Fullscreen HTML5Med Res (??MB)HTML4

Simple drivers for radio control servos Build one of these simple servo drivers & you can run the devil out of your servos. You can use them for testing servos or for direct control applications where a radio link is not required. The circuit parts are cheap and readily available. By NENAD STOJADINOVIC As anyone who has been reading Bob Young’s excellent radio control column in this magazine will know, servos are the muscle behind any radio control system. These devices are a minor elec­tronic miracle: small, powerful and cheap, but until now have always been lumbered with a radio control system to drive them. Think of how useful they would be if you could drive them direct­ly from a simple pot or pair of pots controlled by a joystick. These were my thoughts, one dark and stormy night, as I was casting about for a good way of remotely controlling a pair of mirrors to be used in a laser light show. A quick perusal of some modelling magazines and the current circuit was born. After fashioning some suitable metalwork I 24  Silicon Chip 2.1ms 30ms soon had laser beams flying about the lab with gay abandon. Lately, I’ve been using servos in place of mechanical link­ages in my car and at around $20 per servo there is little in­ centive to fiddle around with cables, rods and so on. What’s more, running controls into areas of very high or low pressure is made easy by the availability of watertight bulkhead electrical connectors from your friendly local marine chandler. Anyway, whatever our field of endeavour, that old worn-out cliche about the applications are only limited by your imagina­tion must surely apply. So without further ado, on with the circuit. How it works The standard servo has three input This photograph shows the author’s prototype of the circuit featured in Fig.2. Note that the final version differs from this prototype in terms of board layout. 0.7ms Fig.1: the control signal for a servo consists of a continuous fixedfrequency pulse stream. The pulse width controls the servo position. pins and these are +5V (power), 0V (GND) and control. The control signal is a continuous pulse stream which is shown in Fig.1. It is important to note that the frequency of these pulses does not intentionally vary (it is not critical) and a period of between 20 and 50ms will do the job with most servos. The movement information is contained in the width of the pulses, which is why this sort of control system is referred to as Pulse Width Modulation (PWM). A pulse width of 0.7ms will usually give fully coun­ terclockwise movement and 2.1ms will give fully clock­wise rotation. This alternative version is based on the circuit shown in Fig.5. Once again, the final version differs in layout from this prototype – see Fig.6. 82k 11 180k 2 4 14 IC1a NE556 6 5 3 7 VR1 10k LIN Q1 100k VN10KM D 0.1 .01 G S 2.7k +5V PARTS LIST VR3 100k Circuit One (Fig.2) VR2 5k 100 .01 .01 10 13 0.1 .01 9 IC1b 12 DG S VIEWED FROM BELOW 1 PC board, code 09105942 1 556 dual timer (IC1) 1 VN10KM Mosfet (Q1) 1 10kΩ linear potentiometer (VR1) 1 5kΩ trimpot (VR2) 1 100kΩ trimpot (VR3) OUTPUT 11 8 .01 .01 0.22 Fig.2: the pulse frequency for the servo driver is derived using astable oscillator IC1a. Its output at pin 5 is differentiated & then used to trigger monostable IC1b via buffer stage Q1. VR1 varies the pulse width produced by IC1b. The duration of this output pulse is set by potentiometer VR1 which is calculated to give the required 0.72.1ms range. Taming the duty cycle As presented, the free-running oscillator has a duty cycle of about 60%, meaning that its positive output pulses will be about 18ms long. This is much longer than can be used to trigger the 556 monostable (IC1b), so some means had to be used to obtain short negative pulses. My solution was to differentiate the oscillator output and this produces a series of positive and negative spikes about zero volts at every transition. These spikes don’t have much energy and so are buffered by a Mosfet (Q1) which has a very high input impedance. Being an N-channel device, it only conducts on the positive pulses and so produces negative-going pulses at its drain (D). These pulses are coupled to pin 8 to trigger the mono­stable. It produces 100k VR2 1 +5V 1 PC board, code 09105941 1 4011 or 4001 quad gate package (IC1) 1 1N914 signal diode (D1) 1 0.1µF MKT capacitor 1 10kΩ linear potentiometer (VR1) 1 10kΩ trimpot (VR2) Resistors (0.25W, 1%) 1 1.8MΩ 1 1kΩ 1 150kΩ short positive pulses which can be set to vary between 0.7 and 2.1ms long. Building the circuit A small PC board was designed to accommodate the components and this is shown in Fig.4. With only a handful of components, construction is very simple. You could use a small piece of Veroboard as an alternative to a PC board. OUTPUT 0.1 GND 180k .01 VR1 Circuit two (Fig.5) Q1 0.1 2.7k IC1 556 82k VR3 Resistors (0.25W, 1%) 1 180kΩ 1 82kΩ 1 100kΩ 1 2.7kΩ 0.22 .01 100  The circuit to achieve this uses a free-running oscilla­ t or coupled to a monostable or “one shot”, as shown in Fig.2. It is based on a 556 dual timer which can be regarded as two 555 timers in the one package. The free-running oscillator has its frequency determining components connected to pins 1, 2 and 6 and these give a frequency of around 34Hz, corresponding to a period of about 30 milliseconds. The oscillator output is taken from pin 5 and it is used to trigger the monostable section of the cir­cuit. The monostable or “one shot” is the second half of the 556 and its pulse length is determined by the components connected to pins 12 and 13; ie, trimpots VR2 & VR3, control pot VR1 and the 0.22µF capacitor. The output pulse stream appears at pin 9. A monostable produces an output pulse of programmable duration each time it is triggered, the only proviso being that the trigger pulse must be of shorter duration than the output pulse. Capacitors 1 0.22µF MKT capacitor 2 0.1µF MKT capacitor 2 .01µF MKT capacitor GND Fig.3: install the parts on the PC board as shown in this diagram, taking care to ensure that the IC is oriented correctly. Fig.4: the full-size etching pattern for the PC board. It is coded 09105942 & measures 51 x 40mm. May 1994  25 SATELLITE SUPPLIES Aussat systems from under $850 +5V 1 4001 IC1a 2 1.8M 1.8M ON CONTROL VR2 10k FEEDHORNS C.BAND FROM .........$95 150k 150k 4 IC1b 8 9 IC1c 10 12 13 14 IC1d 11 1k OUTPUT 7 0.1 .01 D1 1N914 LNB’s Ku FROM ..............................$229 FEEDHORNS Ku BAND FROM ......$45 6 VR1 10k SATELLITE RECEIVERS FROM .$280 LNB’s C FROM .................................$330 5 3 Fig.5: this alternative circuit from Bob Young was originally featured in the April 1993 issue. IC1a & IC1b form an astable oscillator, with pulse width set by VR1, VR2 and the 0.1µF capacitor. DISHES 60m to 3.7m FROM ...........$130 0V +5V 1k 1 OUTPUT 0.1 IC1 4001 150k 1.8M VR2 D1 VR1 Fig.6 (left): the circuit of Fig.5 is assembled as shown in this diagram. Note that the output frequency also varies with this unit but not enough to affect servo operation. Fig.7 at right shows the full-size etching pattern for the PC board. LOTS OF OTHER ITEMS FROM COAXIAL CABLE, DECODERS, ANGLE METERS, IN-LINE COAX AMPS, PAY-TV DECODER FOR JAPANESE, NTSC TO PAL TRANSCODERS, E-PAL DECODERS, PLUS MANY MORE For a free catalogue, fill in & mail or fax this coupon. ✍     Please send me a free catalog on your satellite systems. Name:____________________________ Street:____________________________ Suburb:_________________________ P/code________Phone_____________ L&M Satellite Supplies 33-35 Wickham Rd, Moorabin 3189 Ph (03) 553 1763; Fax (03) 532 2957 26  Silicon Chip The lead for your particular servo can usually be obtained from any good hobby shop. If not, just ask who repairs that particular brand and call them. While you’re at it, find out how the servo pins are arranged; the manual might have the informa­tion. If not, it’s simply a matter of checking the output voltag­es from the receiver with a multimeter. Ground and +5V should be fairly obvious and the lead with some small voltage will be the control output. Adjustment The servo travel limits are adjusted by VR2 and VR3. VR2 should be the anticlockwise limit and this is set by moving VR1 fully anticlockwise and then adjusting VR2 so that the servo is not stalled. It is important not to stall servos because they will draw high currents and get very hot, ultimately burning out the motor. The clockwise limit is then set in the same way using VR3. Second circuit Having designed the above circuit, I then came across a small circuit from Bob Young that does the same job as mine! It was featured in the April 1993 issue of SILICON CHIP. The obvious solution, of course, was to present his circuit as well, complete with a PC board and the addition of the suggested trimpot, VR2. Bob’s version is shown in Fig.5 while the PC component wiring diagram is shown in Fig.6. Take care to ensure that the IC and diode are correctly oriented during the PC board assembly. This workings of this circuit are less apparent than the circuit shown in Fig.2 but essentially IC1a and IC1b are connected as a free-running oscillator with an uneven duty cycle. The pulse duration is mainly a function of VR1, VR2 and the 0.1µF capacitor. There is also an essential difference in its operation in that when you change the settings of VR1 to set the pulse output, the frequency changes too, although not markedly. However, this does not affect the servo operation at all and so the circuit is quite valid SC for test purposes.