Incredible stepper motor phenomena

I'm finding that a stepper motor drive on a standard nema 17 motor gives a fixed amount of holding torque (say, 5L when still.
If you put the motor in motion, say, 5 steps a second, or even 500 steps a second, all the sudden it takes 5LB of force to stall it if you push WITH the motor, but only about 0.5LB to stall it if you push against the motor.

In other words the motor is stalled much easier in the direction it needs the most force. I've seen this problem on every 3D printer. Is there any way to reverse is so that the maximum torque can be provided against the direction of motion? What is this kind of toruqe difference called (detent torque vs. direction of motion)?

It is similar with microstepping

Edited 1 time(s). Last edit at 10/18/2013 11:54AM by Simba.

what motors are you using?

"a standard nema 17"

also testing the ones on affinia, does the same thing.
It should point of that detect torque rating is reasonable against the direction of motion, with the direction of motion it feels 5-10 times above the rating....

At low speeds I would expect close to the holding torque in each direction.

With a spring balance pulling my X carriage I get about 3kg if I retard it and 4kg if pull in the same direction when moving very slowly but about 3.5kg in each direction when stationary. The speed I used was F60 which is 5 full steps per second but using x 16 microstepping.

Holding torque with a 0.43Nm motor rated at 1.7A, driven with 1.2A peak is 0.43(1.2/1.7) / sqrt(2) = 0.22Nm.

One rev of the pulley is 40mm, so the effective radius is 40/2pi = 6.37mm, so the holding force should be 0.22 * (1000 / 6.37) / 9.8 = 3.52Kg, which is a good match.

So I get some effect but nowhere near the 10:1 you report.

Detent torque is the holding force when no power is applied. Not sure if it is relevant when powered but maybe it works with you in one direction and against in the other.

---

[www.hydraraptor.blogspot.com]

@Simba, I don't understand how it can stall in both directions. Perhaps you mean that in one direction you can make the motor overspeed (and miss steps) and that takes more force than if you try to slow the motor? By overspeed I mean the sort of thing that would happen if you were using the motor to lower a weight and the weight was too heavy. And by slowing I mean trying to lift a weight that is too heavy.

This is different from the measurements that @nophead made which, I think, are equivalent to trying to lift a weight with the motor turning clockwise and anticlockwise.

I have not experienced the phenomenon but I may experiment over the next few days. What stepper drivers are you using and what software?

Then again if the motor has enough torque in the "normal" direction does it matter if it has 10x in the other direction?

...R

Robin2 Wrote:
-------------------------------------------------------
> @Simba, I don't understand how it can stall in
> both directions. Perhaps you mean that in one
> direction you can make the motor overspeed (and
> miss steps) and that takes more force than if you
> try to slow the motor? By overspeed I mean the
> sort of thing that would happen if you were using
> the motor to lower a weight and the weight was too
> heavy. And by slowing I mean trying to lift a
> weight that is too heavy.
>
> This is different from the measurements that
> @nophead made which, I think, are equivalent to
> trying to lift a weight with the motor turning
> clockwise and anticlockwise.
>
> I have not experienced the phenomenon but I may
> experiment over the next few days. What stepper
> drivers are you using and what software?
>
> Then again if the motor has enough torque in the
> "normal" direction does it matter if it has 10x in
> the other direction?
>
> ...R

Yes you said it right.
When you try to overstep the motor, it takes 5-10X the force. I'm using Ti DRV with fullstep. Microstepping made no difference. GOING slow didn't make a difference. It's allowing this to happen the section there is motion in a particular direction. Again, held still, the stall force IS equal in both directions.

I should probably restate what I said earlier as the stall torque is as rated, but while moving the stall torque against the direction of motion is far, far below the rating. It skips too easily

It is the effect I measured. I.e. making the motor skips steps by holding it back or pulling it forward to over speed it. I didn't change the direction the motor was moving.

---

[www.hydraraptor.blogspot.com]

Try measuring it the way I did. When you turn the pulley by hand it is easy to get a false impression because the torque is proportional the the sine of the displacement and it skips when the displacement is two full steps and the torque is then zero and switching sign.

---

[www.hydraraptor.blogspot.com]

Based on what @nophead said, and if you are using well proven driver software without any DIY amendments I would suspect that the motor is wired up incorrectly so that there is some problem with the phasing of the magnetic forces. And if it is wired up correctly I would be looking for a software problem for the same reason.

I suppose another possibility is a failure in part of a driver board so that it is not giving full power on every pulse on every coil.

...R