Fullscreen HTML5Med Res (??MB)HTML4

REMOTE CONTROL BY BOB YOUNG How to service servos & winches, Pt.2 Some servos must be regarded as throwaway items not able to be serviced but even defunct servos can be cannibalised to keep others going. This month, we continue with this topic & include an interesting if unusual description on how these cir­cuits work. The development of the modern integrated circuit servo goes back a long way. To my knowledge, Orbit Electronics in America commissioned the first IC amplifier in about 1969. I remember bringing back a bag full of them from the World Aerobatic Cham­pionships in 1971. They were a stunning innovation at the time, replacing an 11-transistor discrete amplifier which chewed up lots of space in the servo. As a result of this chip, Orbit introduced the PS-4 servo which rocked the R/C world at the time. We had never dreamed of servos so small. These days they still look small but the new miniature servos are an or- Fig.1: up until a few years ago, the Signetics NE544 was used in many servos but it is no longer available. 88  Silicon Chip der of magnitude down in size again and make them look very ordinary. However, in 1970 they were simply amazing. By modern standards the Orbit amplifier was not very good and it was prone to several shortcomings no longer encountered in the modern IC servo. They tended to dither around neutral due to the dead band being too small. This tended to raise servo current and made the neutralising a little less precise than it should have been. They also exhibited temperature drift which we ul­timately cured with the addition of a diode in the feedback path. They were also inclined to non linearity, a very serious fault in a servo. New ICs followed in quick succession as other manufacturers scrambled onto the bandwagon in order to maintain their position in the R/C marketplace. Each learned from the preceding and gradually the flaws disappeared. With the arrival of the NE544 IC (Fig.1), servo design came of age but not however without a struggle! The first two NE544 masks were duds. The “B” mask in particular was prone to latch up on the output bridge. This resulted in a short circuit through the bridge and the heat generated usually blew the top off the IC. With the arrival of the “C” mask, all difficulties and de­fects were overcome and I had a long and happy association with this amplifier. Linear, accurate and reliable, it was all we could ask for in a servo amplifier. Then they went and discontinued it. Why do manufacturers do these things? This is particularly upsetting when they offer no direct replacement and people with equipment designed around these devices are left stranded with no alternative. The Japanese in the meantime were pressing on with their own development and Futaba came up with a twochip solution, as shown in Fig.2. The logic was in one chip and the bridge in another. These were also prone to blowing out the side of the chip. (They were vertical mount). Manufacturers do tend to skimp on epoxy at times. They seem to forget that electronic devices are driven by smoke under pressure and that if the case ruptures and the smoke escapes, then the device will no longer function. (Editor’s note: we are indebted to Bob Young for this illuminating explanation of the workings of electronic components. Designing circuits will now be so much easier!) Futaba finally came up with a new single chip solution (with thicker epoxy?) and went on to produce some very popular and reliable servos. These days surface mount components have reduced the servo amplifier to a mere shadow of the old massive shoulder-to-shoulder discrete servo amplifiers. A couple of SM resistors and capacitors and a teensy IC, and that is it. Not like the good old days at all. From a servicing point of view, there have been some nasty techniques introduced by modern assembly methods which make servicing very tedious. Mounting the servo motor directly on the PCB is probably the nastiest. This means complete stripping of the servo to get inside to the amplifier. The completely sealed pot which has now become a throwaway item is another although the scales seem to have been balanced by the improved reliability of these pots. All in all, there is little that can be serviced in the modern servo and one must be careful not to be drawn into servic­ing something that is really a throwaway item. All you can do with cheap servos is to cannibalise them for parts. Fig.2: Futaba’s first integrated circuit servo used two ICs, one for the logic & the other for the motor drive bridge. the servo for crash damage, etc (see last issue) and then plug the opened servo into the analyser. Servo neutral and travel length are checked against the manufac­turer’s specs and a note taken of the servo motor current con­ sumption. Modern servo standards usually call for 1-2ms at extremes with a 1.5ms neutral. The old Futaba was 0.65-1.90ms and I have seen examples of sets swinging around 1.2-2.5ms although this was rare. If in doubt, check the manufacturer’s specifications – if you can get them, that is. with a lint free cloth and a smear of Vase­line will help minimise wear on the track. Check the wiper for tension and cleanliness. Servo motor current consumption is a bit of a headache as motors of various types draw widely differing current. Your best bet is to note the current on a new servo and use it as a guide. What you are looking for is a marked increase in motor current when the motor is free running. Typically, a new 11Ω permag motor will run free (unloaded) at around 80-100mA. As “All in all, there is little that can be serviced in the modern servo and one must be careful not to be drawn into servic­ing something that is really a throwaway item. All you can do with cheap servos is to cannibalise them for parts”. Routine chores That said, there are routine chores which should be carried out regularly and there are some not so routine techniques which may be helpful in unusual circumstances. At Silvertone, we use a servo analyser which consists of a pulse generator with variable rate auto-sweep, a pulse width meter with LED display and a current meter. We begin the service with a visual inspection of The servo is then checked for smoothness over the entire arc of travel. This will pick up any flaws in the potentiometer track. If the servo jumps or dithers around one spot this usually indicates a hole in the track or a dirty pot element. The pot is replaced or cleaned as appropriate and re-neutralised. Where it can be done, pots are cleaned routinely as part of a general service. Clean them they age, the current will creep up, sometimes to as high as 300mA. The causes are many and varied and include dry sintered bronze bearings, dirty commutator, bent shaft, broken brushes and sometimes pin­ions which have been pushed against the bearings. Lubrication Some motors can be stripped down for inspection and repair or cleaning, May 1994  89 REMOTE CONTROL – Servicing the servos and some cannot. A simple way to brighten up a tired motor is to start the motor running and spray CRC-226 onto the bearings (front and rear). This will soak into the bronze bushes and some will work its way into the commutator, cleaning it as well. Be sure to run the motor in both directions for about five minutes. I have seen motors respond well to this treatment, with current consumption falling from around 170mA to 100mA. In a model with four servos, this amounts to a significant reduction in battery load. Motor problems Motor problems are still a worry due to engine vibration pounding away at delicate brushes. One rule to remember in this regard is to never connect the case of the servo motor directly to ground. If the armature insulation breaks down then you have a dead short from motor drive positive to ground. Scratch one servo amplifier. The servo amplifier IC case will rupture immediately, fill the model with escaping smoke, blind whilst monitor­ing the detector output with an oscilloscope. Noisy servos will show up as noise bursts in the sync pause or even obliterated control pulses. If you lack a scope, then remove the Tx antenna and do a range check. With the model at the extreme of control­lable range, move each servo in turn. A noisy servo will kill the signal and control will be lost, whereas good servos will function normally. Once you are satisfied that all is well electronically, reassemble the servo and check for smooth operation and that the servo neutral and current draw have not changed. Sometimes tight­ening the screws in the servo case can distort the case or load the gears, causing increased current drain. Finally, before we close on the servo, servicemen are often asked routine questions or required to perform non-standard modifications on servos for special projects. Here are a few hints on these problems. Reversing servos One common inquiry is how do you reverse a servo. Virtually all transmit- poten­tiometer element. Do not touch the wiper wire. The tricky bit comes about because the wiper usually does not sit in the centre of the pot element. Therefore, after reversing, the wiper must be reset so that the angle/resistance (whichever is easier for you to meas­ure) is the same between the wiper and whichever of the two leads you used as a reference. Once the wiring has been swapped, reset the wiper so that the angle/resistance between the wiper and the reference end or colour is the same as the original. If you get this wrong, the servo will jam at one end of the track, possibly damaging the gears, amplifier or the motor. Plug the servo in and switch the Tx and Rx on. The servo should take up a position roughly around neutral and work in the reverse rotation. Reset neutral and seal up the servo. Stay alert during this procedure and take the phone off the hook. If you get interrupted, all hell can break loose. I have seen people come into my workshop with servos in pieces and unable to get normal operation in either direction. Usually they have reversed one pair of wires and not the other, forgotten to reset the wiper angle or moved the wiper wire by accident. Changing angle of rotation “One rule to remember is to never connect the case of the servo motor directly to ground. If the armature insulation breaks down then you have a dead short from motor drive positive to ground”. the pilot and almost certainly result in a crash. Checking for interference Finally, one last note on servo motors. One common cause of radio interference is servo noise getting into the receiver. This can be caused by a dirty commutator but is more usually caused by broken noise suppression capacitors on the servo motor. These sometimes get broken in a crash and can cause problems to the less experienced or alert serviceman. The best way to check for this is to reduce the Tx signal level to a minimum and run each servo separately 90  Silicon Chip ters these days are fitted with servo reversing switches and these are a good thing too. There are, however, many old transmitters still in use without this feature and the re­quest still comes at regular intervals. The change is tricky at times and some care is called for. Firstly, write down or draw the location of the original wiring before you disturb it. Next, before you touch the pot wiring, measure the angle/resistance between the wiper and one end of the pot element – this becomes your reference end or colour. Now reverse the two commutator leads on the motor and the two wires at the ends of the Another common request is for the angle of rotation of the servo to be changed. Some applications call for very small or very large angles of rotation. The request for 180° of travel for flaps, undercarriage, etc is still a common one. Again, modern transmitters can sometimes accommodate this or servos can be purchased with 180° of rotation as standard. If, however, you live on a desert island and need to doctor an existing stan­dard servo, the procedure is as follows. Most servo ICs have external components to set such parame­ters as travel length, minimum impulse and pulse stretching, so check the specifications for this information. Small varia­tions of rotation angle can be achieved by changing the value of the one-shot timing resistor. Large variations are best done by placing a resistor in series with the pot element. Done very carefully, rotation angles of up to 250° can be achieved. Do not forget to remove the output gear over-travel stops in SC the gearbox.