Re: [ATM] Servo Motors: Gear Reduction and Encoder Resolution

["Ashley Roll" <ash@digitalnemesis.com>]

Hi Martin

This is my first ATM post - I've been lurking for a few months. 

I am not certain if this will answer your question - I hope it sheds some light on the subject of driving the servos though.

You have two different modes of operation - positioning and tracking. With positioning you want to go to an absolute position, with tracking you want to apply correction at a computed speed. 

It is possible to think of a servo motor being controlled similar to a stepper in that you "step" it to the next encoder location each time period, but the control circuit for this would be complex I think and you loose the inherent smooth motion afforded by a servo motor. I suspect that it would also "jitter" around the position as the only feedback is the encoder so the motor driver would only be able to constantly hunt around the "edge" of the signal transition.

Perhaps the more common way a servo motor is used is "constant speed". 

Sorry I don't have the time to do the numbers at the moment, but basically you calculate the desired rotor speed in rpm. This translates to pulses per second (frequency) which can easily be measured accurately with a frequency reference and the motor drive PWM signal modified to keep the speed as programmed. You then have a trade off between the speed of positioning verses the accuracy of the tracking control. If you can run the motor faster for tracking you get higher frequency and faster response to correct the speed, but at the penalty of slower maximum speed after the gearing so slower positioning.

You would want to probably have two "loops" of control. The inner loop monitors and controls the speed of the motor; the outer loop monitors the encoder to count absolute position. The outer loop can then drive the inner loop to slew position fast by ramping up and down the target speed based on the current and desired position error and then go into track mode after computing (and adjusting periodically) the required correction speeds.

The servo motor solution will require more complex electronics to drive it. But there are commercial (industrial) motor drives that will do all this for you, but you pay for it :)

Cheers,
Ash.

> I want to drive my new alt/az telescope with servo motors.
> My motors have 1000 encoder cycles per turn and
> a 3.7:1 gear so each turn at the gear output gives
> 3700 encoder cycles or 14800 signal changes.
> This is more resolution than a stepper with 200 steps per
> turn and 40 microsteps per fullstep.
> But: Is it possible to use a servo instead of a stepper
> and control it precisely enough ?
> Or should I use more gear reduction ?
> The planned gear is 1400:1 so one encoder cycle
> is 0.25 arcsec, one signal change is 0.06 arcsec.



_______________________________________________
ATM mailing list http://www.atmlist.net/