Re: Going home / more stepper speed!

Jon,

Hi, What I did to extract a usable signal was to lobotomize a drive;
meaning I took out everything except the bridge drivers, which were
slaved to the drive under test.

I had it drive an identical motor except I superglued its rotor. This
way one motor was running while the other was pernamently stalled and
thus only showed the inductive current. Both were being driven under
identical conditions.

A differential amplifier subtracted the current of one from the other
and revealed the underlying back EMF. At low speeds things needed to
be trimmed out very carefully since the signal I was looking for was
at least 50 db down at 100 full steps per second.

I did design a simple feedback circuit that effectively damped low-
speed resonance and gave an output proportional to load. This was
effective down to 1/2 rev per second.

What has occured to me since the subect has come up again now is to
replace the crude (though effective) stalled motor with a current
source driven capacitor to model the stalled motor and act in its
stead as the other differential input.

I may as time allows to persue this line of reasoning again. A change
in perspective always modifies ones understanding of things.

Mariss

--- In CAD_CAM_EDM_DRO@y..., Jon Elson <jmelson@a...> wrote:
>
>
> mariss92705@y... wrote:
>
> > Alan,
> >
> > I respectfully disagree with Jon. You cannot see ringing with
either
> > a voltage probe or a current probe when the motor windings are
driven
> > with a full bridge at low speed. This is because any back-EMF
signal
> > is completely washed out by the motor's inductive current and the
low
> > impedence voltage source that constitutes the bridge.
>
> Yup, I realized you are right!
>
> > I have tried very hard to come up with differential canceling
circuit
> > schemes to accomplish this because a real breakthrough would be
the
> > result.
> >
> > If it were possible to reliably retrieve this signal then it
would be
> > very simple to use it to use it to actively damp the ringing and
also
> > have direct missed step feedback.
>
> I think this can be done with some motors with split windings that
are
> on different poles of the stator. (Obviously, windings that are
distributed
> over the same stator poles would couple the same flux.)
>
> > The signal is there alright, at higher speeds (>100 full steps per
> > sec), it is easy to find. Unfortunately as speed approaches zero,
so
> > does the back EMF.
>
> Yes, right there, too.
>
> Jon