Re: chopper drive - resonance problems - need ideas

Joel Jacobs wrote:

> From: "Joel Jacobs" <jj@...>
>
> Could use some ideas here. I'm having some resonance problems with my
> chopper drive. Maybe someone else has been down this road. First a
> topology review...
>
> Looking at 1/2 the circuit - a centertap and two windings L1 and L2, it's
> kinda like a unipolar exited bipolar drive - the chopper is applied to the
> center tap with a high side switch and when it's on the current ramps up in
> only the one winding whose low side switch is on but when it's off (90% of
> the time) the current freewheels back through the other winding so It
> actually flows in both windings during that time. The two low side switches
> share a current sense resistor as they are never both on at the same time.
>
> The high side switch is turned on every 62us (16khz) and remains on until
> current limit is reached. I'm running a 40volt supply and when it steps,
> the current ramps up to 4 amps in ~600us and then begins chopping. The
> steps are very 'forceful - when it's stepping slowly step - step - step -
> step, the motor goes TwANG - TwANG - TwANG - TwANG.
>
> There are certain speeds where the resonance of the armature wreaks havoc so
> severely that it looses steps with no load on the motor. Particularly bad
> at about 4000 spm.

This is not out of the ordinary, that's 67 steps/sec.

> The motor is not mounted so the case and the armature resonate - if I hold
> the case down tightly to the table it improves.

This is an abnormal situation. The motor housing should always be
tightly constrained, as it would be in a real application.

> If I apply slight drag by holding the pulley it stops resonating and runs
> good.

Yes. When the natural frequency of the rotor's rotational inertia and the
springiness of the magnetic fields that drive it are near the frequency that
the motor is being driven, resonant motion will build until steps are lost.
You have two choices. One, don't go near these frequencies, or two,
damp out those resonances. You can do it mechanically, and there are
hydrodynamic dampers that have a hollow, disc-shaped chamber, with
a heavy (like steel) disc floating inside, with a viscous fluid to couple the
two. Or, you can do it electrically, with RC networks across the motor
windngs, to absorb the resonant energy through the back-EMF generated
by the motor. I would think you could start with 1 uF 100 Volt caps and
a 10 - 22 Ohm resistor.

I haven't done this, but I have heard of it working quite well. When the
motor is stiffly connected to a load, like a leadscrew, it should perform
much better.

Jon