Motor Madness -- solved

rjo_

Hi guys,

I have a question about resonant damping and stepper motors and a question about high end torque...and this is probably going to be clear as mud[noparse]:)[/noparse]

I have found that as I vary the dwell time between steps, there are some dwell periods that will cause the motor to misbehave... at lower speeds this looks like the motor is epileptic and at higher RPMs it looks like the motor goes into shock (stalls). I have found that if I quickly step over these nodes of misbehavior... the motor is extremely well behaved. The approach that seems to work best is that when I want to accelerate to a given RPM... I simply decrease the dwell time by a fixed amount each time through the loop that is driving the steps. This works very nicely. For example, if I just begin a session with a dwell time that should produce a high RPM... the high end will be around 120 RPMs... but if I use a ramp on the dwell times I can get the same unit up to around 600 RPM's, reliably and occasionally as much as twice that.

I have thought of finding a way to register the behavior of the motor to specific dwell times as I am ramping up and then simply bypass those dwell times when I need less acceleration.

Is this equivalent to resonant damping?

I've read a little about poor high end torque in stepper motors and have looked at some of the strategies that
are used to increase high end torque... and they are all either above my head or I don't understand them at all[noparse]:)[/noparse]

Why can't we gently coax a stepper up to max speed and then ramp the voltage?... I haven't done it because I'm in one of my rare non-destructive periods[noparse]:)[/noparse]

Thanks,

Rich

Post Edited (rjo_) : 4/3/2009 1:52:49 AM GMT

rjo_

By the way... I originally thought there was something wrong with my implementation or motor... but in looking on the web I found that this is normal stepper motor behavior.

I also found that by attaching a medal bar to the shaft with the flat side perpendicular to the shaft of the motor and then hanging various weights off of it... I can convert a stepper motor into a musical instrument[noparse]:)[/noparse]

You engineers have been having all the fun...


Rich

Post Edited (rjo_) : 4/2/2009 11:10:06 PM GMT

Peter Jakacki

Stepper motors step, that's what they do, they are never going to behave like a DC motor where we can just vary the voltage and have it respond. Microstepping though does give us some damping because it's a bit like a heavy-handed push-pull action that clamps the armature into an in-between position. I haven't played with stepper motors for quite some time but when I did they were big ones with unbalanced loads that needed very fine positioning and rapid response, this was achieved with microstepping and patience. Now with my newer Prop based controllers and drivers I will be "playing" (can never get an engineer to do anything serious) with steppers again, both big ones and little ones so I will try to provide some feedback to the forum on this.

BTW, I am not always familiar with a lot of THE terminology and I don't always agree with it but I do my own practical research and experiment and observe. It's always way more fun and rather than bandying words about you can cut through the confusion because you then have first hand knowledge and understanding, then it's all just child's play

Just read your latest post, see...

*Peter*

rjo_

Peter...

Thanks for responding so quickly.

I warned you I was probably going to be clear as mud... and I was!!!!

I have a basic handle on steppers... what I was really trying to dry-lab before I sat down to build was whether high end torque can be substantially improved by simply applying higher voltages after reaching the high RPMs.

With my little stepper, I can get it going and it is so smooth on the high end that you would swear that it is a regular dc motor.
But... it has absolutely no torque at high speed.... well... I guess that's not completely accurate, I am moving some metal around... but you know what I mean[noparse]:)[/noparse]

I understand your point about micro-stepping and I'll look at that next... it is standard and it is very useful.

Thanks,

Rich

Peter Jakacki

RIch,

I guess that's why I responded to this particular thread about steppers rather than others, I am patient, but not that patient...

Yes, I can get them to run very smoothly too (darn that resonance thing) but at the higher speeds there is a big problem with coil inductance limiting the power that can be delivered thus the speed and torque. The higher voltage will compensate to a certain extent but normally we would be running at a higher voltage anyway and relying upon the current sensing and PWM to regulate the power. Forgive me as I may be lacking a little in appreciation at present until I get back into the saddle with these motors but I feel that an adaptive system is required that senses the position and compensates accordingly. Once we have an encoder in position though we start thinking about DC servos so we have a choice, do we want to keep it simple, or do we need a DC servo? A perfect stepper and a perfect servo, wouldn't they be the same thing?

*Peter*

Philldapill

Rich, I've "played" with some steppers I have, as well. I know exactly what you're talking about when you call it "epileptic" behavior.

If you sit down and really analyze what happens during a "step", it's intersting. When you start pushing current through a winding, a magnetic field starts being produced. This field begins to pull the rotor into position. However, the rotor has a bit of momentum. This momentum causes the rotor to align with the field, but then overshoot a little, then it starts rotating the other way back to field. The rotor continues to oscillate back and forth until the momentum is finally dampened, and it reaches a steady state - aligned with the magnetic field. The period of oscillation seems like it would be related to the magnetic field strength(current) and the moment of inertia of the rotor. If you increase the field intensity, the period of oscillation should decrease. However, I think the time it takes for the rotor to align and the momentum to be dampened, stays fairly constant.

My point about all that, is this... If you are stepping at different frequencies, then when you initiate the next step, the rotor's momentum could either be clockwise or counter-clockwise. The position could be at the farthest CW extreme, or it could be at the CCW extreme. I believe to achieve the highest speed/torque, your stepping frequency would be such that when your next step comes around, your rotor is aligned with the current magnetic field, but have it's highest momentum. This way, the rotor will not ever be locked in any position, but the rotor momentum will keep it flying.

The epeliptic behavior you talked about is probably due the next step being initiated when the momentum is just barely zero, but the rotor is at one of it's extreme positions. Depending on a number of conditions, this may cause the rotor to sort of randomly move about - Clockwise, for a bit, then counterclockwise - both with virtually no torque.

Anyway, I've attached a simple 1 minute paintbrush picture to help illiustrate what I mean. I hope this helps your thinking process of what is going on in those 10-20uS that it takes for the motor to step.

JasonDorie

I overcame this on my CNC machine by using rubber caster wheels mounted to the rear stepper shafts.· Follow the directions here:

http://solsylva.com/cnc/dampers.shtml

I was floored at how well this works.· My machine used to max out at about 50 or so IPM (inches per minute), and now it'll hit 200.· Triple the speed for a couple of $4 rubber wheels.

The wheels absorb some of the vibration outright, and because they're 'slip clutched' they're always slightly out of sync with the stepper.· It makes a very surprising amount of difference.

Jason

Peter Jakacki

Phil,

This is what happens for sure but that is why the microstep when it is tugging both ways helps to dampen the resonance. But once your microstep is being controlled in an almost sinusoidal mode then the motor is gliding rather than stepping and the resonance and "noise!" disappears, that's when steppers are beautiful whereas full and half-step are ugly. Talking about all this is making me yearn to fire up some big steppers and put them through their paces. Ah, but work first because that's my next project.

*Peter*

rjo_

Peter,

Thank you, thank you. That's exactly the question I was trying to ask ++.

Phil

I was watching and listening at various speeds. Phil's description exactly matches my intuitive sense.

Jason,

You were getting so much good feedback from the guys that I didn't join in on the thread. But your really nice project is what about 99% of the guys would really like to do... but just never get around to for one reason or another.

The rig I'm using actually has such a weight... but I didn't know exactly why. I had a retired machinist who used to come in
and churn out parts for my little research group... this motor and mount is left over from a project that got dropped when my old guy just couldn't do it anymore.

Thanks everyone

Rich