|
[DPRG] Motor Filter Capacitors
Subject: [DPRG] Motor Filter Capacitors
From: Randy M. Dumse
rmd at newmicros.com
Date: Wed Mar 21 15:18:54 CDT 2007
> In the robot I am building for SMU, the servo controller
> crashes whenever the motor has an instantaneous load (hard
> start, hits a bump, etc). I don't have any sort of filter
> capacitor across the motor leads. I have several to choose
> from, how can I determine what the proper value is?
Eric the answer depends on many things and it is not easy to
understand exactly what you're asking. If you mean you are using
RC Servos, and you need to put capacitance outside the case,
then, you need to use as big a electrolytic as you can get.
There's a formula, and you can figure the current use of the RC
Servo, which will be amps, and the RC time constant of the cap
going down, but realistically, you won't be able to keep the
voltage very high with tens of thousand of uF. Also, it helps to
use several caps. Different caps will have different effective
response ranges. You very often see a .1 monolythic in parallel
with 100uF (or bigger) electrolytic. Tantalums are even better
than electrolytics, but much more expensive as you get to larger
sizes.
Now if you are instead talking about putting caps across the
motor, you don't want much. Since the motor is being PWM'd at
some fixed frequency, the larger the capacitance you put on the
motor, the more of the PWM bleeds straight through the cap and
makes waste heat. The caps right across the motor are basically
for noise supression, and may have an effect on brush life, etc.
A .1 is about as big as you should dare there. Smaller, .01, is
more likely appropriate. Again, it depends highly on your PWM
rate. You can look up how much AC current will pass through the
cap at a given frequency, and use that to estimate how much
heating of the cap you can tolerate, vs. the noise suppression
you're after by essentially "grounding" out the higher frequency
components through the cap.
One more thing, you might be asking about the cap that goes on
the H-bridge. These supply the kick when the motor "starts" as
well, but the caps aren't across the motor, but across the power
input to the H-bridge transistors. Again, there are formulas.
Key thing for these caps are moderate size, high voltage
(because of back EMF transients), and very very low ESR's so
they don't toast off from dealing with the transients.
So without knowing really the details of your question, I hope
that at least helps you understand which of the problems you are
up against.
Randy
www.newmicros.com
More information about the DPRG mailing list
|
