DPRG
DPRG List  



[DPRG] PWM vs. voltage motor control

Subject: [DPRG] PWM vs. voltage motor control
From: Mike McCarty jmccarty at ssd.usa.alcatel.com
Date: Thu Jun 14 15:10:47 CDT 2001

On Thu, 14 Jun 2001, Joe Whitmore wrote:

> Could it be how the PWM is being implemented?
> Servos - 2us pulse and 1ms repeat? Leaving a lot of off time if driving
> a motor instead. The amount of off time all at once is what I see to be
> the real killer here.
> 
> What I'm asking/suggesting may be already be the case but it never hurts
> to ask again.
> 
> What about making the cycle time smaller. For easy math in my head I'm
> going to use big times.
> Say you wanted 50%, so you'd be on for .5 seconds and off for .5
> seconds. This would be a lot of off time. and going lower even more off
> time.
> Could you then change the cycle rate to be on/off for .25 seconds or
> even .125 seconds. Over the 1 second period the motor would be on for
> the .5 seconds, and off for the .5 seconds, just spread out further. 
> 
> Could this sort of logic go into your PWM algorythm. So that whenever
> the duty cycle dropped below 50% of the total cycle, it would chop up
> the cycle. So that in effect in one cycle it would get 2 or 4 pulses
> spread out, rather than just one in the beginning?
> 
> Just a thought.
> Joe

By my reasoning (presented in an earlier message), this makes things
worse. I suspect that what one needs is a minimum pulse duration. If,
in order to get the duty cycle down low enough, one violates the
minimum pulse duration, then one should fix the pulse duration and
start reducing the frequency of the pulses. This should provide the
torque needed to continue motion at low speeds, while still reducing
the duty cycle.

Eventually, one would start to violate a minimum pulse repetition rate,
and just have to stop. I suspect one pulse every ten seconds is not
acceptable.

It seems to me that one could build a resonant tank from the inductance
of the motor coils and a parallel capactor, and make it resonant at the
pulse repetition frequency. The Q would be low unless the motor is
under heavy load due to the equivalent resistance reflected back into
the electronics domain. One might need to add a parallel Q spoiler for
when the motor is under heavy load and the reflected resistance is very
low. I dunno what that would require, but I suspect it would not waste
lots of power.

Mike
-- 
char *p="char *p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
I can explain it for you, but I can't understand it for you.
I don't speak for Alcatel      <- They make me say that.



More information about the DPRG mailing list