DPRG List  

[DPRG] Servo resolution

Subject: [DPRG] Servo resolution
From: Ed Koffeman edk at kinetric.com
Date: Sat Feb 26 14:12:28 CST 2000

> That is, if the servo is driving a shaft, what is the smallest angular
> change that can be commanded by a digital signal?

The signal is normally only digital in that is is high or low. It is
actually interpreted entirely in analog circuitry in almost all servos,

So, if you are talking about incremental positioning without worrying about
static friction, loading, backlash, loop gain  and other nasty aspects, then
the answer is "infinitesimal" or whatever the carbon grain size and
potentiometer wiper dictate as the smallest change in resistance available.
There are no actual "steps" in the resolution.  Some servos will overshoot
under no load and be less predictable than servos with a unidirectional load
(which takes out all the backlash and output play).

However, if you are talking about reliably returning to a desired position
>from any other position, and especially if the loading could be different,
then there is probably about a degree or so for a cheap servo, mostly from
play in the output bushings and final gears.

I think you could probably see the servo move in consecutive same-direction
steps less than a tenth of a degree with a light unidirectional load.  But
if you changed anything like the direction of the load or the direction of
movement I think it would be off by around a degree.

The Multiplex comapany makes digital servos that are very stiff, with a
proper PID loop and these would have much better repeatability.  The JR
company also has stiff servos but I don't know if they are digital inside.
Both of these still use the standard 1 to 2 ms pulse input.

Ed Koffeman

More information about the DPRG mailing list

Copyright © 1984 - 2006 Dallas Personal Robotics Group. All rights reserved.
Website Design by NCC

For the latest robot news visit robots.net