Re: Basic stepper control question

--- In CAD_CAM_EDM_DRO@e..., Chris Stratton <stratton@m...> wrote:

> Tom,
>
> Sounds like you are going to have to find the stable positions of

the

> stepper and restrict the control to moving to one of them before
> powering down. I would have thought there are 200/revolution, but
> maybe there are only the fifty you can feel.

I think you're right about the 200 steps, but the very last thing I
did before leaving work Friday was to check that out. Other than the
counting, I turned a motor by hand while it was still on the device.
This thing is essentially a "tuner", and I monitored the output with
a frequency counter. Under CPU control, I saw 6KHz steps. By hand,
the best I could get was something like 25KHz. It was pretty
inconsistant, but definately averaging around 25KHz. If I tried
really hard, I could get it to settle elsewhere, but it was not a
sharply defined position.

Given that we're doing full steps, I figure I was seeing the step
size increase by a factor of 4, which implied we were only stable at
one of the 4 possible positions.

FYI: I just counted a motor here at home, and it has 200 very
defined stable positions per rev. The plot thickens...

Something else to chew on: The guy who's going to argue with me
about this also took it upon himself to disassemble/reassemble the
motors in order to get the coil/rotor combo he wanted, despite my
warnings about what manufacturers think about that sort of thing.
Could that explain what I'm seeing?

(Note to the group: I don't really want to debate the risks of
stepper disassembly. At home, I'll screw around like everyone else.
At work, I follow all the manufacturer's rules. Better safe than
sorry, especially when it's the boss's money).

> One certain thing: none of the 2-phase-on positions are stable, as
> they position the rotor halfway in between where either phase alone
> would want it.

This much I gathered from the Compumotor doc that Ballendo posted a
link to. It also makes many references to the fact that you can lose
more steps in the power-up cycle, as opposed to power-down.

> I'd suggest you put a long pointer on a motor shaft, get a bench
> supply set for appropriate current, and run the step sequences by

hand -

> power the winding, then disconnect.

I just may try that. Depends how much time I can spare and how much convincing I need to do.

-Tom Kulaga