval timer 12V •see text The drawback of most analogue timers (monostable circuits) is that, in order to obtain reasonably long intervals, the RC time constant must be correspondingly large. This invariably means resistor values in excess of 1 M12, which can give timing errors due to stray leakage resistance in the circuit, or large electrolytic capacitors, which again can introduce timing errors due to their leakage resistance. The circuit given here achieves timing intervals up to 100 times longer than those obtainable with standard circuits. It does this by reducing the charging current of the capacitor by a factor of 1 00, thus increasing the charging time, without the need for high value charging resistors. The circuit operates as follows: when the start/reset button is pressed Cl is discharged and the output of IC1, which is connected as a voltage follower, is at zero volts. The inverting input of comparator IC2 is at a lower potential than the non-inverting input, so the output of IC2 goes high. The voltage across R4 is approximately 120 mV, so Cl charges through R2 at a current of around 120 nA, which is 100 times lower than could be achieved if R2 were connected direct to positive supply. Of course, if Cl were charged from a constant 1 20 mV it would quickly reach thiss voltage and would cease to charge. However, the bottom end of R4 is returned to the output of IC1, and as the voltage across Cl rises so does the output voltage and hence the charging voltage applied to R2. When the output voltage has risen to about 7.5 volts it will exceed the voltage set on the non-inverting input of IC2 by R6 and R7, and the output of IC2 will go low. A small amount of positive feedback provided by R8 prevents any noise present on the output of IC1 from being amplified by IC2 as it passes through the trigger point, as this could otherwise give rise to spurious outpi : pulses. The timing interval is given by the equation: T=R 2 C! (1 + |^ + |- s )ln(l+|?) k 4 k 2 k 6 This may seem a little complicated, but with the component values given the interval is 100 • Cl, where Cl is in microfarads, e.g. if Cl is 1 n the interval is 100 seconds. It is evident from the equation that the timing interval can be varied linearly by replacing R2 with a 1 M potentiometer, or logarith- mically by replacing R6 and R7 with, say, a 10 k potentiometer. '■C. % 1 voltage- frequency converter Using only a few components and an inte- grated switching circuit it is possible to construct a high-performance voltage- frequency converter. With the component values shown in the diagram, the conversion ratio has a linearity of approx. 1%. An input voltage from 0 V . . . 1 0 V will produce a corresponding 0 . . . 10 kHz squarewave output voltage. By means of potentiometer PI, the circuit can be adjusted so that an input voltage of 0 V will produce an output frequency of 0 Hz. The components which determine the fre- quency are resistors R2, R3, R5, PI and capacitor C2. Using the formulae shown in the diagram, the conversion ratio of the circuit can be altered so that the circuit can be used for a number of different applications. When calculating the product of T = 1 .1 R 3 C 2 care should be taken to ensure that this is always less than half the minimum output period, i.e. the positive output pulse should always be at least as long as the negative pulse. fo 0.486- (R5 + PI) . [ kHz ] Ujn R2 • R3 • C2 L V J T-1,1 • R3- C2 elektor july/august 1977 RA YTHEON product specifications a i capacitance- frequency converter G. Geiger This simple circuit will produce a pulse train whose (average) frequency is directly pro- portional to the value of a capacitor which is to be tested. This waveform can thus be fed direct to a frequency counter which will then indicate the capacitance value in (tens of) picofarads. The circuit is based on a monostable multi- vibrator whose pulse width is given by: t = C x • R x * In 2. If R x is fixed then t is obviously proportional to C x . The output pulse of the monostable is used to gate pulses from a stable oscillator through to the input of a frequency counter. The number of pulses allowed through to be counted is thus proportional to t which is proportional to C x . The reference oscillator uses a normal '27 MHz (model control band) third overtone crystal, but in this circuit it oscillates at its fundamental of around 9 MHz. The monostable IC1 is triggered by gate pulses from the frequency counter and allows pulses from the oscillator through N4 to the counter input. For the circuit to func- tion correctly the counter gate period must be longer than the longest period of the monostable, which with the values shown is around 20 ms. To calibrate the circuit a capacitor of accurately known value is connected across the C x terminals and the preset is adjusted until the count indicated is equal to the capacitance value in tens of picofarads, e.g. a 1 0 n capacitor should give a count of 1 000 (not 10 000!). Any 1% silver mica capacitor of greater than 1 000 pF can be used for this purpose. In order that the calibration of the circuit should not be affected by tem- perature R x b should be a good-quality multi- turn cermet trimmer. With the values shown the circuit will measure capacitance values from 1000 pF to 1 juF. The range may be extended to include higher values by reducing the value of R x . ft 1 electronic scarecrow This circuit produces a good imitation of machine-gun fire, an effect which should be sufficient to deter the most persistent of birds. The system consists of three squarewave oscillators using CMOS inverters, all running at different frequencies. Oscillator N1/N2 speaker. The circuit can also be used to produce background noises for devotees of war games. For this purpose a mixing amplifier stage, Tl, may be provided, and inputs from several different oscillators combined. The use of different values for the mixer input IC1 = N1. ,.N6 = 4049 elektor july/august 1977 gates oscillator N3/N4, which in turn gates oscillator N5/N6. The result is that this last oscillator produces intermittent bursts of pulses, which makes the sound very realistic. The output can be taken from the right- hand side of C4 and be fed to an audio amplifier and suitable (weatherproof) loud- resistors R7, R13 and R14 means that the input signals will be amplified by different amounts, thus producing a more varied sound. PI, P2 and P3 can be adjusted experimentally to give the most pleasing results. & I temperature- to-voltage converter This circuit provides a simple means of constructing an electronic thermometer that will operate over the range 0 to 24 C (32 to 75°F). The circuit produces an output of approximately 500 mV/°C, which can be read off on a voltmeter suitably calibrated in degrees. In order that the circuit should be kept simple the temperature sensing element is a negative temperature coefficient thermistor (NTC). This has the advantage that the temperature coefficient of resistance is fairly large, but unfortunately it has the disadvan- tage that the temperature coefficient is not constant and the temperature-voltage output of the circuit is thus non-linear. However, over the range 0 to 24°C the linearity is sufficiently good for a simple thermometer. Op-amp IC1 is connected as a differential amplifier whose inputs are fed from a bridge circuit consisting of R1 to R4. Rl, R2, R3 and PI form the fixed arms of the bridge, while R4 forms the variable arm. The voltage at the junction of Rl and R2 is about 3.4 volts. With the NTC at 0 C PI is adjusted so that the output from the op-amp is zero, when the voltage at the junction of R3 and R4 will also be 3.4 V. With increasing temperature the resistance of the NTC decreases and the voltage across it falls, so the output of the op-amp increases. If the output is not exactly 0.5 V/°C then the values of R8 and R9 may be increased or decreased accordingly, but they should both be the same value. The IC can be a general purpose op-amp such as a 741, 3130 or 3140. The compen- sation capacitor C2 is not required if a 741 is used since this IC is internally compensated, i Almost any 10 k NTC thermistor may be used for R4, but the smaller types wilNj obviously give a faster response since they have a lower thermal inertia. 5 k or 15k types could also be used, but the values of PI and R3 would have to be altered in proportion. hours. If a shorter cycle time is required it is necessary that the counters be reset when the required count is reached. As an example suppose that the desired cycle time is 24 hours. The counter must therefore coun,s~' up to 24 x 60 x 60 x 50 = 4320000, whici in binary is 1000001 1 1 10101 100000000> Where a 1 occurs in this number the corre- sponding counter output is connected to one of the inputs of the diode AND gate D6 to D 1 3. When the desired count is reached these outputs will all be high simultaneously and monostable N1/N7 will be triggered, giving multi- purpose time switch Using two CMOS counters it is a simple matter to construct a versatile time switch. The total cycle time of the switch can be set between zero and 93.2 hours, and the time switch can be made to switch equipment on and off at any time during this cycle. The reference frequency for the timer is the 50 Hz mains frequency. Two 4040 counters are connected in cascade and count the 50 Hz pulses. Each of these ICs is a 1 2-bit counter, so the maximum time that the counters will count to is 0.02 x 2 s4 seconds, where 0.02 seconds is the period of the mains waveform. This is equal to 93.206 elektor july/august 1977 the counter a reset pulse. A manual reset button is also provided. Any other desired cycle time up to the previously mentioned maximum may also be accommo- dated, but obviously some counts will require more or less diodes in the AND gate. The switch-on and switch-off times of the equipment to be controlled are also deter- mined in the same manner. The binary equivalents of the on and off times are calculated and the appropriate counter out- puts are connected to AND gate inputs B1 to B4 for switch-on and Cl to C4 for switch- off. At switch-on monostable N2/N5 is triggered, which sets flip-flop FF1 , turning on T1 to activate the relay. At switch-off monostable N3/N6 is triggered, which resets FF1. Manual controls are also provided. If several circuits are to be controlled with different switch-on and switch-off times then N2, N3, N5, N6, FF1 and T1 may be duplicated. The one disadvantage of this circuit is that initially it must be reset at the time that the timing cycle is required to start, i.e. there is no time-setting facility, so in the event of a power failure it would be necessary to wait until the correct start time before resetting the circuit. For this reason it is best to make the start of the timing sequence occur at a convenient moment, such as in the morning or early evening. To make the clock input of the counter less susceptible to interference pulses on the mains waveform it may be a good idea to precede it by a Schmitt-trigger using two CMOS NAND gates as described elsewhere in this issue. B I super zener S I thermostat elektor july/august 1977 This circuit is intended primarily to produce a stable reference voltage in battery operated equipment designed for minimum current consumption. Despite the fact that only 1 mA flows through the zener the output voltage showed a fluctuation of less than 1 mV for supply voltage variations of 10 to 30 volts. The reference voltage from the zener is applied to the non-inverting input of a 741 op-amp, and the output voltage is the zener voltage multiplied by the op-amp gain i.e. V 0 = V z x — - *3 This approach has two advantages. Firstly, a low temperature coefficient zener (5.6 V) can be used to provide any desired reference voltage simply by altering the op-amp gain. Secondly, since no significant current is ‘robbed’ from the zener by the op-amp input, the zener need only be fed by a small current. So that the resistance of the zener does not affect the output voltage the zener current must be fairly constant. This is achieved by feeding the zener via R1 from the output of the op-amp. The zener current i s - ° ^ z , so R1 should be chosen to give a K-l zener current of about 1 mA. The reference voltage obtained from the op-amp output can supply currents of up to 15 mA. One point to note when using this circuit is that the supply voltage must be at least 2 V greater than the output voltage of the circuit. 15V©- 2o°c=ov r 30° C= 1 n j R7n J R6 IjJ w ( u- 1 R4 4 2 R3 — Q— | lOOkT-l " 20>L_ GH b L l A1...A3= 3 4 324 Although primarily designed to keep the water in an aquarium at a constant tempera- ture, this circuit is also suitable for a number of other applications. The circuit described here represents only the control section of the thermostat. In addition a temperature sensor and a triac relay, which at periodic intervals supplies the heating element with voltage, are necessary to complete the thermostat proper. A simple temperature sensor is provided by the NTC sensor described elsewhere in this issue, or by the temperature-voltage converter in Elektor 5, July/ August 1975. A suitable triac circuit which triggers at the zero-crossing point (i.e. when the load voltage and current are small, thus preventing interference and contact wear) is the solid state triac relay described in Elektor 1 1, March 1976, or, the ‘solid state relay’ described elsewhere in this issue. The thermostat functions as follows: the water temperature is measured by the sensor (the NTC or diode), which is fixed to the glass on the outside of the aquarium by insulating tape. Since only three amplifiers are needed for the control circuit of the thermostat, the remaining amplifier can be used to construct the NTC sensor. The voltage supplied by this circuit is compared in A2 with the preset value of PI and the LDR, and then amplified by a factor of 10. The amplified voltage is then compared in A1 with the triangular voltage produced in A3. The result is a squarewave output volt- age, which triggers the triac circuit for longer or shorter periods. The desired reference temperature can be set by means of PI . The circuit also contains a second input which is sensitive to light. This has the effect of raising the reference voltage of the thermostat so that the aquarium is allowed to get warmer during the day. With the component values shown in the diagram, the increase in temperature (the size of which depends on R1 and R2) is approx. 2°C. The LDR may also be omitted if required. The 1 5 V supply is not critical, and providing that it is properly smoothed it need not be stabilised. The current consumption for the circuit is 3 mA, which rises to 6 mA when light falls upon the LDR, and to 15 mA when the LED at the output lights up. A certain amount of attention should be paid to the safety of the circuit; for this reason the NTC is placed on the outside of the tank, and the triac relay should be fitted with an opto-isolator so that there is no direct electrical connection between the input and the mains. a I UAA170 economiser iQ triac relay elektor july/august 1977 The Siemens UAA170, which has frequently been featured in this magazine, is an LED voltmeter which will light one LED in a column of sixteen according to the input voltage. The object of the circuit described here is to obtain the maximum use from a single UAA170 IC in applications where the full sixteen LED resolution is not required. For example, if a column of eight LEDs will suffice then a single UAA170 can be made to drive two such columns in response to two independent input voltages. LEDs 1 to 8 connected to the UAA170 outputs form one display channel, and LEDs 9 to 16 form the other channel. The two input voltages are switched alter- nately to the UAA170 input by an elec- tronic switch analogous to an oscilloscope beam-switch. This consists of an astable multivibrator and two inverting buffers driving two diode switches. When the output of N1 is high the voltage on input 1 is al- lowed to pass to the output, but the output of N2 is low and the anodes of D3 and D7 are held to about 0.6 V by D5, so no voltage from input 2 appears at the output. When the output of N2 is high the reverse occurs. For the circuit to work the two input volt- ages must never overlap ;e.g. if input 1 .which causes LEDs 1 to 8 to light, has a range from 0 to 5 V, then input 2 must have a range from 5 to 1 0 V. This is easily accomplished by feeding the voltage to input 2 through the level shifter described elsewhere in this issue. The only other constraint is that the input 2 voltage should not exceed about 1 1 .4 V at any time as this is the maximum that can be handled by ihe channel switch. The input impedance of each channel is 1 0 k, determined by R1 and R2, and if a higher input impedance is required then buffer stages must be used. PI adjusts the duty-cycle of the astable multivibrator and hence the relative bright- ness of the two sets of LEDs. In some cases it may be necessary to reverse D1 to obtain the range of adjustment required to make the brightness equal. It is well-known that, for minimum gen- eration of interference pulses, triacs should be triggered near the zero-crossing point of the mains waveform. There are a number of zero-voltage switch ICs in existence, but unfortunately these are expensive and often difficult to obtain. As these ICs need a number of external components in order to function, a discrete component version of the circuit need not take up much more space, and is certainly cheaper. The mains input is rectified by Dl, dropped by R1 and stabilised by D2 to provide a 24 V DC supply. The control input is con- nected via an opto-isolator to isolate the control circuits from the mains. When no control voltage is applied the LED is not lit, the phototransistor is turned off and T1 is turned on. T5 is thus turned off and the triac cannot fire. When a control signal is present the LED is lit and T1 is turned off. The voltage comparator comprising T2 to T4 now compares a portion of the mains wave- form from the potential divider R4, R5, PI with the zero volt reference at the base of T4. As the mains waveform crosses zero the com- parator turns on T5, which fires the triac. The duration of the trigger pulse can be varied up to a maximum of 1.5 ms by PI (750 /ns each side of zero-crossing) and it is essential that by the end of this time the current taken by the load is greater than the holding current of the triac, otherwise the triac may turn off when the trigger pulse ceases. This places a constraint on the mini- mum load current that the circuit will switch, which depends on the holding current of the triac used. The maximum load current is determined by the maximum current rating of the triac. With R1 = 47 k, when the circuit triggers the 24 V supply will drop to between 10 and 20 V, depending on the width of the trigger pulse, and the triac gate current will be between 18 and 40 mA. This may be in- creased by reducing the value of Rl. As R1 has virtually the full mains voltage across it, for safety and reliability it is best to make up Rl from two series resistors, and to make the wattage rating fairly generous. Even with Rl = 22 k the dissipation in Rl is only one watt, but nevertheless it is rated at 3 W for reliability. I simple transistor tester A/D converter elektor july/august 1977 This simple circuit checks the functioning and measures the current gain (hpE) of PNP or NPN bipolar transistors. It operates by feeding a known constant current into the base of the transistor and measuring the collector current. Since the collector current of a non-saturated transistor is hpE times the base current (which is known) it is a simple matter to calculate the value of hpp, and in fact the meter which measures the collector current can be calibrated directly in hpp. Since both PNP and NPN transistors must be tested, two constant current sources are required, to provide a negative base current for PNP transistors and a positive base current for NPN transistors. The voltage dropped across the LED causes a constant current to flow through the emitter resistor of the TUP and a corresponding constant collector current, which flows into the base of the NPN transistor under test. This current can be set to 10 fiA by connecting a 50 fiA meter between points B and E and adjusting PI . The lower LED and TUN constitute the negative current source. Here again, this may be set to 10 fjA by connecting a micro- ammeter between the lower points B and E, and adjusting P2. When a transistor is plugged into the appropriate socket a current of 10 /jlA will thus flow into the base and a current of hpp times this will flow through the milli- ammeter. The full-scale deflection of the milliammeter depends on the maximum hpp to be measured. Since the collector current is hpE times the base current (which is .01 mA) a reading of 1 mA corresponds to an hpE of 100, so if a 5 mA meter is to hand it can be calibrated in hpp values from 0 to ^22 (X^Kj fpNP 'V 1 4 ? w) ! 500, which should be adequate for most run- of-the-mill transistors. However, for testing ‘C’ versions of small-signal transistors, which can have gains up to 800, a 10 mA meter ! calibrated 0 to 1000 could be used, or a lower f.s.d. meter shunted to read 8 mA an d>l calibrated 0 to 800. Readers may have noticed that it is actually the emitter current of the PNP transistor that is measured, which is of course 1 + hpp times the base current. However, since few transistors have gains less than 50 the worst error introduced by this is less than 2%, which is probably less than the error of the milliammeter. Note. The use of LEDs as reference diodes for constant current sources is dealt with elsewhere in this issue. A •'i 12V This circuit will deliver a number of pulses which is directly proportional to a negative input voltage. The conversion takes place in two stages. With SI in position 1 capacitor Cl will charge until the voltage across it is V a + Vf + V x , where V a is the voltage at the slider of PI, Vf is the forward voltage of D1 and V x is the (negative) input voltage. SI is then switched to position 2, when this voltage will appear on the left-hand side of R4. C2 will charge from Cl through R4 until the emitter voltage of T1 equals Vf, + Vbe T1 . when T1 will conduct. Since T1 and T2 are connected in the familiar thyristor tetrode configuration T1 and T2 both turn hard on and C2 discharges rapidly to Vb e T1 • C2 will then recharge from Cl until T1 again conducts. Each time C2 is charged Cl loses a charge AQ equal to Vb C2. This cycle repeats until the voltage on Cl has fallen below Vb + VbeTl . when T1 will no longer conduct. If PI and P2 are adjusted so that V a + Vf = Vb + VbeTl then this obviously occurs when the voltage on Cl has fallen to V a + Vf, i.e. the portion of the voltage on Cl due to the input voltage V x has been dissipated. The number of pulses appearing at the collector of T2 is thus proportional to this input voltage. Since Cl loses a charge AQ for each pulse the total number of pulses is obviously given by Qx N = , where Q x is the charge on Cl due to V x , which is Cl V x . ThusN = ^v b xV * With the component values given the conversion factor is about 100 pulses per volt. Source: Electronic Engineering, September 1976. automatic NiCad charger H. Knote l two-TUN voltage doubler elektor july/august 1977 It is not generally appreciated that, if Nickel-Cadmium batteries are subjected to prolonged overcharging from chargers of the constant current type, their life may be considerably reduced. The charger described here overcomes this problem by charging at a constant current but switching off the charger when the terminal voltage of the battery rises, which indicates a fully-charged condition. The basic circuit described is intended to charge a single 500 mAh ‘AA’ cell at the recommended charge rate of around 50 mA, but it can easily be extended at little cost to charge more than one cell. Power for the circuit is provided by a trans- former, bridge rectifier and 5 V IC regulator. The cell is charged by a constant current source T1 which is controlled by a voltage comparator based on a TTL Schmitt trigger N 1 . While the cell is charging the terminal voltage remains at around 1.25 V, which is below the positive trigger threshold of Nl. The output of Nl is thus high, the output of N2 is low and T1 receives a base bias voltage from the potential divider R4/R5. While the cell is being charged D1 is lit. When the cell approaches the fully-charged state the terminal voltage rises to about 1.45 V, the positive trigger threshold of Nl is exceeded and the output of N2 goes high, turning off T1 . The cell ceases to charge and D1 is extinguished. As the positive trigger threshold of Nl is about 1.7 V and is subject to a certain tolerance, R3 and PI are included to adjust it to 1 .45 V. The negative trigger threshold of the Schmitt trigger is about 0.9 V, which is below the terminal voltage of even a fully- discharged cell, so connecting a discharged cell in circuit will not cause charging to begin automatically. For this reason a start button SI is included which, when pressed, takes the input of Nl low. To charge a number of cells the portion of the circuit enclosed in the dotted box must be duplicated. This has the advantage that, unlike chargers in which cells are connected in series, cells in any state of discharge may be placed on the charger and each will be individually charged to the correct level. The disadvantage is that batteries of cells cannot be charged. However, up to ten AA cells may be charged if the circuit is duplicated the appropriate number of times. This little circuit will produce a DC output that is almost twice the supply voltage. A square-wave input is required of sufficient level to turn T1 fully on and off. When T1 is conducting, C2 is charged to just under the supply voltage. When T1 is cut off, T2 starts to conduct and raises the voltage at the negative end of C2 to just under the positive supply level. This implies that the voltage at the positive end of C2 is raised to almost twice the supply voltage, so that C3 will ultimately charge to this level. The circuit is remarkably efficient: the current drawn from the main supply is only marginally greater than twice the out- put current. In the example shown here, the efficiency is approximately 90%. The value for R1 depends on the amplitude of the square-wave input: T1 will require a base current of 0.5 ... 1 mA. (RCA application note) 12V/40mA o 1 TTL voltage doubler m 1 low-noise V.H.F. aerial amplifier elektor july/august 1977 This voltage doubler can be used in circuits that have only a 5 V supply rail, where a higher voltage is required at a low current. Figure la shows the basic circuit, which uses three of the gates in a 7437 quad two-input NAND buffer IC. N1 and N2 are connected as a 20 kHz astable multivibrator, and the output of N2 drives N3, which acts as a buffer between the astable and the doubler circuit. When the output of N3 is low Cl charges through D1 and N3 to about +4.4 V. When the output of N3 goes high the voltage on the positive end of Cl is about 9 V, so Cl discharges through D2 into C2. If no current is drawn from C2 it will eventually charge to about +8.5 V. However, if any significant current is drawn the output voltage will quickly fall, as shown in figure lb. Much better regulation of the output voltage, as shown in figure 2a, can be obtained by using the push-pull circuit of figure 2b. This is driven from an identical astable to that in figure 1 b. While the output of N 1 is low and Cl is charging, the output of N2 is high ^ *•* and C2 is discharging into C3, and vice versa, f y Since C3 is being continually charged the-- . J regulation of the output voltage is much f ' improved. This simple aerial amplifier can be used to boost the level of weak FM signals. It has a gain of 22 dB, and the extremely low noise figure of 1.6 dB, so that it will not unduly degrade the signal-to-noise ratio. The amplifier consists of a single, low-noise BFT66 transistor T1 operating in common emitter configuration. Base bias is provided by a constant current source T2, which stabilises the operating point. The value of LI is nominally 6 /iH, but any r.f. choke of a similar standard value (5.6 /l/H or 6.8 uli) may be used. L2 is a home-made coil consisting of five or six turns of 0.25 mm (33 SWG) enamelled copper wire. This is wound onto a 5 mm diameter former which is then removed and the self-supporting air-cored coil is stretched to about 1 0 mm length. When constructing the preamp care should be taken to keep all component leads as short as possible to avoid stray inductance and capacitance. The circuit should be mounted in a screened metal box located as close as possible to the aerial. Use of an IC voltage regulator has the advantage of reliability and compactness. However, if such an IC is not readily obtainable, it may be replaced by a simpler circuit: a 680 12 resistor between C5 and C6, and a 12V/400mW zener diode plus a 1 0 /it/ 16V electrolytic capacitor in parallel with C5. Specification Frequency range: 1 MHz - 300 MHz. Gain: 22 dB Noise figure: 1.6 dB Input and output impedance: 60 12 Supply voltage: +12 V Supply Current: 4 mA Reference: Siemens Data on BFT66. The disadvantage of most simple speed controllers for model trains or cars is that they simply supply the motor with a fixed voltage. Consequently the speed does not remain constant, since the model slows down when climbing gradients and speeds up when going downhill. With model trains the setting of the control knob to maintain a particular speed also varies with the load that the engine is pulling. The circuit described here eliminates this problem by monitoring the motor speed and keeping it constant for a given control setting, irrespective of load. The circuit will operate with most models which use a DC permanent magnet motor. The terminal voltage of a motor consists of two components, the back e.m.f. generated by the motor and the voltage dropped across the armature resistance. The back e.m.f. is proportional to the motor speed and so motor speed can be sensed by measuring it, but the problem is to separate the back e.m.f. from the resistance voltage. If an external resistor is connected in series with the motor then, since the same current flows through it and through the armature resistance, the voltage drop across the series resistor will be proportional to the drop across the armature resistance. In fact if the two resistances are equal then the two voltages will be equal, and the voltage across the series resistor can be subtracted from the motor voltage, leaving only the back e.m.f. The circuit monitors the back e.m.f. and adjusts the motor current so that, for a given control setting, the back e.m.f., and hence the motor speed, remains constant. To simplify the description of the circuit it is assumed that P2 is set to its mid-position and that R3 is equal to the armature resistance of the motor. The motor voltage is the sum of the back e.m.f. V a and the voltage dropped across the internal resistance V r . Since a voltage V r is dropped across R3 the output voltage V Q equals V a + 2 V r . The voltage at the inverting input of IC1 is V a + V r , and that at the non-inverting input is Vj + + These two voltages are equal, i.e. V a + V r = Vi + (V a + 2V r - Vj) 2 Simplifying this equation gives V a = Vj, which means that the back e.m.f. is always kept equal to the control voltage, so the motor runs at constant speed for a given setting of the speed control PI . P2 is used to compensate for the fact that R3 may not be equal to the armature resistance, by varying the amount of positive feedback to the non- inverting input. To set up the circuit a model is run and P2 is adjusted until the speed just remains constant on gradients and with different loads. If P2 is turned too far towards PI then the model will slow down, but if P2 is turned too far in the opposite direction then the model will actually go faster when climbing a gradient. If the controller is to be used with several different models then they must obviously all be fitted with similar motors, otherwise the circuit would require readjustment whenever a different model was used. The output transistor T1 should be fitted with a heatsink of around 4 C/watt. The signal-to-noise ratio of an FM broadcast received in stereo is considerably worse than that of the same broadcast received in mono. This is most noticeable on weak trans- missions, when switching over from stereo to mono will considerably reduce the noise level. This noise reduction occurs because the left-channel noise is largely in anti-phase to the right-channel noise. Switching to mono sums the two channels and the anti- phase noise signals cancel. By summing only the high-frequency com- ponents of the signal it is possible to elim- inate the annoying high-frequency noise without destroying the stereo image since channel separation is still maintained at middle and low frequencies. Each channel of the circuit consists of a pair of emitter followers in cascade, with highpass filters comprising R3 to R7 and C3 to C5 that allow crosstalk to occur between the two channels above about 8 kHz when switch SI is closed. When SI is open the two channels are isolated, but resistors R9 to R1 1 maintain a DC level on C3 to C5 so that switching clicks do not occur when SI is closed. The stereo-mono crossover frequency can be increased by lowering the values of C3 to C5 or decreased by raising them. tremolo toft thermal touch switch elektor july/august 1977 Tremolo is one of the most popular effects used in electronic music. The tremolo effect is produced by amplitude modulating the music signal with a low-frequency signal of between 1 Hz and 10 Hz. The effect gives warmth and richness to the otherwise ‘flat’ sound of instruments such as electronic organs. The most pleasing effect is produced when the modulation waveform is sinusoidal. The music signal is buffered by emitter- follower T1 and is then fed into op-amp IC1 , whose gain can be varied by means of P 1 . The output of IC1 is fed to the diode modulator D1/D2, the output of which is buffered by a second emitter follower T2. The sinusoidal signal is generated' by an oscillator built around IC2, whose frequency can be varied between 1 Hz and 10 Hz by P2. The output level, and hence the modulation depth, can be varied by P3. Switch SI, when closed, disables the oscillator, which allows the music signal to pass unmodulated. 2&-A This ingenious touch switch is operated, not by skin resistance or capacitance effects, but by heat. The circuit utilises the negative temperature coefficient of silicon diodes. The 741 functions as a comparator with positive feedback and, for the purpose of describing the circuit it is assumed that the voltage drop across each diode is initially the same and that the op-amp output voltage is low. Due to the potentiometer effect of R4 and R5 the voltage on the non-inverting input of the op-amp will be pulled slightly lower than that on the inverting input. If D 1 is heated by touching it, its forward voltage drop will decrease, and so will the voltage on the inverting input of the op-amp. When this voltage falls below that on the non-inverting input the op-amp output will go high. Positive feedback via R5 and R4 will pull the non-inverting input still higher, so that even if D 1 is subsequently allowed to cool to the same temperature as D2 the output will remain high. The switch is returned to its original state by touching D2. This will cause the voltage at the non-inverting input to fall, and when it is less than that on the inverting input the output of the op-amp will go low. The output of the switch can be used to control other circuits, relays etc. To indicate the state of the switch two LEDs may be driven by the circuit. If a brighter display is required than can be driven direct from the op-amp output then a two transistor buffer may be used to drive LEDs at a higher current. Before use the circuit must be nulled to com- pensate for differences in the diode forward voltage drops and for the offset voltage of the 741. This is done by closing switch SI, which inhibits the positive feedback loop, and by adjusting PI until the op-amp output voltage is approximately half supply (4.5 V). The circuit may also be used as a differential thermostat in such applications as solar heating. In this case one diode is mounted on the solar panel and the other on the hot water tank. PI is adjusted so that the circuit will operate when the required temperature difference between the two is reached. This is usually when the solar panel is 25 to 30 C hotter than the tank. rcfr 1 signal powered dynamic compressor timebase input circuit A.M. Bosschaert elektor july/august 1977 This dynamic range compressor will provide approximately 20 dB of compression over the input voltage range 100 mV to 10 V. An unusual feature of the circuit is that it re- quires no power supply, the control voltage for the voltage-controlled attenuator being derived from the input signal. A portion of the input signal is rectified by D1 and D2 and used to charge capacitors Cl and C2. These provide a control voltage to the diode attenuator comprising R3, R5, R6, D3 and D4. The diodes operate on the non- linear portion of their forward conduction curve. At low input signal levels the output signal appears with little attenuation. As the signal level increases so does the rectified voltage on Cl and C2. The control current through the diodes increases and their dynamic resistance decreases, thus atten- uating the output signal. The attack time of the compressor is fixed and depends on the time constant consisting of Cl or C2, R2 and the output impedance of the circuit feeding the compressor, which should be as low as possible. The decay time of the compressor can be varied to a small extent by PI. The input impedance of the circuit that the compressor output feeds should be as high as possible. The circuit works best with germanium diodes, since these have a low forward volt- age threshold and a much smoother and more extended ‘knee’ than silicon types. The accompanying graph shows the response of the compressor using both silicon and germanium diodes, and it is obvious which are better! 2 Uou. Mains frequency is often used as a reference for digital clocks and other digital circuits. Unfortunately problems can arise if the ‘raw’ mains waveform is used without further processing, especially when driving clocks or other timing circuits, as mains transients can cause extra, unwanted counts to occur. The circuit described here will accept a half- or full-wave rectified mains input and provide a 50 or 100 Hz square- wave output, free from interference and suitable for driving CMOS or TTL logic circuits. There are two methods of producing a clean squarewave from a noisy sinusoidal signal. The first approach is to use a Schmitt trigger with a large degree of hysteresis. Once the positive trigger threshold has been exceeded the output will go high and any small negative transients cannot then affect the output state until the input signal has fallen to near the negative trigger level. When the output has returned to the low state small positive transients cannot affect the output until the signal has again risen to near the positive trigger level. The second approach is to use the input signal to trigger a monostable whose pulse length is just less than the period of the input signal. Once the monostable has been triggered any spurious pulses cannot affect the output state until the monostable resets, by which time the monostable is due to be triggered again anyway. The circuit described here combines both these methods in a ‘belt-and-braces’ approach. The 50 Hz signal, taken from the mains transformer secondary of the equip- ment, is either half-wave or full-wave recti- fied to give a 50 Hz or 100 Hz input signal. This is reduced to a level suitable for driving the circuit by a potential divider compris- ing PI and Rl. Two CMOS inverters N1 and N2 are cascaded with positive feedback from output to input via R4 to give a Schmitt trigger. The degree of hysteresis is deter- mined by the ratio of R4 to R2. However, AC positive feedback is also provided via Cl so that, for a short time after the output of N2 has changed state, Cl will be charging or discharging and will present a low impedance across R4. This greatly increases the hysteresis and effectively causes the circuit to latch up so that it cannot be triggered by spurious pulses. A squarewave output is available at the output of N2 and an inverted version of this waveform is available at the output of N3. To set up the circuit PI is adjusted until the circuit triggers reliably. If possible PI should be adjusted until the output is a square- wave with a 50% duty cycle. If the circuit is to be used to drive CMOS logic then either a 4049 or 4069 may be used for N1 to N3, and the supply voltage should be the same as that of the circuits which are being driven (+3 to +15 V). If TTL circuits are being driven then the supply voltage must obvi- ously be the same as the TTL supply (+5 V) and a 4049 should be used for N1 to N3. For reliable triggering the peak value of the input signal should be equal to or greater than the supply voltage. inverter 0....10 V supply J. Borg man elektor july/august 1977 This inverter circuit can be used to power electric razors, stroboscopes and flash tubes, and small fluorescent lamps from a 1 2 volt car battery. In contrast to the usual feed- back oscillator type of inverter, the oscil- lator of this inverter is separate from the output stage, which allows easy adjustment of the oscillator frequency to suit different applications. The oscillator circuit consists of a 555 timer connected as an astable multivibrator. The inclusion of D1 ensures that the duty-cycle of the squarewave output is maintained at about 50%. The output of the 555 drives the base of T1 which switches current through one half of the primary of the transformer. T2 is driven from the collector of T 1 and thus switches current through the other half of the transformer winding on opposite half cycles of the drive waveform. Zener diodes D4 and D5 protect T1 and T2 from any high-voltage spikes generated by the transformer. The voltage applied to the transformer primary is stepped up and the required high output voltage appears across the secondary winding. Depending on the application the secondary voltage may or may not be rectified. Components The transformer is a standard mains trans- former with two identical secondary windings or a single, centre-tapped second- ary. This transformer is, of course, driven in reverse, i.e. the secondary becomes the primary and the output is obtained from the primary (which is now the secondary). It must be borne in mind that, since the inverter produces a squarewave output, the RMS secondary voltage and peak secondary voltage are identical. This affects the choice of transformer for different applications. The required secondary voltage of the mains U m is the normal mains primary volt- age of the transformer. Up is the desired peak secondary volt- age. s s. An electric razor requires 240 V* RMS = 240 V* peak, so if a transformer with a 240 V primary is used the secondary windings should each be 1 2 V or a single 1 2-0-12 wind- ing. For vibrator type (non-rotary) razors the oscillator frequency should be 50-60 Hz, so the value of Cl should be 330 n and PI should be adjusted accordingly. Rotary razors are less critical of mains frequency. When operated from the normal mains supply, fluorescent lamps receive a peak supply voltage of around 340 V, which enables them to strike reliably. The trans- former secondary voltage should be calcu- lated with this in mind, which means that secondary voltages of eight or nine volts will be suitable. Fluorescent lamps can be operated with* » improved efficiency at frequencies greater-^ than 50 Hz, and the transformer will also be more efficient. Choosing a value of 56 n for Cl the oscillator frequency may be set to around 250 Hz. At frequencies much higher than this iron losses make the trans- former less efficient. The current rating of the transformer depends upon the load. For electric razors and small fluorescent tubes up to 8 W, 500 mA secondaries will be adequate. Higher output powers may be obtained by choosing a suitable transformer, replacing T1 and T2 by higher power types and reducing the value of R3 and R4 (minimum 1 20 £2). To power strobes and flash tubes the output ^ must be rectified and used to charge a reservoir capacitor, which should be of a type rated for high discharge currents. The bridge rectifier should be rated to suit the peak output voltage. transformer is given by U s = where 1 2 V is the inverter supply voltage * U.K. only. Overseas readers substitute the appropriate local mains voltage. The 723 is an extremely useful voltage stabiliser. It can be used for a wide range of supply voltages, with one limitation: the output voltage cannot be set at less than 2 V. This limit is set by the built-in differential amplifier. This problem can be solved by adding an ‘outboard’ differential amplifier, such as the 3130. in the circuit given nere tne / z 5 is usea as a fixed-voltage regulator, providing a 14V supply for the 3130. At the same time, the reference voltage output of the 723 is fed (via PI) to P2 to provide a variable reference input to the 3130. The gain of this opamp is set (by means of R7 and R8) at x2, so its output voltage will be twice the voltage at the slider of P2. The opamp output is buffered by Tl, so that the supply can deliver up to 300 mA at any output voltage from 0 ... 10 V. Tl should, of course, be provided with a# adequate heatsink. The current-limiting circuit in the 723 is also used. Its current-sense input is connec- ted across R6; the value of this resistor sets the maximum output current. A 2.2 resistor will limit the output to 300 mA; using a 6.8 12 resistor will bring the maxi- mum output down to 100 mA. The unregulated DC input to the circuit should be 20 V. This can be derived from a EPS. 7 70 5 9 ®i ♦ 0 OO potentiometer here; a multiturn type will prove useful if precise and stable setting of the output voltage is required. Alternatively, the circuit shown in figure 2 can be used. In this case, the output voltage can be altered in 1 V steps by means of the 10-position switch, and P3 allows fine-adjustment within each 1 V step. . ^ * N T1 ^ JL ABD137/TIP31 V ' R5 IC2 3130 0. 0...10V R3Q R8 \7H 10n 6V 77059 1 see text 77059 2 30 spot- frequency sinewave generator This circuit employs an unusual method of producing a sinewave signal and, unlike most sinewave generators, requires no amplitude stabilising components such as thermistors or FETs. N1 and N2 are connected as a Schmitt trigger at whose output appears a squarewave (how this happens will become apparent). The squarewave signal is fed into two cascaded selective filters consisting of IC2/T1, IC3/T2 and their associated com- ponents. The filters remove the harmonic content of the squarewave leaving only the sinusoidal fundamental. This signal is fed back via Cl to the input of the Schmitt trigger. At each zero-crossing of the sinewave the Schmitt trigger changes state, thus producing the original squarewave that is fed to the input of IC2. PI adjusts the trigger point of the Schmitt trigger, which varies the duty-cycle of the squarewave and hence the sinewave purity. By suitable adjustment of PI distortion levels of 0.15% to 0.2% can be achieved. With the ICs used I lower figures are not possible due to the J distortion introduced by IC3 and T2. N3 and N4 also function as a Schmitt trigger, which further speeds up the leading and trailing edges of the squarewave from N2. A squarewave signal with short rise and fall times, synchronised to the sinewave signal, is available at the outputs of N5 and N6. The value of C for a particular frequency f 0 is , 0.34 given by C (fiF) = — - . 2 & compensated light sensor elektor july/august 1977 Circuits which use photosensitive devices such as LDRs, photodiodes or photo- transistors to sense rapid changes in light level (e.g. in optical tachometers) can be affected by changes in ambient light level altering the biasing of the photosensor. If the bias current is fixed this can lead to saturation of the photodevice at high ambient light levels. This problem can be overcome by using a light-dependent current source that will alter the bias current to cope with changes in ambient light level and thus maintain a constant bias voltage across the photosensor. However, if the response time of the current source is made sufficiently long it will not respond to the rapid variations in light level such as those which the circuit is required to sense, and such variations will produce changes in output voltage. In addition, since the circuit is a current source it has an extremely high output impedance and thus does not load the photosensor output. The circuit comprises a current source T 1 , which is driven by the output of op-amp IC 1 . The inverting input of IC1 is biased to about 5 V by R6 and R7 and feedback causes the output voltage of the IC to adjust itself until the current supplied by T1 is such that the 1 voltage dropped across the photosensor is equal to the voltage at the inverting input of the op-amp (about 5 V). This ensures that if the ambient light level (and hence the resistance of the photosensor) varies, the current through the photosensor will be adjusted to maintain a constant output volt- age. However, due to the time constants R4 • C2 and R5 • C3 the circuit cannot respond to variations in light level that occur at fre- quencies greater than about 2 Hz. The current source will thus ‘ignore’ these vari- ations and will continue to supply a constant current, so that if the resistance of the photo- sensor varies due to rapid changes in light level the output voltage will vary in sym- pathy. The circuit will function with a wide range of LDRs, phototransistors^and photodiodes, the only constraint being that the resistance of the device connected between points A and B must not fall below 300 £2. The total current consumption of the circuit is between 20 and 35 mA, depending on the current drawn by the photosensor. flfl 1 threshold box Using a single LM324 quad op-amp IC a versatile threshold comparator may be constructed. The circuit will sound an alarm and/or energise a relay whenever the input voltage falls outside preset levels. The circuit may be adjusted to respond to the three following conditions. a) Input voltage falls below a preset level. b) Input voltage rises above a preset level. c) Input voltage falls inside a range of volt- ages defined by upper and lower limits (window comparator). Op-amps A1 and A2 function as compara- tors. When the voltage on the inverting input of A1 falls below the preset reference voltage on the non-inverting input the output voltage will go high. When the volt- age on the non-inverting input of A2 rises above the preset voltage on the inverting input the output will go high. For use as a window comparator inputs 1 and 2 should be linked. When one or both outputs go high T3 will be turned on, energising the relay. The astable multivibrator built around A3 will begin to oscillate and will feed a train of short pulses at about 1.5 kHz to the amplifier comprising T1 and T2, thus sounding the alarm. A4 provides a stable 6 volt reference for the comparators. Coarse adjustment of the reference voltages is provided by P2 and P5 and fine adjustment by PI and P4. Hysteresis is provided, which is variable by means of P3 and P6. This is particularly useful if the input signals are noisy. The relay should have an operating voltage of 9 V or less, but if it is less then a suitable resistor must be included in series with the relay to drop the excess voltage. The maximum input voltage to inputs 1 and 2 is 25 V and the input resistance is 1 M£2, but the input voltage (and resistance) can be increased by raising the value of R2 and R7. sawtooth- CCO elektor july/august 1977 This sawtooth waveform generator which is , built round a current controlled oscillator is distinguished by its large sweep range. It is suitable for use in electronic music appli- cations, and the narrow output pulse also enables the circuit to be used as a pulse- CCO for sample/hold circuits. The circuit consists of a controllable cur- rent source (Tl, T2), a trigger (Nl, N2) and a switch (T3). As soon as the circuit is switched on capacitor Cl is charged by the current source. When the voltage across Cl reaches the threshold value of Nl, T3 is turned on via Nl and N2, and Cl is dis- charged, after which the whole cycle repeats itself. The sawtooth output signal which is buffered by FET T4 has a peak-to-peak value of approx. 1.3 V. With the component values shown in the diagram, the frequency can be adjusted from approx. 5 to 500 kHz (with PI). Although a higher output frequency is possible, there is a corresponding deterio- ration in the waveform. With Cl = 5n6 and R1 = 1 k, the frequency range runs from 0.5 .. . 500 kHz. In place of NANDs inverters may be used. 1 speed controller for motors M. Junghans Electric motors up to around a quarter horsepower can often be picked up very cheaply (ex-domestic appliances). They are extremely useful for driving bench saws, drilling machines etc. A speed controller that will set and maintain a near-constant speed under varying load conditions is a useful accessory in this type of application. The motor speed is sensed by an optical commutator consisting of a disc with 15 holes or slots around its periphery. This interrupts a light beam falling on a photo- transistor, which turns T1 on and off. Pulses from the collector of T1 are used to trigger a monostable IC1 whose Q output is inte- grated by R6/C4 and R7/C5 to provide a DC output level which is inversely proportional to motor speed. The desired motor speed is set by PI and the 770W V elektor july/august 1977 desired and actual motor speeds are com- pared by T2. T3 and T4 form a trigger pulse generator which produces a pulse to fire the triac once every half cycle of the mains waveform. T2 varies the collector current of T3 and hence the charging current of C6. This in turn controls the point in the cycle at which the trigger pulse occurs. If the motor speed should tend to rise then the voltage on C4 will fall and the base current of T2 will be reduced. The collector voltage of T2 will rise, thus reducing the current through T3, and the trigger pulse will occur later, causing the motor to slow down. If the motor tends to slow down the reverse will occur. The voltage on C4 will rise and the collector voltage of T2 will fall, thus increasing the collector current of T3 and causing the trigger pulse to occur earlier. The trigger stage (T2 . . . T4) is driven from an unsmoothed full-wave rectified supply (A). When T4 fires, it will discharge C6 rapidly and then remain conducting until the supply voltage approaches the next zero- crossing. The result is that C6 is always dis- charged at the start of each half-cycle, so the position of the trigger pulses relative to the zero-crossings is determined solely by the current through T3. In other words, the trigger pulses are synchronised to the mains. The type of triac required will obviously depend on the motor used. It should be rated for at least three times (!) the nominal mains voltage; the current will depend on the maximum motor current, and a reason- able rule-of-thumb is to divide the motor power rating by the mains voltage and multiply the result by two. As an example, a 500 W/245 V motor would require a triac rated at x2*4 amps! The trigger pulse transformer (Trl) can be wound on a type AL250 potcore: primary 80 turns, secondary 40 turns, both 0.1 mm enamelled copper wire (42 S.W.G.). Editorial note: In some cases it may be possible to dispense with the ‘disc-with- holes’. If the LED and phototransistor are mounted inside the motor casing, the motor itself may sometimes be used to periodically interrupt (or reflect) the light beam. I tesoj frequency divider using one TUN $v distortion suppressor This circuit is designed to deliver an output voltage, the frequency of which is half that of the input signal. Since there is no base bias for Tl, this transistor can only be turned on during a positive period of the input signal. The basic circuit is derived from a Colpitts oscillator, and the tank circuit (LI, C4, C5) is tuned to 16.5 kHz. When driven by an input signal in the 30 ... 60 kHz range, it will ‘lock on’ and produce an output signal in the 1 5 ... 30 kHz range. The section of the circuit round D1 and C3 is included to prevent the circuit oscillating spontaneously. C3 ensures that, when the input signal disappears, the base of Tl is biased slightly negative with respect to the emitter, causing Tl to turn off. A remarkable feature of the circuit is that amplitude modulation of the input signal by up to 70% will be reproduced at the output. The usual method of measuring the har- — 18 dB/octave, the total slope thus being » monic distortion of an audio amplifier is to -36 dB/octave. Each stage consists of a pair feed it with an extremely pure sine wave of transistors connected as a highly linear signal and to measure the harmonic products ‘super emitter-follower’. The distortion introduced by the amplifier. Unfortunately, introduced by these transistors is negligible, few commercially available signal generators and the use of T2 and T5 as constant current have distortion figures better than 0.05%, emitter loads makes the distortion still and since modern audio amplifiers better lower. Each stage is essentially identical, but this figure most of the distortion measured to simplify DC biassing the first stage is will be that produced by the generator. built with an NPN/PNP transistor pair while The solution is to filter out the harmonic the second stage uses a PNP/NPN pair, distortion of the filter, leaving only the pure With the component values shown the cutoff fundamental. This can be done using a (-3 dB) frequency of each stage is 1 kHz, so highly selective notch filter that allows only if the input signal frequency is 1 kHz the the fundamental to pass and attenuates the total attenuation of the fundamental will be harmonics. Unfortunately, to provide good 6 dB i.e. it will be halved. However the attenuation of the second harmonic the second harmonic is attenuated by a factor Q-factor of the filter must be extremely of 20 with respect to the attenuated funda- high, which means that the notch is very mental, and the third and higher harmonics narrow. Any frequency drift in the signal are attenuated still more, will then cause the fundamental to be When the circuit was tested with a 6 V p-p seriously attenuated. input at 1 kHz from a generator having This problem can be avoided by using a 0.08% total harmonic distortion the output lowpass filter with a steep slope. Provided voltage was found to be 3 V p-p, with a total the oscillator frequency remains around the harmonic distortion of only 0.002%! cutoff frequency of the filter the funda- The values of Ca, Cb and Cc for a 1 kHz mental will suffer little attenuation, but the cutoff frequency are 22 n, 56 n and 3n9 harmonics will be severely attenuated. respectively. Other values can easily be The circuit consists of two cascaded filter calculated from the equations given with elektor july/august 1977 stages each having an ultimate slope of the diagram. $ knotted handkerchief B3 I automatic car aerial U. Behrendt elektor juiy/august 1977 Since the advent of paper handkerchiefs, that time-honoured method of jogging a forgetful memory, namely tying a knot in one’s hanky, has been faced with practical difficulties. The circuit described here offers a modern answer to this old problem, i.e. an electronic ‘knot’ in the shape of an audible alarm signal which can be set to sound after an interval of up to 60 minutes. The circuit is built round the CMOS IC CD 4060, which consists of a pulse gener- ator and a counter. When switch SI is closed a reset pulse is fed to the IC via C2. At the same time the internal oscillator begins feeding pulses to the counter. After 2 13 pulses the counter output (Q14) will go high, switching on the oscillator round T1 and T2. The result is a piercing 3 kHz signal which is made audible via an 8 ohm miniature loud- speaker or earpiece insert. The circuit is switched off by opening S 1 . With the values given for R2 and Cl, the ‘knot’ will sound approx. 1 hour after the circuit has been switched on. By replacing R2 with a 1 M variable potentiometer, the alarm interval can be varied between 5 minutes and 2% hours, and the potentio- meter suitably calibrated. The circuit consumes very little current I (0.2 mA whilst the counter is running and I 35 mA during the alarm signal) so that a j 9 V battery would be assured of a long life. V 'see text to one side until the aerial is fully extended. To lower the aerial the switch is held over to the other side until the aerial is fully retrac- ted. It is quite easy to forget to lower the aerial when leaving the car, thus losing the vandal-resistant advantage of a motorised aerial. The circuit described here will raise the aerial automatically when the car radio is switched on and lower it when the radio is switched off. SI can be the special switch contact provided for this purpose in some car radios, or an extra lead may be taken from the normal on-off switch, since little extra current is drawn through this contact. T3 is normally turned on. When the radio is switched on (SI closed) T3 is turned off. Current flows from SI, charging up Cl through R4, PI and the base of Tl. T1 turns on, energising Rel and causing the aerial to extend. The time for which the aerial motor runs can be adjusted to the correct value by PI. When the radio is turned off T3 turns on and C2 charges through T3, R5, P2 and the base of T2. T2 turns on, Re2 is energised and the aerial retracts. The time can again be adjusted (by P2). I » wheel of fortune • io • 9 • on off s 5H a 7 • 6 START • a <§> • A • 3 • e * This circuit represents a simple ‘wheel of fortune’ or ten-sided die suitable for many games applications. The wheel is ‘spun’ by pressing button SI, which causes pin 1 of the 7413 to go high. This starts the oscillator with the 7413-gate, and square wave pulses are fed to pin 14 of the four-bit counter IC2. The oscillator stops when SI is released. The counter IC2 counts the total number of pulses that the astable multivibrator has produced during the period that SI was open. A sufficiently high frequency is chosen for the oscillator to ensure that the player cannot cheat by releasing SI at a chosen moment. The four-bit information from the IC2 is fed to a BCD to decimal converter IC3. As long as the state of the counter DCBA is lower than or equal to 1001, one of the 10 outputs of IC3 is low, i.e. one of the LEDs lights up. If, for example the state of the counter is DCBA = 0110, then output 6 of IC3 will be low and D7 will light up. However the 7445 decodes only 10 of the 16 possible counter states; from 1010 ... 1 1 1 1 all 10 outputs of IC3 are high and none of the LEDs are lit. This state may be used to indicate the end of a player’s turn. IC3 may be replaced by a BCD-to-one-of- sixteen decoder, e.g. the 24-pin type 74154, along with hajf a dozen LEDs with series resistors. This will produce an electronic die with sixteen possible results. stereo pan pot This circuit offers the possibility of stereo image width control from stereo, through mono, to reverse stereo. The circuit com- prises two emitter followers and a linear stereo potentiometer. If x is the ratio of the resistance between the sliders of the pots and the lower ends of the pots to the total resistance then it follows that the outputs L' and R' are given by: L' = R (1-x) + Lx R' = Rx + L(l-x) Therefore, when x = 1, L' = L and R' = R (normal stereo); when x = Vi, L' = R' = Vi ( L + R) (mono); when x = 0, L' = R and R' = L (reverse stereo). The low output impedance of the emitter followers ensures that, when the poten- tiometer is in either the extreme clockwise or anticlockwise position, crosstalk travelling along the potentiometer tracks cannot appear at the outputs. Good channel separ- ation in the stereo and reverse stereo modes is thus maintained. 10... 30 V elektor july/august 1977 R.F. amplifier with 100 dB dynamic range and exhibits low intermodulation distortion over the range 100 kHz to 30 MHz. Incorpor- ation of the amplifier into an existing receiver obviously varies, depending on the receiver, and cannot be discussed in detail but must be left to the individual constructor. Figure 1 shows the test circuit which was used to measure the performance of the amplifier using a spectrum analyser. Two signals of equal amplitude at 5.2 MHz and 5.24 MHz were fed into the amplifier and the IM products were examined. With an input amplitude of 1.25 V p-p per input signal, photo 1 clearly shows that the IM products are at least —40 dB with respect to both output signals. The voltage gain of the amplifier is approximately four, with PI set to minimum. It was decided to incorporate a manual r.f. gain control into the amplifier, since this facility was present in the receiver which was to be modified. Varying the gain by changing the working point of the FET (as is common practice) was quickly rejected, 1 •**> ooos nw 3 ua ns 1 tmt Photo 1. The output spectrum of the amplifier with two input signals of 1.25 V pp . (PI set for maximum gain). Photo 2. The output spectrum of the amplifier with the same signals as in photo 1, but with R3 removed. Although there is a reduction in gain, there is a considerable increase in the intermodu- lation products. Photo 3. The same input signals as in photo 1, but with PI set for minimum gain. The intermodu- lation distortion is 10 dB better. The final solution was to apply a variable , amount of negative feedback to the FET source via PI. At DC, Pi is short-circuited by the low resistance of L2, so varying P 1 can have no effect on the DC bias conditions. At radio frequencies, however, L2 presents a very high impedance, and the gain can thus be varied by adjusting P2. Photo 3 shows the low (-50 dB!) intermodulation products even with PI at maximum resist- ance (minimum gain). PI can vary the gain by a factor of 30 dB, and with PI set to minimum gain input signals of up to 6 V p-p may be handled with the same low level of distortion. Figure 2 shows a typical application of the amplifier in a receiver. LI /Cl, C2 and L4/C5, C6 are the existing first and second tuned r.f. circuits of the receiver, the active part of the circuit (r.f. amplifier) being replaced by the MOSFET stage. In view of the relatively large dissipation in the MOSFET (typically 300 mW) it should be fitted with a heatsink. oooswj khz ns t i t > > t. 1 mf.M 000 5 HHZ 3 KHZ ns 104 9/ 300HZ 50k H? elektor july/august 1977 & program- mable frequency divider W. Schwinn e one-shot for immersion heater elektor july/august 1977 This circuit will divide the frequency of a TTL compatible squarewave signal by a factor from 0 to 999. The circuit comprises three decade counters IC1 to IC3, and a few NAND gates. The three 7490’s are cascaded and count the input signal. When the desired count is reached all the inputs of IC4 become high, so the output goes low. This triggers the monostable consisting of N2 and N3, which provides a short output pulse. The output of N1 goes high, resetting the three counters, and the count then begins again. To programme the counter it is first necessary to work out the binary coded decimal (BCD) equivalent of the required division ratio. Then, wherever a ‘1’ occurs in this number the corresponding output of the counter is connected to the input of IC4. In the example shown the division ratio is 283, or in BCD 00 1 0 1 000 00 1 1 , so outputs B of IC3, D of IC2, and A and B of IC1 are connected to IC4. The unused inputs of IC4 are connected to +5 V via a 1 k resistor. Note that the circuit will not divide by 777, which in BCD is 01 1 1 01110111, since this would require 9 inputs to IC4, and only 8 are available. Immersion heaters, which are used to heat water during the summer months when the central heating system is shut down, consume considerable amounts of power if left on continuously, even when the hot water tank is well-lagged. It is quite easy to switch on the immersion heater and forget about it, particularly if the switch is in some out-of-the-way place such as the airing cupboard. The circuit described here provides one-shot operation of the immer- sion heater, so that water can be heated as required. When the one-shot function is initiated the immersion heater will heat the water to the temperature determined by the tank thermostat and will then shut down. The heater will not operate again, even when the thermostat closes, until the one-shot button is again pressed. Operation of the circuit is very simple. When the circuit is off Cl is charged via the immersion heater element, the thermostat and D1 to 320 V. When S2 is pressed the thyristor Thl is triggered, and current flows through LED1. The initial surge current flowing through the LED is limited by R1 and D3. The LED is optically coupled to LDR1 , so the resistance of the LDR falls and the TriacTri is triggered every half cycle. The immersion heater current flows through Tri. Since the triac triggers at a point on the AC waveform corresponding to the diac break- down voltage, this is the maximum voltage that appears across the triac before it triggers, so the voltage on Cl falls to about 20-30 V. This gives a steady state current through the LED of about 20 mA. When the thermostat opens no further current can flow into Cl, so it discharges rapidly, Thl turns off and the LED is extinguished. The resistance of LDR1 becomes very high, so the triac can no longer trigger, even when the thermostat closes again. The circuit will not operate again until Thl is triggered by closing S2. Note. Suitable LDRs that will withstand 240 V AC are made by Heimann and are available in U.K. from Guest Distribution. The dynamic range of a receiver is largely dependent upon the characteristics of the mixer. It is therefore crucial that large signals are processed by the mixer with a minimum of distortion, and that at the same time its noise figure should be as low as possible. When the noise figure is less than 10 dB for the frequency range from 0.1 to 30 MHz, and there is a conversion gain of 6 dB, then the mixer can be used directly as the input stage of the receiver. For superhets with a fairly high intermediate frequency (e.g. 9 or 10.7 MHz) it simplifies matters if the RF- and oscillator-inputs of the mixer have a high impedance. The cir- cuit described here fulfils these requirements: Two input signals, each 2.5 Vpp, produce an output signal with a third order IMD of -45 dB. With the optimum setting for PI, Zs and Zp the conversion gain is approx. 6 dB and the noise figure approx. 4 dB. Although the circuit looks symmetrical, this is not quite the case. The input signal at G1 is present in attenuated form at both sources and this produces two drain signals in antiphase. However, these signals are not quite equal. This is equally true for the oscillator signal. This slight degree of asymmetry means that the input and oscil- lator signals will not completely cancel. The ratio Zp:Zs needs to be between 6 and 10 at the intermediate frequency. The simplest solution is to make Zs smaller than is thought necessary and to place it in series with a resistor. Although this ‘long-tailed pair’ mixer can be found in most radio design handbooks, what is not often mentioned is that the circuit will fail to function satisfactorily without Z s . This impedance is minimal at the inter- mediate frequency, and this results in narrow-band response of the mixer. This is more of an advantage than a disadvantage, since most IF-filters have a higher impedance outside the pass-band than in the pass-band, a fact which can often cause overloading at the mixer output. The tuning procedure is as follows: Zs is omitted and an RF signal which is large enough for cross-modulation to be just audible, is fed in. Pi is then adjusted to minimum cross-modulation (if necessary the RF-signal should be increased further). Zs is then added. If the component values are correct, then this should increase the value of the IF-signal without affecting the cross- modulation. Care should be taken to ensure that no IM occurs after the mixer stage! If high-frequency measuring equipment is available, the procedure can be simplified. An HF signal generator, for instance, can be used to produce any desired amount of IMD; a spectrum analyser is an invaluable aid for those perfectionists who want absolute ‘spot-on’ alignment. ftp O V voltage reference This circuit will provide an accurate 0 V reference output. The input voltage can vary between -1000 V and +1000 V without detriment to the precision of the reference output. This particular design approach is a major improvement over more conventional cir- cuits, in that it results in an exceptionally low output impedance while at the same time being completely short-circuit proof. elektor july/august 1977 MISSING LINK: Last-minute lab tests have shown that the performance can be improved still further be omitting D 1 , D2 and R 1 . u .»0 ©- -® m I frequency doubler using 4069 A.M. Bosschaert I digital capacitance meter with 555 timer J. Borgman elektor july/a ugust 1977 30 ©~ ii r ©innnnr Using a single 4069 hex inverter IC, a fre- quency doubler can be constructed to give an output pulse train whose frequency is twice that of a squarewave input signal. The signal is applied to the input of Nl. It should be a squarewave with a duty-cycle of approximately 50% at a level compatible with CMOS logic (3 - 1 5 V peak-to-peak depending on supply voltage). The input signal is buffered and inverted by Nl, and inverted again by N2, so the outputs (A and B) of Nl and N2 are squarewave signals 1 80 out-of-phase. The output of Nl is differen- tiated by Cl and R1 and the output of N2 is differentiated by C2 and R2, giving two spike waveforms (C and D) 1 80 out-of- phase. These signals are buffered, inverted and ‘squared up’ by N3 and N4 to give waveforms E and F. These are then com- bined in a NOR gate consisting of Dl, D2, R3 and N5, and finally inverted by N6 to give the output waveform G, which has a frequency twice that of the input signal. The circuit will operate over a wide fre- quency range. The upper frequency restric- tion is imposed by the fact that the width of the negative-going pulses E and F must be greater than the minimum pulse width that N3 and N4 will reliably transmit. Assuming that waveforms E and F have the minimum possible pulse width, as the frequency of the input signal increases the duty-cycle of the output signal will approach 50% as the pulses come closer together. When this situation is reached then the width of the positive output pulses is also the minimum that the 4069 will handle. With the component values shown the width of pulses E and F is about 500 ns, so the duty-cycle of the output will be 50% when the frequency is 1 MHz, i.e. when the input frequency is 500 kHz. Anyone possessing a frequency counter with a facility for period measurement can build this simple add-on unit to measure capaci- tance direct. In the circuit shown the 555 is connected as an astable multivibrator, the period of which is given by T = 0.7 (Ra + 2Rb)C x . If the un- known capacitor is connected in the C x position then, since Ra and Rb are fixed, the period is proportional to C x , the un- known capacitor. The period of the multi- vibrator can be measured by the period meter and, if Ra and Rb are suitably chosen, this reading can be made to equal the capacitance in picofarads, nanofarads or microfarads. As an example, suppose the period meter has a maximum reading of one second and this is to be the reading for a capacitance of 1 /iF. Then the total value of RA + 2 Rb should be 1.43 M£2. A slight problem exists when measuring electrolytic capacitors around the 1 juF value. As shown above, for a reading of one second the resistance required is fairly high and errors may occur due to the capacitor leakage resistance. In this case it is probably better to opt for a reading of 1 second = 1000/iF, since the resistance values can be 1 000 times smaller. If a seven decade counter is used that gives a reading of 1.000000 for one second = 1 000 [jF, then 1 [J.F will give a reading of 0.001000, which is still better than the accuracy of the circuit. Some suitable values for Ra and Rb are" given in the table. 1% metal oxide resistors should be used for these to give a reasonable accuracy. Other values can be calculated to suit personal taste and the counter used. When using the circuit the wiring capaci- tance of any jigs and fixtures used to hold the capacitor should be taken into account and subtracted from the reading. For this reason, such jigs should be of rigid mechan- ical construction so that the capacitance does not vary. For example, in the proto- type is was found that the circuit was reading consistently 36 pF high due to wiring capacitance. This was therefore noted on the test jig so that it could be subtracted from all readings. Of course, when testing large value capacitors this small error is not significant. Ra Rb Cx T 1 k 220 n 1 000 AtF 1 s 1 M 220 k ll nF (non-electrolytic) 1 s 40 add-on timer for snooze- alarm-radio- clock P.C.M. Verhoosel This circuit will enable the digital clock published in Elektor No. 20 to be used as an interval timer that will switch on an ap- pliance at a particular time and switch it off after a preset interval. The circuit requires only three connections to the existing clock circuit, to points j, m and position 1 of S5, as shown. In addition, S6 must be replaced by a single-pole change- over switch with centre off position, for reasons which will become apparent. The circuit functions as follows: when the clock is to be used as an interval timer S6 is set in the centre off position (this prevents the alarm output from activating the buzzer or the relay) and S7 is open. To set the interval timer, the start time is set (with S8 open) by turning S5 to position 3 and using SI and S2 to set the desired alarm (start) time. S5 is then returned to position 1, S8 is closed and SI and S2 are used to set the ‘radio delay’ time (max. 59 minutes), which is used as the interval time. Until the desired start time is reached, ‘alarm output’ m will be low, T1 will be turned off and input j will be held high via D1 and Rl, so the clock will display the selected interval time. When the start time is reached the alarm output will go high, turning on Tl. The radio relay output will go high, energising the relay, and the ‘radio delay’ timer will start to run. At the end of the selected interval the ‘radio delay’ output will go low and the relay will drop out. During the timing interval the clock will revert to normal time display, but the state of the interval timer can be examined by briefly setting S5 to position 4. The one disadvantage of this system is that it is not possible to read the actual time from the clock during the time the relay is not actuated, as opening S8 to revert to normal time display will also cause the relay to pull in. However, assuming that the interval timer is to be used while one is out of the house, this is not a great disadvantage. I To avoid cutting an extra hole in the clock case for S8 an alternative would be to replace S5 with a six-way switch. The sixth position of this would be connected to the top end of Rl in place of S8 and could then be labelled ‘interval timer’. TV games Wherever there is a mass market for a complex digital circuit, an LSI chip is sure to appear. It should not come as a surprise that ‘TV games’ are no exception. The chip used here, the AY-3-8500, offers six different games and includes scoring and sound effects. To simplify matters, all out- put signals are derived from a single clock frequency. Divider stages in the IC are used to ‘produce’ players, ball, boundary lines etc. The positions of the players are deter- mined by the period times of two on-board monostable multivibrators. The circuit shown here will give four differ- ent games: Pelota, Squash, Hockey and Tennis. The chip itself is capable of two further games, but since they require a special (optical) gun it was decided not to include them in this simple circuit. The corresponding ‘game select’ pins (18 and 19) and the connections to the small additional circuit required (pins 26 and 27) are brought out, however. The complete unit, including an audio amplifier (Tl), the clock oscillator (T2) and a VHF/UHF oscillator (T3) can be mounted' on a small printed circuit board (EPS 77084) The power supply should be capable of delivering 9 V/100 mA. PI and P2 are the ‘player position’ controls. Initially, Cl 2 can be set in the mid-position and the television set can be tuned through the low UHF band until the signal is located. If necessary, the signal can be moved up or down the band by readjusting Cl 2. The next step is to adjust CIO until a stable picture is obtained; finally, the contrast can be adjusted according to personal taste with P3. Note that it is not advisable to set too bright a picture, as this may eventually cause damage to the picture tube. The desired game is selected with SI; S2 is elektor july/august 1977 L2:4turns 77084 fin 0,486 (R4 h T1 = 1,1 ■ R3- C2 frequency- voltage converter elektor july/august 1977 This frequency-voltage converter is dis- tinguished by its markedly linear conversion ratio. With the given component values the conversion ratio of the circuit is 1 V/kHz. If a DC voltage is applied to the input (0 Hz) then the output voltage is 0 V. The duty cycle of the squarewave input signal has no effect upon the conversion ratio. However, if sinusoidal signals are to be converted to a DC-voltage, then the converter-IC should be preceded by a Schmitt trigger. Other conversion ratios can be calculated using the formulae shown in the diagram. The circuit can also be connected to the output of a voltage-frequency converter and used as a means of transmitting DC signals over a long cable without the cable resistance attenuating the signal. RA YTHEON product specifications aft electronic weathercock Until recently, finding out which way the wind is blowing has always necessitated putting on one’s shoes and stepping outside the door, thereby exposing oneself to the vagaries of the British climate. However with a little technical ingenuity, it is possible nowadays to know the precise direction of the wind without leaving the comfort of one’s fireside. The electronic weathercock functions by connecting the vane to a potentiometer which turns with the vane. The voltage at the slider of the potentio- meter is then proportional to the angle through which the vane is turned by the wind. The size of this voltage (and hence the direction of the wind) may be displayed in digital form using a UAA 170 and 16 LEDs. The circuit is designed so that there is a smooth interchange between the LEDs. Potentiometer PI controls the brightness of the LEDs, whilst P2 is set such that, when the voltage at the slider of P3 (which is connected to the vane) is at a maximum, then D16 lights up. Further details regarding the UAA 170 may be found in Elektor 12, April 1976. Potentiometer P3 may present a slight problem, in that it must be of a type which can be adjusted through 360 . If such a potentiometer proves difficult to find, then l one solution is to use sixteen reed relays, each of which is enabled whenever a magnet connected to the vane passes over the relay. In this case a resistance divider replaces the potentiometer. Readers who are adept at making very small printed circuit boards, may like to replace the carbon track of a conventional potentiometer by a small 1 6-segment circuit board and connect each segment to the resistance divider. The supply does not need to be stabilised, since the IC has an internal reference voltage output (pin 14) which is (gratefully) utilised. The maximum current through an LED is approx. 50 mA, thus a suitable supply would be a transformer producing 100 mA with a voltage of 9 or 12 V. The circuit is completed by a bridge rectifier and a 470 /i 25 V electrolytic capacitor. ± 15 volt regulator Using IC 1568 or 1468 (from, among others, EXAR) and only a small number of external components, it is possible to produce a symmetrical, current-regulated supply voltage of plus and minus 1 5 volts. The circuit is intended as an ‘on card’ supply, and is not particularly suited tor experimen- tation, since the maximum dissipation of the IC is not particularly high (max. 1 W). With the circuit arrangement as shown in the figure, it is not advisable, in view of this dissipation value, to select input voltages much greater than those indicated (i.e. 3 V above the output voltage). It goes without saying that the IC will not be able to tolerate shorting the outputs for long. The current is limited as soon as the voltage drop across either of the two resistors R1 and R2 exceeds 0.6 V. By means of PI the output voltage may be varied between 14.5 and 20 V (always assuming that the input voltage is sufficiently high). The positive and negative voltage can be matched exactly using P2. Capacitors Cl ... C4 are needed to guarantee the stability of the supply, and should be positioned as near as possible to the IC. Some further details: the 1568 differs from the 1468 in having a slightly narrower tolerance with respect to the value (0.2 compared to 0.5 V) and match of the output voltages (150 compared to 300 mV). The maximum input voltage is 30 V, and the 17.5. ..23V 17.5.. .23V max. current 100 mA. A change in the load of 50 mA causes a variation of approx. 3 mV in the output voltage. For a current of 50 mA the minimum voltage drop across the regulator is at least 2.5 V. The noise suppression is then around 75 dB, and the stability of the output voltage is better than 1% for a variation in temperature of 75 C. The noise at both outputs is less than 0.1 mV. elektor july/august 1977 phasemeter sv AC touchswitch A.M. Bosschaert elektor july/august 1977 This phasemeter will measure the phase angle between two signals over the fre- quency range 10 Hz to 100 kHz, and is thus extremely useful for measuring the phase response of audio systems. The principle of operation is as follows: the meter is calibrated so that when the out- puts of N3 and N4 are high it reads full-scale. Flip-flop N1/N2 is set at the positive-going zero-crossing of waveform A and reset at the positive zero-crossing of waveform B. While the flip-flop is in the set condition the outputs of N3 and N4 will be high. If the waveforms are in phase the flip-flop will be reset as soon as it is set and the out- puts of N3 and N4 will remain low, so the meter will read zero. When the phase angle approaches 360° the flip-flop will remain set practically all the time and the meter will read full-scale. At a phase angle of 180° the flip-flop will be set for half the time and reset for half the time so the meter will read half-scale. To ensure that the flip-flop is triggered at exactly the right points independent of the input waveshape or amplitude the input signals must be processed. The two input channels are identical to ensure that any phase shift introduced by the signal- processing is the same in each channel and will thus cancel. Each channel consists of a source follower FET to provide a high input ’ impedance (approx. 1 M£2/10pF). R1 and diodes D1 and D2 protect the input against excessive voltages. This is followed by a xlO gain stage T2, with limiting diodes on the output, and a second xlO gain stage T3. The amplified and limited waveform is then fed to a comparator (IC1), which is connec- ted as a Schmitt trigger. The output of this goes low on each positive transition of the input and high on each negative transition. The negative output pulses of IC1 are differentiated by C8, D5 and R17 and used to trigger the flip-flop. A The meter is best calibrated at the 180° point, since it is easy to obtain signals 180° ' out of phase from the two halves of a centre- tapped transformer secondary. The signals are fed into inputs A and B and PI is ad- justed until the meter reads half-scale or 180°. Alternatively the meter can be cali- brated at 360° by grounding the input of N 1 and adjusting PI until the meter reads full- scale. The meter scale should, of course, be marked out linearly from 0 to 360°. The meter will function with input voltages greater than a few millivolts RMS and is protected against input voltages in excess of 250 V, and can thus operate over an extremely wide dynamic range without adjustment. Many designs for touchswitches have pre- viously been featured in Elektor. However, most of these operated on skin resistance and thus required a double contact that could be bridged by a finger. Single contact operation is possible using a capacitive pick- up of mains hum, but this is not very reliable, and will not work at all with battery- powered equipment! The design given here overcomes these difficulties and provides a reliable single-point touch switch. N3 and N4 form a 1 MHz oscillator. When the contact is not touched the signal from the output of N4 is fed via C2 and C3 to the input of Nl, which causes the output of N1 to go high and low at a 1 MHz rate. This charges up C4 via Dl, holding the input of N2 high which causes the output to remain low. When the contact is touched, body capaci- tance ‘shorts out’ the 1 MHz signal. The input of Nl is pulled high by R3 and the output goes low. C4 discharges through R2 N3 N4 and the output of N2 goes high. One oscillator will provide a 1 MHz signal for several touch switches, which may be connected to point A. m3 audible logic probe H. Kaser This logic probe provides an audible rather than a visual indication of logic state by producing a high-frequency audio tone for a logic T’ state and a low-frequency tone for a logic ‘0’ state. The logic input signal is fed to N1 and N2. If the input is high then N2 will pass the high frequency signal from the oscillator built around Al. If the input is low then N2 will block, but the output of N1 will be high so N3 will pass the low frequency signal from the oscillator built around A2. Depending on the input state one or other of these signals is fed through N4 to the input of a differentiator built around A4. This produces a train of short pulses from the squarewave input signal and these are fed to an audio amplifier comprising T1 and T2. The use of short pulses ensures a high peak audio output while keeping the average current consumption low. To avoid the annoying ‘bleeping’ of the circuit when measurements are not being taken both oscillators may be switched on and off by a flip-flop constructed around A3, which is controlled by two push buttons SI and S2. If the circuit is to be used exclusively with TTL circuits then N1 to N4 should be a 7400 IC and the supply voltage should be +5 V, which can be derived from the circuit under test. If it is to be used with CMOS ICs then N1 to N4 should be a 401 1 IC, and the circuit will operate over supply voltages of 5 to 10 V at a current consumption of between 4 and 10 mA. comple- mentary emitter follower elektor july/august 1977 This circuit presents an interesting alterna- tive method of constructing a low-distortion buffer or output stage for use at low output powers. The quiescent current flowing through T1 and T2 is determined solely by the value of U and of R1 and R2 respect- ively. This contrasts with conventional cir- cuits where the bases of T1 and T2 are connected to one another by means of a diode network. The current supply of the diodes normally has an unfavourable influ- ence on the input impedance (unless bootstrapping is used) causing variations in the quiescent current. In this circuit the quiescent current through T1 equals and that through T2 is Ri — — , assuming that the current gain of T1 R2 and T2 are so high (or closely matched) that the voltage drop across R3 is negligible. Normally Rl is given the same value as R2. The relative values of C2, C3 and R4 deter- mine the lowest frequency at which the circuit will function. If T1 and T2 have the same current gain and Rl equals R2, then no DC voltage is pro- duced across R3, and Cl may be omitted. If the circuit is fed from an op-amp then both Cl and R3 may be omitted. u u The circuit is intended as a class-A buffer or output stage. The maximum class-A output power dissipated in R4 is I 2 R4, where I is assuming that R4 is smaller than R = Rl = R2. reaction speed tester telephone ’tell-tale’ one of the most popular types of electronic remains lit. game, namely a reaction tester. A new game can be started after pressing As soon as the ‘start’ button is pressed, IC1 the reset button. feeds a train of pulses to the counter IC3, With the component values given in the causing LEDs 1 . . . 10 to light up one after diagram the circuit consumes 120 mA with another. The sooner the ‘stop’ button is a 5 V stabilised supply. The oscillator fre- pressed, the smaller the number of LEDs quency may be adjusted by means of PI which light up; the last LED to light up between 10 and 80 Hz. burns continuously. If the oscillator which If desired, an additional LED with a 220 12 generates the clock pulses is set so that a series resistor can be included between the pulse is produced say, every 20 ms, then the output of N3 and positive supply. This will reaction time of the players can be calcu- light up as soon as the opponent presses lated quite simply by observing which LED the ‘start’ button. This circuit is designed to indicate whether around 1 second, P2 should be adjusted so anyone has attempted to telephone the that the MMV remains triggered for just' householder during his absence. If the tele- under a second. This ensures that only phone bell rings at least 8 times, an LED signals of approx. 1 second or longer will be (Dl) will light up until the reset switch SI recognised by the circuit and clock the is pressed. counter. The signal from the bell is picked up by a If a signal lasts for less than the preset time, microphone capsule fixed to the bottom of the output of N1 will be low and the pulse the phone, and amplified by IC1 . As long as from N6 will be passed through N7 and N8 no signal from the bell is present, C2 will to reset the counter. 9V 77108 remain charged. The moment the signal In order to further decrease the sensitivity sounds T1 is turned on, discharging C2. The of the circuit to spurious signals the last output of N 1 goes high, the MMV round N2 output of the counter is taken to drive the and N3 is set and remains in the triggered LED, so that the phone must ring at least state for a certain time (adjustable by means 8 times without interruption before the of P2). When the MMV resets, a pulse is fed LED will light up. This number can naturally via N6 to N5 and N7. If the signal is still be reduced by using one of the other present (the output of N1 is still high), this outputs. pulse will be passed through N5 to the With a supply voltage of 9 V the circuit clock input of the counter IC4. draws a maximum current of 5 mA. A 9 V elektor juiy/august 1977 Since the duration of the signal is normally battery provides a suitable supply. tri-state voltage comparator y i-ilCI -M- P.rY 1- (}► DUS (22k)vf ( + HC2 Q / t p2 rV (}► . . n 25ok \A $ | (220k yM level shifter This circuit will compare an unknown input voltage with two preset reference voltages and display the result on one of three LEDs. IC1 and IC2 function as comparators. If U is less than U2 then the output of IC2 will be high and D2 will be lit. If U is greater than U2 but less than U1 then the outputs of IC1 and IC2 will both be low and D3 will be lit. If U is greater than U1 then the out- put of IC1 will be high, IC2 will be low and D1 will be lit. Of course, the foregoing assumes that U1 is greater than U2. This can be achieved by correct adjustment of PI and P2. Alterna- tively, to ensure that U1 is always greater than U2 PI and P2 can be arranged as shown It is often necessary, particularly when experimenting with circuits, to make connection between the output of one circuit and the input of another which is at a different DC level. If the signals involved in the circuit are AC signals this is no problem, a capacitor can be used to isolate the DC levels while allowing AC signals to pass. However, when dealing with DC or very low frequency AC signals the solution is not so easy, and it is in these cases that this little gimmick will prove useful. The circuit consists simply of an op-amp connected as a voltage follower whose quiescent output voltage can be set to any desired level within the output range of the op-amp. Input A is connected to the output of the circuit in question while the output is connected to the input of the circuit which it is feeding. Input C is grounded, while input B is connected to a DC voltage equal to the difference between the input voltage of the second circuit and the output voltage of the first. It can easily be proved that this works! Firstly, voltages appearing at the non- inverting input of the op-amp are amplified by a factor R3 + R4 _ - Secondly, suppose the output voltage of the first circuit is Va and the input voltage of the second circuit is Vj. The voltage Vg applied to input B is thus V] - Va. The voltage appearing at the junction of R1 and R2 is thus V A + VB ~ Va . The voltage appearing at the op-amp output is twice this, i.e. Va + Vg But since Vg = Vi - Va this equals Vj , the ■ input voltage of the second circuit, elektor july/augurt 1977 I Obviously, if V x is less than Va then Vg with P2 deriving its supply from the slider of PI. Note that P2 is now 250 k. With this arrangement there will be some interaction between the two potentiometers. The circuit makes an ideal battery state indicator for a car. PI and P2 can be adjusted so that D2 lights if the battery voltage falls below say 1 1 V, D3 lights between 1 1 and 1 3 V and D 1 lights above this. Many types of op-amp IC will work in this circuit, but if a quad Norton op-amp such as LM3900 is used then a 1 00 k resistor should be placed in series with the + and - inputs of each op-amp. will be a negative voltage. Despite the difference in input and output levels the circuit functions as a voltage follower in that any change in the voltage at input A will produce the same voltage change at the output. The circuit can also be used as an inverter. In this case the signal is fed to input C, B is grounded and A is fed with a DC reference voltage. To see what voltage must be applied to A it is simplest to treat the circuit as a unity gain differential amplifier. The output voltage V Q is equal to the difference between the voltages at the non-inverting and inverting inputs i.e. V G = Va-Vc so Va = V 0 + Vc, i.e. input A must be fed with a voltage that is the sum of the voltage at C and the required output voltage. Any change in the input voltage at C will produce the same change at the output, but of opposite polarity. Two points must be noted when using this circuit. Firstly, care must be taken not to exceed the common-mode input rating of the op-amp used, especially with a single- ended (asymmetric) supply. Secondly, the values of R1 to R4 should be at least ten times the output resistance of the circuit feeding the level shifter to avoid excessive loading of the output. ultrasonic receiver This receiver is intended to be used with the US transmitter described elsewhere in this issue. The input signal is fed to a cascode amplifier comprising T1 and T2, which amplifies it approximately 2000 times. T3 functions as a rectifier, and T4 amplifies the rectified signal, which is used to trigger a flip-flop comprising T5 and T6. This will be set on one US burst and reset on the next, thus turning T7 on and off to success- ively energise and de-energise the relay. To reduce the circuit’s sensitivity to multiple triggering (caused principally by Doppler- shift) positive feedback is provided via C8 causing T3 and T4 to operate as a monostable. This ensures that only one output pulse is generated for each received pulse and prevents spurious triggering of the flip-flop. To set up the receiver the following pro- cedure should be observed. Turn the slider of PI towards the 0 V rail. This should cause the relay to pull in and drop out at random. PI is then adjusted until this just ceases, when the receiver sensitivity will be at a maximum. Activating the transmitter should now energise the relay. The system will oper- ate at distances of up to eight metres. If this range is too great then it may be reduced by suitable adjustment of PI. The receiver should be set to the minimum range required for a particular application, as too great a sensitivity may cause spurious triggering by normal sounds which have an ultrasonic content, e.g. handclaps, rustling paper etc. ultrasonic transmitter for remote control This simple ultrasonic (US) transmitter uses only one CMOS IC and a few discrete com- ponents, and will generate a short pulse of ultrasound. This can be used to trigger an ultrasonic receiver (such as the one described elsewhere in this issue) to activate a relay or other circuit. When the touch contacts on the input of N 1 are bridged by a finger the output of N 1 will go low and the output of N2 will go high, holding the input of N3 high via C2 for about 60 milliseconds. During this time the astable multivibrator comprising N3 and N4 will oscillate at about 40kHz (adjustable by PI). Since the output of N4 can supply little cuirent, drive to the US transducer is provided by Tl, which feeds the resonant circuit L2/C5. This is tuned to around 40 kHz, which is the resonant frequency of most US transducers. Although only a 9 V supply is used, the ‘Q’ of the resonant circuit ensures that a high drive voltage appears across the transducer (up to 100 V), and a range of up to eight metres can be achieved in conjunction with the aforementioned receiver. As the circuit consumes virtually no power except when actually transmitting, no on-off switch is required. Almost any 40 kHz US transducer can be used. elektor july/august 1977 I This simple equaliser is intended mainly for tailoring of room acoustics in such appli- cations as disco and p.a. work, and for sound effects in electronic music. It is not intended for hi-fi use as more sophisticated circuits are needed to give satisfactory results in this application. The circuit comprises an emitter follower T1 feeding a number of Wien networks, each of which passes a band of frequencies about its centre frequency. PI, P2 . . . etc. vary the proportion of the total signal that is fed to each Wien network and hence the pro- portion of each selected frequency band that appears at the output. By using a number of Wien networks with centre frequencies spaced at suitable intervals throughout the audio spectrum it is possible to boost or cut selected bands of frequencies, and thus adjust the response of the equaliser to compensate for room acoustics etc. The output of each Wien network is fed to a summing amplifier consisting of T2 and T3, which has a gain of three to overcome the attenuation of three introduced by the Wien networks at their centre frequencies. For most purposes five Wien networks should be sufficient with centre frequencies of 40 Hz (C = 39 n), 155 Hz (10 n), 625 Hz (2n2 in parallel with 330 p), 2.5 kHz (680 p), 10 kHz (160 p). EPS 77071 I j| shoo dog! \ elektor july/august 1977 This ‘inaudible’ car horn is intended for warning four-footed pedestrians of imminent danger, without needlessly scaring the (hopefully) better-behaved two-footed variety. Basically, the unit is a power multivibrator that oscillates at a frequency above the human hearing range - but clearly audible to the canine ear. The correct frequency can be set with PI. Bear in mind that young people can usually hear higher frequencies! R7, R9, C2 and D1 are required to start up the oscillator by passing a short current pulse into the base of Tl. The only dis- advantage of this type of circuit is that power must be applied ‘instantaneously’ — it will not work if the supply voltage rises slowly to its final value. However, if the unit is connected to a car battery, depressing SI will reliably start the multivib. The minimum loudspeaker impedance is 4 £2. On a 1 2 . . . 14V supply the unit will deliver 5 W into this impedance. Alternatively, on a 40 V supply it will deliver 25 W into 8 12. Not many tweeters can withstand this sort of drive level for long, so it is not advisable to use the horn for longer than a few seconds at a time. As a final note: the unit can also be used to chase stray dogs out of the garden - and it is certainly more humane than throwing stones 12. ..40V minireg power supply 63 non-inverting integrator elektor july/august 1977 Although it utilises only two transistors and eight other components, this simple stabil- izer will supply currents up to about 4 A and is equipped with foldback current limiting. The circuit operates as follows: ignoring K1 for a moment, current from the output of the supply flows through Dl, D2, R3 and PI to ground. Due to the forward voltage drop of Dl and D2 the emitter of T2 is always biased about 1 .2 V lower than the output voltage. Should the output voltage tend to rise the emitter voltage of T2 would rise by the same amount, but since the base is fed from the potential divider P2/R4 the base R 4 voltage will rise by only 5 — , - 5 - times this •r 2 K 4 amount. The base-emitter voltage of T2 will thus tend to fall and T2 will draw less current from R2. The base-emitter voltage of T1 will thus fall and T1 will tend to turn off, so the output voltage of the supply will fall. Should the output voltage tend to fall the reverse will occur. The base-emitter volt- age of T2 will tend to increase. T2 will turn on harder, which in turn will turn T1 on harder, and the output voltage will be restored to its original value. The output voltage of the power supply is given by V o “(V z -0.6).(~ + l) Where V z is the forward voltage drop of Dl plus D2. As the output current of the supply increases the current through T2 also increases as it turns on T1 harder and harder. The voltage drop across R3 and PI will therefore increase, and the current through Dl and D2 will decrease. When the current has de- creased so far that Dl and D2 are in the non- linear region of their forward transfer curve the voltage across Dl and D2 will begin to fall, and hence the output voltage will fall since the V z term in the above equation is decreasing. The output current will fold back as shown in the graph. The maximum output current is given approximately by (V 0 - V z ) hpE Ti Imax Ra + Pl (A) where hpE Tl is the DC current gain of Tl. With an average BD242 having a gain of about 25 the maximum output current can be adjusted from a few hundred mA to about 4 A with the component values shown. However, using the equations given it is not difficult to redesign the power supply for other output voltages and currents. Resistor R1 is necessary to ensure that the circuit starts up reliably. If R1 were omitted then, on switch-on, Tl would be turned off and the circuit would fail to start. R1 pro- vides base current to T2 which ensures that the circuit starts. The value of R1 should be the largest value at which the circuit still starts reliably. If the circuit is to be used for lower currents than that specified then a lower power device may be used for Tl. For currents up to about 50 mA a TUP may be used, pro- vided the difference between the input and output voltages is not more than two or three volts. For currents from 50 to a few ^ hundred milliamps a medium power, device ' such as a BC143 could be used. A drawback of conventional integrator cir- cuits (figure a) is that the R-C junction is at virtual earth; this means that C appears as a capacitive load across the op-amp output, a fact that may adversely affect the stability and slew rate of the op-amp. Since the non- inverting character of an integrator is of minor importance in many applications the circuit shown in figure b offers a viable alternative to conventional arrangements. This integrator, unlike that in figure a, is non-inverting. The time constants R 1 C 1 and R 2 C 2 should be equal. If both Ri and C,, and R 2 and C 2 are transposed then the result is a non-inverting differentiator. For correct offset-compensation Ri and R 2 should have the same value. m l short-wave converter v \ 77092 t tes positive- triggered set-reset flip-flop using inverters elektor july/august 1977 This frequency converter enables a normal medium wave radio to tune the short wave bands from 2 to 35 MHz. A MOSFET mixer and BFO are used to convert the frequency of the input signal to a frequency at the top end of the medium wave band (800 to 1 600 kHz) which can be picked up by the ferrite aerial of the receiver. A telescopic aerial of about one metre length should be used for the converter. A defunct car aerial with the cable removed is ideal, and can be picked up for next to nothing at any car scrapyard. The shortwave signal is fed from the aerial via Cl to the resonant circuit L1/C2 which tunes the aerial input, and thence to gate 1 of a dual gate MOSFET Tl, which functions as a mixer. T2 is the BFO and functions as a Clapp oscillator tuned by L2 and C8 ... Cl 1. The BFO signal is fed to the second gate of Tl . The mixer output is taken to a coil which consists of a few turns of insulated wire wrapped around the case of the MW receiver to feed the signal into the receiver’s ferrite aerial. The MW receiver should not be placed too close to the converter as harmonics from The standard set-reset flip-flop circuit con- sists of two cross-coupled NAND gates and is set and reset by applying a logic ‘0’ level to the appropriate input. The circuit shown in the figure is triggered by a logic *1* and uses inverters. Assume that initially both inputs are low and the Q output is high. The input of N1 is also pulled high via the 1 80 k resistor, so the Q output is low, which holds the input of N2 low._If a logic ‘1’ is applied to the S input the Q output will go low, pulling the input of N1 low, and the Q output will go high, thus holding the input of N2 high even if the S input subsequently goes low. Apply- ing a logic ‘1’ to the R input will reverse the procedure and reset the flip-flop. The circuits feeding the inputs of the flip- flop should be capable of providing a logic ‘1’ level into a 180 k load, which normal CMOS circuits are capable of doing. Reference: RCA Application Notes. the receiver’s local oscillator may be picked up by the converter. To use the converter the MW receiver is tuned to about 1200 kHz and C9 is adjusted until short wave stations can be heard. C9 is used to tune the BFO so that 1200 kHz on the MW receiver corresponds to the centre of the desired short wave band i.e. by tuning the receiver from 800 kHz to 1600 kHz the entire band can be covered. Once a station is tuned in C2 can be adjusted to give maximum signal strength. Values of LI and L2 for the various short wave bands are given in table 1 . Table 1 Frequency band (MHz) L, (mH) L 2 (mH) 2 . . 2,8 120 100 2,8 . . 4,0 56 56 4,0 . . 5,8 27 33 5,8 . . 8,3 15 18 8,3 . . 1 1 ,9 6,8 8,2 11,9 . . 17,0 3,3 4,7 17,0 . . 24,4 1.5 2,2 24,4 . . 35,0 0,82 1 fe6 auto trigger level control left I fuzzbox elektor july/august 1977 Oscilloscopes, frequency counters and other instruments triggered by AC signals almost invariably have a manual trigger level control, to adjust the point on the waveform at which triggering occurs. When making measurements where the signal level varies, for example at different places in a circuit, it is tedious to have to make frequent adjustments to this control. The circuit described here provides a trigger signal at a fixed percentage of the peak input level, irrespective of what that level is, so the frustration of having the trace disappear from an oscilloscope when the signal level falls below the trigger level is avoided. The circuit consists basically of a peak rectifier that provides one input of a comparator with a DC voltage equal to a fixed percentage of the peak signal level. The other input of the comparator is fed with the signal. When the signal level exceeds the DC reference level the comparator output will go low. When it falls below the reference level the comparator output will go high. The peak rectifier consists of IC1 and Tl. On positive half cycles of the signal waveform the output of IC1 will swing positive until Tl starts to conduct, after which IC1/T1 will act as a voltage follower, charging up Cl to the peak value of the signal. A portion of this voltage is taken from the slider of PI and applied to the non-inverting input of IC2, which functions as a comparator. The AC signal is fed to the inverting input. When the signal level exceeds the reference voltage the comparator output will go low; when the signal level falls below the reference level the comparator output will go high, (see figure lb) PI may be used to set the trigger level to any desired percentage of the signal level. The DC level at the slider of PI may also be fed \ to the comparator input of an existing trigger level circuit. In this case this circuit should have a high input impedance to avoid discharging Cl. Alternatively the output from PI can be buffered by an op-amp connected as a voltage follower. lb The fuzzbox is an indispensable piece of equipment for the electric guitarist. This electronic device limits the guitar signal and produces a clipped waveform which, as is well known, contains a large number of harmonics. The guitar therefore produces a much richer sound. Many commercial fuzz amps suffer from the drawback that they may only be used with guitars which have pickups with low output impedance. This is not the case with the circuit described here. At the input is an emitter follower which acts as a buffer between the guitar output and the limiting amplifier and ensures a high input impedance. Operational amplifier IC1 is used to limit the input signal. The gain of this amp and therefore the degree to which it limits the signal may be adjusted by potentiometer PI. When reverse parallel connected diodes D1 and D2 are conducting then the fuzz effect is produced. At the output there is another emitter follower which feeds the distorted guitar signal to the power amplifier at low impedance. Thus cable capacitance, even with a long cable, does not affect the high frequency response. By means of P2 the output signal of the limiting amp may be adjusted to suit any guitar amplifier. Switch SI , which is used to switch the fuzzbox on and off, can be conveniently mounted in a foot pedal. 68 LED logic flasher 68 CMOS alarm circuits elektor july /august 1977 The condition of the LED is determined by the logic states of the two inputs A and B. If A is low and B is high then the LED will be lit continuously. If B is low then the LED will be extinguished, irrespective of the state of A. If A and B are both high then the astable multivibrator comprising Nl, N2 and N3 will start to oscillate and the LED will flash at about 3.5 Hz. Component values are given for supply voltages of 3, 10 and 15 V. At the maximum supply voltage of 15 V the current consumption is less than 25 mA. Source: RCA CMOS Application and design ideas. Because of their low cost, high input resist- ance, good noise immunity and wide supply voltage range CMOS logic circuits lend themselves to the construction of cheap and reliable alarm circuits for various appli- cations. Figure 1 shows a basic alarm oscillator constructed around two CMOS NAND gates. As long as input Q is low the circuit remains inactive. When input Q goes high the circuit begins to oscillate, switching T1 and T2 on and off and producing an alarm signal from the loudspeaker. Figure 2 shows a circuit for triggering the alarm after a preset time delay that can be varied between one second and one minute. On switch-on the input of N2 is briefly held low by C2, so the flip-flop is reset and the Q output is low. Cl now charges via PI so that the input voltage of Nl falls until the flip-flop is set and the Q output goes high. Figure 3 is an alarm circuit that is triggered when a circuit is broken, this is particularly useful in burglar alarm systems where several switches can be connected in series. The input of Nl is normally pulled high via the switches, but if a switch is opened the input of Nl will be pulled down by the resistor, setting the flip-flop and triggering the alarm. If the resistor is connected up to + supply and the switches are connected in parallel down to ground then the alarm can be triggered by closing a switch. Figure 4 shows an alarm circuit that responds to light. In darkness the resistance of the LDR is high and the input of Nl is pulled high by R1 and PI. If light falls on the LDR the resistance falls and the input of Nl is pulled down, setting the flip-flop and triggering the alarm. ^=180°-2arctan _p^< HRiaK > O ° C Hna f o «|Sgj o y ofHs~|o O ” O jmq h o o <>H l-oo S O O ^RIQ t o o H l-oo o-| l-oo ^ 1 o |pi k > O Hr3 1 -0 £ annnnnn, rcCtr O O. O O O -m- r ♦ o • elektor july/august 1977 I The input stage is followed by a tone control stage which possesses bass, middle and treble controls. As the frequency response of many guitar pick-ups is far from flat these controls can be used to compensate for any peaks or dips. The response of the tone control networks for different settings of the control pots is shown in the accompanying graph. The oscillogram, the lower trace of which shows the preamp output when fed with a 1 kHz squarewave (upper trace) illustrates that the h.f. response of the preamp is fairly good. Indeed, the performance of the preamp is so good that, as well as its intended use, it may also be used in hi-fi systems. Some children have less difficulty falling asleep if the bedroom light is left on. How- ever, this means that one of the parents has to go up after half an hour or so to turn the lights off - hopefully without waking the children up again . . . The circuit described here will fade out the lights very slowly, either completely off or to a preset mini- mum (night-light) level. As long as SI is closed the light(s) will burn at full brightness. As soon as SI is opened, the lights start to fade out very gradually until they reach a certain level (preset by means of PI). The fade-out time is deter- mined by the value of C4 and by the setting of PI. As an example, if C4 = 100/i and PI is set at minimum it will take approximately half an hour for the lamp to fade out. If 90 CMOS PLL elektor july/august 1977 As PLL (phase-locked loop) ICs are still somewhat expensive it seems reasonable to look around for a cheaper alternative, par- ticularly for non-critical applications that do not require such high specifications. Using two CMOS NAND gates it is possible to construct a CCO (current controlled oscillator) as described elsewhere in this issue. If a 401 1 quad two-input NAND gate IC is used this leaves one gate to act as a phase comparator and another as an input amplifier. The circuit shows a complete PLL using one 4011 and a few discrete components. Con- sidering the simplicity and low cost of the circuit the results obtained were surprisingly good, and using a typical 401 1 the following measurements were taken. CCO frequency range (adjusted by P2): 25 kHz - 800 kHz. Hold range: 20% of CCO free-running frequency. Output level: 45 mV measured at fj n = 500 kHz, deviation = ± 30 kHz, modulation frequency = 1 kHz. AM suppression for 30% AM: better than 40 dB. Minimum input level: less than 2 mV from 50 £2 source. These measurements were taken at a supply voltage of 6 V, when the current consump- tion was 600 juA. Since different IC manufacturers use differ- ent processes and different chip geometries it might be expected that results would vary when using different types of IC. The best results were obtained using ICs in which the gates had a steep transfer characteristic (better approach to an ideal switch) and lowest crosstalk between gates. In our experience, the Solid State Scientific SCL 4011 is a good example of this type of chip. desired, C4 can be increased; however, it is not advisable to go above about 470 p. The circuit must be mounted in a well-**' insulated case, and PI should be a potentio-J^ meter with a plastic spindle. The type of triac required will depend on the load, of course. It is advisable to select a type that is capable of handling a current of up to where PL is the nominal ‘wattage’ of the lamp and UM is the nominal mains voltage. A ‘cold’ filament draws a relatively heavy current! (Plessey application) The 4011 PLL is particularly suitable for narrow-band FM demodulation, and in fact proved superior, in terms of s/n ratio and impulse noise rejection, to several mono- lithic PLL ICs. osculometer i- ‘Lie Detector’ machines, which measure skin resistance, can provide great amusement at parties, especially if two participants each hold one electrode and indulge in some form of physical contact (e.g. kissing). The meter can then be calibrated in degrees of passion. A variation on this theme is a circuit that produces an audible output rather than a meter indication, which is even more amusing. The circuit consists of two current controlled oscillators (described elsewhere in this issue). The output of oscillator N1/N2 gates oscillator N3/N4 which produces some interesting effects. The output of N4 is used to drive an audio amplifier comprising T9 and T10. The circuit has provision for eight electrodes for up to four pairs of participants. As the resistance between a pair of electrodes decreases then the frequency of the corre- sponding oscillator will rise, so the more ardent the embrace the higher the oscillator frequency. The gating effect between the two oscillators produces some unusual sounds. For safety reasons the circuit should be battery powered by a 9 V transistor ‘power pack’, such as a PP3, PP6, PP9 etc. elektor july/august 1977 Using two CMOS NAND gates (or inverters) and two transistors it is possible to construct a simple current-controlled oscillator (CCO). The circuit of figure 1 is based on a normal two-inverter astable multivibrator. When the output of N1 is high the output of N2 will be low and Cl will charge through T1 until the threshold voltage of N1 is exceeded, when the output of N1 will go low and the output of N2 high. T1 will now operate in a reverse direction, i.e. the collector will function as the emitter and vice versa, and Cl will charge in the opposite direction. When the voltage at the collector of T1 falls below the threshold of Nl, then the output of Nl will go high and the cycle will repeat. T1 and T2 form a current mirror, i.e. the collector current of T1 (which is the charging current of Cl) tracks or ‘mirrors’ the collector current of T2, which is, of course, controlled by the base current. If the two transistors were identical then the collector currents would be the same. A frequency range of about 4 kHz to 100 kHz is obtainable with the component values shown. When T1 is conducting in the reverse direc- tion T2 will be turned off and its base- emitter junction reverse-biassed. If the supply voltage is greater than +5 V then this junction may break down, but as the volt- ages and currents involved are fairly small no damage will occur. A circuit that avoids the unusual mode of operation of T1 and possible breakdown of T2 is given in figure 2. Here a diode bridge D1 to D4 ensures that the current through T1 and T2 always flows in the correct direction. The advantage of this circuit is that the astable may also be controlled by other asymmetric devices such as photo- diodes and phototransistors. s N1.N2 = Vi 401 1 (V 3 4049) key board decoder “5 i J si el ?i sJiofnT 0 1 23456789 7442 D C B A 12 13 14| 15 ©— ■ * 1= 11 8 9ll2 0 c B A bd in iJ 14 C a in 7490 R om l ’oO) R 9(1I R 9I2I I 7 Jr ’,ovi 1N4148 Slf v 21 1N4148 • 6| T 2 *V 2 '4123 . -I I 74123 L J n 74 123 iol 8 clear ° I When one of the keys SO . . . S9 is pressed the keyboard decoder generates the BCD code which corresponds to that key. To ensure that only the desired BCD infor- mation is read out the circuit also produces a strobe pulse which indicates that the information has been accessed. The diagram illustrates how clock pulses are fed from a free running squarewave oscil- lator 0/27413) to a decade counter (7490). The state of the counter is decoded and fed to the contact keys SO . . . S9. The outputs of the decoder go successively low. Thus, the keys are scanned rapidly and sequentially until one of them is pressed ./ The corresponding output of the decode^ will, after a certain time, go low. This ‘O’ stops the oscillator and the counter, which then remains in the state which coincides with that of the key which has been pressed. To prevent possible mistakes arising as a result of contact bounce, the circuit includes a monostable multivibrator, which after 0.3 ms triggers a second monostable. This in turn supplies the pulse which reads in the BCD information from the 7490. car headlight alarm R . T rost This circuit is designed to give an audible and visible warning to the absent-minded motorist who forgets to switch off the headlamps when leaving his car. The power supply to the circuit is taken from the headlamp switch, represented by S2. If the headlamps are switched off then the alarm obviously receives no power and does not operate. The actual light switch in the car will be more complex than S2 since it also controls the sidelights. How- ever, examination of the car wiring diagram and a little probing with a multimeter will soon show which terminal of the switch acquires a positive voltage when the head- lamps are switched on. SI represents the car ignition switch. As long as the ignition is switched on pins 2 and 6 of IC1 are pulled high via D1 and Rl. When the ignition is switched off, however, this voltage will fall as Cl discharges through PI. PI sets the time allowed for switching off the headlamps before the alarm sounds. When the voltage on Cl falls below the trigger threshold of IC1 then, assuming the headlamps are still switched on, the output (pin 3) of IC1 will go high, turning on Tl. This lights the seven segment LED display which gives an ‘L’ indication. If the expense of a LED display is not thought to be justified then a single LED or lamp could be used. Tl also triggers IC2, which is connected as a monostable multivibrator. The output of IC2 goes high, thus activating astable multivibrator IC3, which begins to oscillate, producing an alarm signal from the loudspeaker. The length of time for which the alarm sounds is determined by the period of the monostable IC2, which may be adjusted by means of P2. P3 adjusts the volume of thf alarm signal. To prevent the possibility of IC2 being spuriously triggered when the headlamps are switched on, T2 is provided. When the headlamp switch is closed C5 begins to charge through R6 and R8, which momen- tarily turns on T2. T2 thus holds the reset input (pin 4) of IC2 low so that it cannot trigger. self-shifting register A.M. Bosschaert Q1(J Q2Q 030 Q< The unusual feature of this shift register is that it will transfer pulses from its input, through several stages, to the output without the need for an external clock generator. The shift speed is fixed, and is determined by the component values in the circuit. When the input goes high, the output of N1 will go low and the input of N2 will be held low for a period determined by the time constant R1 • Cl. During this time the out- put of N2 will be high. When the input of N2 goes high again, the output of N2 will hold the input of N3 low for a time deter- mined by R2 • C2. In this way the pulse is shifted through the register. When the input LED tuner W. Auffermann goes low, Cl will simply discharge through R 1 and the output of N 1 , ready for the next pulse. It is apparent that, provided the length of the input pulse is longer than the time constant R1 • Cl, the length of the output pulse is determined solely by the circuit time constants. In general the time constants R1 * Cl, R2 • C2 etc. will all be equal, and in this case the maximum input pulse rate is determined by the fact that the interval between two pulses may not be less than the time constant R1 • Cl, otherwise pulses may overlap. t~ T'c — rtl R 3f*| A?' R1 P -| 100k | — }y< Cl Ta _jii— < R7 H ^^HTUN T , ■ 25V R8 P A. . r T2 TUN 1C 2 s NLrf7T-iT (r\ 3 741 27 k 1 * til rf x R12 "gP K 1 H 1M h x\ R10 |ci Cr Is . 73 , |S 8 N1...N4=IC3=CD4011 M N4 IQ — | 27k j elektor july/august 1977 This circuit can be used as a tuning indicator, instead of the more common pointer instru- ment. It gives a three-LED indication of correct tuning: ‘off-to-one-side’, ‘correctly tuned’, ‘off-to-the-other-side’. A voltage is derived from the AFC control voltage in the FM receiver and fed to two comparators (IC1 and IC2). The divider chain R2, R3, PI and R4 produces two reference voltages. If the input voltage is higher than the greater of the two reference voltages T 1 will be turned on and LED D1 will light. In the other extreme case, where the input voltage is lower than the lower reference voltage, T2 will be turned on and LED D2 will light. In the in-between range, where the receiver is correctly tuned and the input voltage is somewhere between the two reference voltages, T1 and T2 will both be turned off. In this case the output of N1 will go low, the trigger circuit N2/N3 will switch, the output of N4 will go high and T3 will be turned on - lighting LED D3. Since the AFC voltage corresponding to ‘correctly tuned’ varies considerably from one receiver to the next, the values of R2, R3, R4 and PI are not given in the circuit. It is a simple matter to calculate these values for any particular application. If the total resistance is to be 20 ... 30 k (a reasonable assumption), the voltage midway along R3 should correspond to the AFC voltage for correct tuning. To give two examples: - assume that the ‘correct’ AFC voltage is 9.5 V. In this case the voltage across R2 + ViR3 should equal 2.5 V and the volt- age across the rest of the chain should equal 9.5 V. If R2 is selected as 4k7 and a 1 k preset is used for R3, the sum of PI plus R4 should be approximately 20 k with PI in the mid position. A good choice in this case would be R4 = 18 k and PI = 4k7 (preset). - assume that the correct AFC voltage is 5.6 V (as for the CA 3089!). In this case the voltage across the upper half of the divider chain should be approximately 6.5 V; reason- able values are R2=12k and R3 = 2k2. R4 + V4P1 should be approximately 10 k, so R4 can be 8k2 and PI can be 4k7. Note that R3 sets the sensitivity of the indicator, whereas PI is used for correct calibration. Some FM detectors, notably ratio discrimi- nators, give a 0 V output when correctly tuned. In this case the circuit shown in figure 2 can be added, between the AFC output and the input to the circuit shown in figure 1 . m ohmmeter LED peak indicator elektor july/august 1977 Using a CA 3140 FET op-amp it is easy to construct a simple, linear-scale ohmmeter. The op-amp is connected in the non- inverting mode, with the non-inverting input fed from a 3.9 V zener. The op-amp output voltage is thus given by Since one end of the meter is returned to the zener the meter voltage is R x R 2 Ry _J: X 39 + — x39 — 39ie _ x 3 9 R 2 R 2 R 2 Since the zener voltage and R2 are fixed the voltage measured by the meter is pro- portional to R2. The full-scale deflection of the meter is about 3.9 V, but the exact value will depend on the tolerance of the zener. Three ranges are provided by using different values of R2. With R2 = 1 k the full-scale reading of 3.9 V is obviously obtained when R x = 1 k. With R2 equal to 10 k and 100 k full-scale readings are obtained at 10 k and 100 k. The voltmeter is simply a 1 mA meter with a nominal 3k9 series resistor, so the 0 to 1 mA scale can easily be converted to read 0 to 1 k. 0 to 10 k and 0 to 100 k. The germanium diode connected across the meter protects it in the event of an overload. To calibrate the meter it is first necessary to zero it by nulling the op-amp offset voltage. To do this P2 is first set to minimum resist- ance to make the meter most sensitive, and a wire link is connected across the R x ter- minals. PI is then adjusted to give a zero reading on the meter. The meter may then be calibrated by con-Vr, necting a close tolerance resistor of known*' -. value (e.g. 1 00 k 1 %) across the R x terminals and adjusting P2 until the meter reads correctly. To ensure good accuracy on all ranges R2, R2' and R2" should be close tolerance types, 2% or better. The maximum value which can be used for R2 and hence the maximum value of R x that can be measured depends on the input resistance of the op-amp, since any current flowing into the op-amp input will cause errors. However, with the 1.5 Tfi input resistance of the 3140 it should be possible to use values up to 10 M, assuming 10 M close tolerance resistors can be obtained. A disadvantage of the VU meters frequently used in hi-fi equipment is their inability to respond to short transients, which can lead to overrecording when used with tape recorders, and clipping when used with power amplifiers. Even many would-be peak- reading meters fail to cope with transients. Since inertia-free mechanical meters do not exist the obvious solution would seem to be a completely electronic indicator to indicate the onset of overload, and indeed many cassette recorders are now fitted with peak indicator lamps. The circuit described here, for a stereo peak indicator, uses only one IC and a few other components. The circuit is based on a 3900 quad Norton amplifier. Two of the amplifiers in the IC are used as comparators. Since the inputs of a Norton amplifier are current fed a resistor must be inserted in series with each input to allow them to be voltage driven. When the left channel input voltage exceeds the volt- age at the slider of PI the output of A1 will swing negative. This triggers a monostable multivibrator built around A2, which causes the LED D2 to light for a few hundred milli- seconds, no matter how short the transient. In the event of a continuous overload the monostable will be retriggered continuously and the LED will appear to be continuously lit. The circuit of the right channel, com- prising A3 and A4, operates in exactly the same fashion. A suitable power supply for the circuit is given in figure 2. Figure 3 shows a printed circuit board and component layout for the indicator. The bridge rectifier for the power supply is not mounted on the board since, if the indicator is built into a piece of existing equipment it is quite likely that a suitable unregulated DC supply may already be available. PI and P2 are simply connected direct to the positive supply line of the equipment. If the indicator is to be used with a tape recorder then calibration is quite simple. With the tape deck set to record a 1 kHz signal can be fed in and the record level adjusted until the record level meters move into the red. PI and P2 may then be adjusted until D2 and D4 just light. If the indicator is to be used with a power amplifier then it can be calibrated in several ways. If an oscillator and oscilloscope are available the amplifier can be fed with a 1 kHz signal which is adjusted so that the output of the amplifier is just below the onset of clipping. PI and P2 can then be adjusted until D2 and D4 light. If no test gear is available then the peak output voltage of the amplifier may be calculated from the equation Vpeak = V2WR where W is the amplifier output power and R is the specified load impedance. The slider voltage of PI and P2 can then be adjusted to just below this value, using a multimeter of not less than 20,000 fi/V. elektor july/august 1977 transistor solar cell H. Januschkowetz elektor july/august 1977 In order to receive an SSB signal it is first necessary to recombine the carrier wave (which was suppressed in the transmitter) with the SSB signal. The simplest way of doing this is to feed an oscillator signal to the diode detector in the receiver. However a much better solution is to use a separate demodulator for SSB signals. The diagram shows a design for a demodulator which will adapt any shortwave receiver with an inter- mediate frequency of 455 kHz for SSB reception. The advantage of a dual-gate MOSFET is that it can function simultaneously as a demodulator and an oscillator. Such a circuit is in effect a self-oscillating mixer stage, with the difference that the i.f. output signal has the same frequency as the oscillator signal. As is usual with ‘direct conversion’ (i.e. oscillator and input signal have the same frequency), there is the danger that the oscillator will be pulled off frequency by the input signal (forced resonance). With most dual-gate MOSFETs however, this sort of back-coupling is reduced to negligible proportions. On the other hand it is also possible to take advantage of this effect and (as is the case here) construct an SSB- demodulator which can also be used as a synchronous demodulator for AM. As is apparent from the diagram, the lower section (gate 1 and source) of the MOSFET is connected as a Clapp-oscillator with a frequency of 455 kHz; the upper section functions as a mixer stage for the oscillator and input signals. The optimum input level for SSB is approx. 15mVpp, when the out- put level is lOmVpp. For synchronous AM detection the input level has a value of 300 mVpp, and with 30% modulation the value of the output signal is approx. 60 mVpp. The circuit can be switched from AM to SSB as indicated in the block diagram. The adapter consumes approx. 2.3 mA from a 12 V supply. The supply voltage depen- dence of the oscillator frequency is approx. 35 Hz/volt. The synchronous range for AM is approx. 1 kHz. The measurement points shown in the diagram facilitate adjustment procedures. to 20 mA. The graph shows output voltage plotted against load current. As the area of the silicon chip is small compared to a normal solar cell a magnifying glass may be used to focus light onto the junction and so increase the output current. However, this is not to be recommended in very strong sunlight or the junction may be destroyed! If a good transistor is used then the output current may be doubled by connecting the collector-base and emitter-base junction in parallel, as shown in the diagram. This should not be done with faulty transistors, however, since if the faulty junction is short-circuit it would short the output of the solar cell. Warning: It is not advisable to use old Germanium power transistors, since these may contain highly poisonous substances. However, a major semiconductor manu- facturer has assured us that the more modern silicon devices, such as the 2N3055, are completely safe. Many circuits using, for example, both op- amps and logic circuits, require more than one supply voltage. The circuit described here is designed to supply four voltages of + 1 2, +5, -7 and - 1 2 volts, with a maximum current of 50, 300, 50 and again 50 mA respectively. The positive supply voltages are produced in the normal fashion, using positive voltage regulator ICs; for the negative voltages it would be possible to use the special ICs which have been designed for this purpose, however these are both fairly expensive and often difficult to obtain. For this reason an alternative solution was sought. Although the 723 was designed for positive voltages, it can also be adapted for negative output voltages if, instead of being used as a series- regulator, it is connected as a shunt stabiliser (IC3 and IC4). Shunt stabilisers suffer from the dis- advantage that a constant power is taken from the mains transformer, irrespective of whether they are feeding a load. This means that this type of circuit is not particularly efficient; however in this case, where the maximum current is only 50 mA, the power loss is negligible. The negative output voltages can be adjusted by means of PI and P2. After adjustment, the series-connected potentiometer and resistor can be replaced by two series- connected resistors. All the voltages supplied by the circuit are short-circuit proof; that is to say that shorting the outputs will not damage the supply. The positive outputs are provided with the usual current limiting. In the case of the shunt regulators for the negative voltages, the short-circuit current is determined by the dropper resistors R7 and R1 3. These should be rated at 2 W (or more) to prevent overheating. Note that it will not always be necessary to use such a complicated transformer (8-0-8-16 V). If the 5 V supply does not have to deliver much current, a 0-8-16 V (i.e. an 8-0-8 V!) transformer can be used. D2 and C2 are omitted in this case. 77050 012V ■74 KQ* I distortion meter elektor july/august 1977 The standard method of measuring the har- monic distortion of an amplifier is to feed the amplifier with a pure sinewave signal and to feed the distorted signal from the ampli- fier output into a notch filter which rejects the fundamental, leaving only the harmonic distortion products introduced by the ampli- fier, which can then be measured or exam- ined on an oscilloscope. A twin-T network tuned to the fundamental frequency of the signal will provide a very large degree of attenuation of the fundamen- tal. However, since the Q-factor of a twin-T network is fairly low the harmonics, in par- ticular the second, will also be attenuated, giving a too optimistic picture of the distor- tion. A solution to this problem is to include the twin-T network in an active filter arrange- ment as shown in figure 1 . Instead of con- necting point A to ground in the normal IF T1+T2 2 fashion, bootstrapping is applied via emitter follower T3. This increases the Q of the network by a factor -5—, which virtually K9 eliminates attenuation of the 'second and other harmonics. The complete circuit of a distortion filter using this principle is given in figure 2. The twin-T network consists of resistors R3 to R7, capacitors C2a to C5b and poten- tiometers PI to P6, which allow fine tuning of the filter. The filter is not designed to be tuned over the entire audio spectrum since it is intended mainly for check distortion at spot frequencies of 100 Hz, 1 kHz and 10 kHz. To this end capacitors C2a to C5b are mounted on a plug-in module so that they can easily be changed. Capacitor values for frequencies of 100 Hz, 1 kHz and 10 kHz are given, but values for other frequencies can easily be calculated from the equation: f ° 2*RC C = 2?f^R The nominal value of R is 10 k, and the equation gives the combined value of C4a / C4b which equals C5a/C5b. The combined value of C2a to C3b is twice this. The potentiometers are used for fine tuning the filter and give a frequency adjustment of about ± 10%. For distortion measurements on low impedance sources such as power amplifiers the signal can be fetf direct to the twin-T network with SI in position 2. The input impedance of the twin-T network at f 0 is about 7 k. For measurements on circuits with a high output impedance a ‘super emitter-follower’ buffer stage T7/T8 is provided. T9 is a constant current emitter load for this stage. If this facility is not required then T7 to T9, R19, R20 and Cl 2 may be omitted. The output of the twin-T network is fed to a second super emitter-follower stage T1/T2. Bootstrapping from the junction of R9 and RIO is applied to the twin-T network by T3. The distortion signal is available at output B1 and also, amplified ten times, at output B2. To use the distortion meter the amplifier to be tested is fed with a low distortion sine- wave signal at the required frequency. With the filter out of circuit the amplifier output is monitored on a ‘scope’ and the signal level is adjusted to give a suitable output from the amplifier. With all the poten- tiometers in their mid-positions the filter is switched in and the signal frequency adjusted slightly to give the minimum signal at output Bl. The potentiometers can then be adjusted to null out the fundamental still more until no further reduction can be obtained. The peak value of the distortion is then given by 100 %. v in p-p If the scope is not sufficiently sensitive to measure the distortion using the B 1 output then the B2 output may be used. In this case the distortion is given by v out p-p ^in p-p x 10%. However, during the initial adjustment stages where a large proportion of the fundamental is present, the Bl output should be used as the B2 output will probably be driven into clipping. If two distortion filters are cascaded then it is possible to measure distortion figures as low as .005%. By slightly offsetting the null frequencies of the two filters it is also poss- ible to obtain a wider notch so that the effect of oscillator frequency drift is not so troublesome. In this case, if the B2 output of the second filter is used the percentage distortion will be given by VtP -P-x 1%. v in p-p A printed circuit board and component layout for the filter are given in figure 3 (EPS 77005). Capacitors C2a to C5b are mounted on a dual-in-line component module that plugs into a DIL socket on the board. elektor july/august 1977 \03 LC resonance meter \Q» drill speed control J. Becela elektor july/august 1977 This circuit is intended to perform the same function as a conventional grid-dip meter i.e. measurement of the resonant frequency of LC-tuned circuits. Unlike a normal grid-dip meter it is not, in itself, a complete instru- ment, but can be used in conjunction with a frequency counter to give a direct reading of resonant frequency. The circuit consists of a difference amplifier comprising T1 to T4 and a pair of coils LI and L2. These coils are wound on the same former but are spaced apart so that, when the circuit is not coupled to an LC circuit, no spontaneous oscillation occurs. When the coils are brought close to an LC circuit then - 1 oscillation will occur at the resonant fre-r quency of the LC circuit, and this frequency > can be measured by feeding the collector signal of T3 to a frequency counter. No direct connection to the LC circuit is required. It must be stressed that this is a design idea that has not been fully developed but is printed here for the benefit of the exper- imenter. In consequence no constructional details are given for LI and L2. Most drill speed controllers suffer from one or more drawbacks. These include poor speed stability, excessive instability at low speeds, and high power dissipation in the series resistor used to sense motor current. The circuit described here suffers from none of these drawbacks, and in addition is extremely simple. The mains input is rectified by D1 and dropped by Rl. The current drawn by T1 can be controlled by means of PI, thus also controlling the DC voltage that appears across C2, and hence at the base of T2. T2 is connected as an emitter follower, and the voltage appearing at the cathode of D3 is about 1 .5 V less than the base voltage of T2. Assuming that the motor is turning but that the triac is turned off, the back e.m.f. generated by the motor will appear at the T1 pin of the triac. So long as this voltage exceeds the cathode voltage of D3 the triac will remain turned off, but as the motor slows down this voltage will fall and the tria^ will trigger. If the load on the motor increases, thus tending to slow it down, the back e.m.f. will fall more quickly and the triac will trigger sooner, thus bringing the motor back up to speed. Since the triac • can be triggered only on positive half-cycles of the mains waveform the controller will not vary the motor speed continuously from zero to full speed, and for normal full-speed running SI is included, which turns the triac on permanently. However, the circuit exhibits good speed control characteristics over the important low speed range. LI and Cl provide suppression of r.f. interference generated by the triac. LI can be a commercially available r.f. suppression choke of a few microhenries inductance. The current rating of LI should be from two to four amps, depending on the current rating of the drill motor. Almost any 600 V 6 A triac can be used in the circuit. Using the 3V2-digit-A/D-converter LD110/ 1 1 1 and a minimum of external components, it is possible to construct a universal digital voltmeter. The accuracy of the meter is approx. 0.05% + 1 digit. The scale of the meter runs from -2 V ... +2 V and by means of additional voltage dividers can be extended as required. The polarity of the measured voltage is indicated by the sign in front of the highest digit. When the range of the meter is ex- ceeded all 7 segments will flash on and off (over-range indication). If no voltage is present at the input then the meter auto- matically indicates 0 V (auto-zeroing). The circuit board is designed for use with Hewlett-Packard 7-segment displays 5082- 7730/5082-7732 or 5082-7750/5082-7752, although other pin-compatible common- anode displays may also be used. The input resistance of the circuit is greater than 1 M, and the input current is approx. 4 pA. The reference voltage is produced by the FET-constant current source T6 and a transistor which is reverse biased and used as a zener diode*. Construction and calibration It is recommended that the meter is used with a stabilised voltage supply. A cermet trimmer should be used for potentiometer PI. After applying the supply voltage, the input should be short-circuited. Zero adjust- ment is then carried out using trimmer capacitor C5. Finally the meter is calibrated by means of PI against a standard voltage. The decimal points of the displays are brought out separately. It should be noted that the series resistors in the cathode connections are not shown on the printed circuit board and should be added as required. (Siliconix Application) * Note that for optimum performance T6 and T7 ought to be selected types. A simpler solution is to use a 1 k resistor instead of T6 and a 5V6 volt- age reference diode instead of T7. ° O ^ § m ■a o> 3 CL O 3 "O o ■1 elektor july/august 1977 tup-tun-dug- dus elektor july/august 1977 — 7-79 Wherever possible in Elektor circuits, transis- tors and diodes are simply marked 'TUP' (Transistor, Universal PNP),'TUN' (Transistor, Universal NPN), 'DUG' (Diode, Universal Ger- manium) or 'DUS' (Diode, Universal Silicon). This indicates that a large group of similar devices can be used, provided they meet the minimum specifications listed in tables la and 1b. For further information, see the article 'TUP- TUN-DUG-DUS' in Elektor 1, p. 9. type Uce 0 •c hfe Ptot fT max max min. max min. TUN NPN 20 V 100 mA 100 100 mW 100 MHz TUP PNP 20 V 100 mA 100 100 mW 100 MHz Table la. Minimum specifications for TUP and TUN. Table 1b. Minimum specifications for DUS and DUG. type UR max IF max IR max Ptot max CD max DUS Si 25 V 100 mA 1 /JA 250 mW 5 pF DUG Ge 20 V 35 mA 100 IdA 250 mW 10pF Table 2. Various transistor types that meet the TUN specifications. ‘tun BC 107 BC 208 BC 384 BC 108 BC 209 BC 407 BC 109 BC 237 BC 408 BC 147 BC 238 BC 409 BC- 148 BC 239 BC 413 Bd 149 BC 317 BC 414 BC 171 BC 318 BC 547 BC 172 BC 319 BC 548 BC 173 BC 347 BC 549 BC 182 BC 348 BC 582 BC 183 BC 349 BC 583 BC 184 BC 382 BC 584 BC 207 BC 383 Table 3. Various transistor types that meet the TUP specifications. TUP BC 157 BC 253 BC 352 BC 158 BC 261 BC 415 BC 177 BC 262 BC 416 BC 178 BC 263 BC 417 BC 204 BC 307 BC 418 BC 205 BC 308 BC 419 BC 206 BC 309 BC 512 BC 212 BC 320 BC 513 BC 213 BC 321 BC 514 SC 214 BC 322 BC 557 BC 251 BC 350 BC 558 BC 252 BC 351 BC 559 Table 4. Various diodes that meet the DUS or DUG specifications. DUS DUG BA 127 BA 217 BA 218 BA 221 BA 222 BA 317 i BA 318 BAX 13 BAY 61 1N914 1N4148 OA 85 OA 91 OA 95 AA 116 Table 5. Minimum specifications for the BC107, -108, -109 and BC177, -178, -179 families (according to the Pro-Electron standard). Note that the BC179 does not necessarily meet the TUP specification Oc.max = 50 mA). NPN PNP BC 107 BC 177 BC 108 BC 178 BC 109 BC 179 v ce 0 45 V 45 V max 20 V 25 V 20 V 20 V v eb 0 6 V 5 V max 5 V 5 V 5 V 5 V 100 m A 100 mA max 1 00 mA 100 mA 100 mA 50 mA p tot. 300 mW 300 mW max 300 mW 300 mW 300 mW 300 mW 'T 150 MHz 130 MHz min. 150 MHz 130 MHz 150 MHz 130 MHz F 10dB 10 dB max 10 dB 10 dB 4 dB 4 dB The letters after the type number denote the current gain: A: a' (0. h fe ) = 125-260 B: a' = 240-500 C: a' = 450-900. Table 6. Various equivalents for the BC107, -108, . . . families. The data are those given by the Pro-Electron standard; individual manu- facturers will sometimes give better specifi- cations for their own products. NPN PNP Case Remarks BC 107 BC 108 BC 109 BC 177 BC 178 BC 179 ■Q BC 147 BC 148 BC 149 BC 157 BC 158 BC 159 1 c ' ■g Pmax = 250 mW BC 207 BC 208 BC 209 BC 204 BC 205 BC 206 ■© BC 237 BC 238 BC 239 BC 307 BC 308 BC 309 c *0 BC 317 BC 318 BC 319 BC 320 BC 321 BC 322 (Di •crnax = 150 mA BC 347 BC 348 BC 349 BC 350 BC 351 BC 352 m BC 407 BC 408 BC 409 BC 417 BC 418 BC 419 ■Gl Pmax = 250 mW BC 547 BC 548 BC 549 BC 557 BC 558 BC 559 Pmax = 500 mW BC 167 BC 168 BC 169 BC 257 BC 258 BC 259 ®! 169/259 Icmax = 50 mA BC 171 BC 172 BC 173 BC 251 BC 252 BC 253 0 251 . . .253 low noise BC 182 BC 183 BC 184 BC 212 BC 213 BC 214 ■0 *cmax = 200 mA BC 582 BC 583 BC 584 BC 512 BC 513 BC 514 0. Icmax = 200 mA BC 414 BC 414 BC 414 BC 416 BC 416 BC 416 0; low noise BC 413 BC 413 BC 415 BC 415 0 low noise BC 382 BC 383 BC 384 ■0 BC 437 BC 438 BC 439 Pmax = 220 mW BC 467 BC 468 BC 469 01 Pmax = 220 mW BC 261 BC 262 BC 263 •Q low noise 7-80 — elektor july/august 1977 transistors OObLIN 4011 prunsrunsnerwr TuuniniTiiriijTinir iininiJiiniriinir TiniJiirMiiriinTr ‘rTT" '■ " » » » <■«> grlrviiv fe! to Data Inputs 2 + Pin-compatible CMOS equivalents available from Teledyne Semiconductor and National Semiconductor Beginners Digitals, Nor, Latching etc. ELEKTOR MAGAZINES BINDERS PRINTED CIRCUIT BOARDS Electronic books, Projects, data etc. Crystals & Sockets Piezo Electric Ceramics PHIURQN Electronic Components Specialists 325 Dl ITOIT STREET, P.O. BOX 2749, PRETORIA, 0001 TRANSVAAL. REPUBLIC OF SOUTH AFRICA Telegraph address "T RINIT RON" ArKlUA ^ Stockists of Heathkit - Siemens — Monsanto — Texas, — Interface — Linear — MOS - Signetic? Fairchild °RODUCTS Capacitors Tantalum \ Polyester £_ Electrolytic — I Co-axial & J P.C. Mount | LED'S U a Hit u ?anel Meters