Some poorly documented facts about driving servo motors

[selbstdual]

I found some interesting things regarding driving servo motors being not or poorly documented I want to share here:

- Put a LOW Port.Number before Pulsout to make sure that it does not invert the signal the wrong way which makes the motor not work at all
- Use INTCON = 0
- Use CMCON = 7
- Don't expect the motor to perceive 1ms as 0° and 2ms as 180°. Test it first. Most servos accept values less and greater than that.
- Keep on sending signals to keep the motor's maximum torque.

[BrianT]

ALL servos WILL work correctly with pulse widths between 1 mSec and 2 mSecs. This may give you +/- 30, +/- 45, +/- 60 degrees or in some rare cases +/- 90 degrees or odd values in between. There is NO standard angle of rotation, not even within the same brand of servo.

SOME servos may allow you go to lower or higher pulse widths but at the risk of hitting internal mechanical stops. Don't expect to be able to dial in +/-120% for a digital servo travel.

DIGITAL servos WILL SELF DESTRUCT if they hit the stops. With the JR digitals they get so hot so quickly that the coreless motor armature melts the plastic armature former. They may cool down and work afterwards but will never be the same. I blew up 8 digital servos at vast expense with code that was sending < 1 mSec for about 5 seconds before the autopilot properly initialised.

Price is no guide to life. On test, we had expensive German servos (Vxxx) fail mechanically days before the cheapest Taiwanese GWS servos of comparable rating.

Futaba metal gear micro servos seem almost bullet proof and can be attached directly to the vibrating engine for throttle control.

Metal gear servos wear and develop increasing backlash with age. Nylon gear servos have constant backlash throughout their life provided they are not abused.

From my experience working with a small UAV company that had a ten watt 450 MHz transceiver in the airframe, metal gear servos were much more susceptable to RFI as the metal gearing seems to act as an antenna.

The pulse repetition rate determines the speed and holding torque. Above about 40 frames per second the servo will deliver rated torque and speed. As the frame rate comes down the holding torque diminishes but an unloaded servo will hold position down to below 5 frames per second. This can save a lot of power on servos that do not have to constantly deliver torque.

FWIW
BrianT