Re: Closed loop steppers

Ian

There are trade-offs between steppers and servos.

With steppers, one problem is lost steps. This is what Arne and a few are
mulling over. A second problem is repetitive starting and stopping rather
than simple smooth velocity control. This leads to problems with resonance
and such.

With servos, the biggest problem that I've encountered is with tuning. You
are right, a servo system can overshoot, following error can be different
between different motors, and a motor/amp can palsy when stalled. All of
these problems are curable in a well designed system with proper tuning.

Perhaps a part of the uncertainty about servos is caused by the fact that
someone long ago decided to call the difference between where the servo was
and where it was commanded to be "servo following error" or "servo lag."
It doesn't bother me a whole lot, when I get to the edge of town and press
resume on my car's cruise, that it takes it a bit to get back to speed.
But somehow we get the notion stuck in our heads that this will be a
problem with an axis drive. Following error bothers me more when a deer
(around here they are used by the govt for revenue enhancement so more is
better) jumps out in the road and my car does not slow down as fast as I'd
like. This also seems to bother my insurance adjuster. The equivalent
with a servo drive bothers you and many others.

At modest cutting speeds, I try to tune following error down to a couple
tenths of thousandths. At rapid speeds of 200 + ipm, following error may
be as much as a couple tenth. Most servo systems will use following error
for stopping distance although some have internal ramps for acceleration
and braking and reduce following error to near zero while at constant
speed. Any decent servo system, even the old ones (50's and 60's) that I
have to tune up quite often, exhibit very predictable motion.

With a well tuned servo system, I can cut a 3-15 degree taper with a
surface finish and appearance that we'd never get with a stepper system
that is as capable.

And as for stopping, the old GE drive and Reliance motors that run the axis
on my lathes, still stop within 0.0001 of where they are commanded.
Following error is a percent of speed. So at zero speed, it is pretty
small. :-) And at zero speed, the current will ramp to full six amp,
without a loss of position, if I apply a load to the shaft.

That same adjustable power that prevents motion can be used to stop motion.
In that case it's dynamic braking, the servo uses the
electricity/magnetism generated by the systems own inertia to stop it. My
3kw spindle motor will stop the spindle and a 1-3 inch part from 3600 rpm
in less than a second.

I've been playing with steppers here a bit and have almost as much
trepidation about them as you express about servos.

Ray

From: "Ian Wright" <Ian@...>

>Subject: Re: Close loop steppers
>

<snip>

>I must say I'm not yet totally convinced about the benefits of servo motors.
>I can't see how they can reliably stop in an exact position - I can
>understand how feedback from the encoders works and even how the software
>can ramp down the speed towards the end of a move but how do they get rid of
>flywheel effect running the motor on beyond its desired stop position and
>what happens if the cutter hits a hard bit near the end of the cut, does the
>motor stall early?