Re: microstep accuracy was what comprises as step

Hi again Jon, and Mariss,

I also got with the Servo's (auction) a Super Vexta microstepper
system. The driver is UDX5114, and the stepper is UPH596-A. 500 spr, I
think. The manual is *ALL* in Japanese, but I see 11.8kgcm and
700gcm^2, and 1.5kg in one table for this motor/driver. 1.5kg looks
like weight, and I'm guessing that 11.8kgcm will translate out to
holding torque, and 700 gcm^2 will be the running torque. It's a NEMA
#34, and I've only got one, otherwise I'd use them on my PCB router
project. Not much info on this on the 'net, either!

My question is, if it's 500 steps (more?), can I count on each and every
one of them? I'm told these are matched systems. 100VAC to the driver
(I can fix that w/ a bucking xfmr), and CW/CCW inputs. Logic can easily
fix that, or, as I'm working with TurboCNC and my own program, I think
I'll probably just modify the code to drive them! The driver came with
it, otherwise, I'd be tempted to play around with a 5-phase stepping
sequence!

Alan KM6VV


Jon Elson wrote:

>
> mariss92705 wrote:
>
> > Jon,
> >
> > I think you are confusing servoes with steppers here. A microstep
> > drive outputs sine and cosine currents to the motor. Between steps
> > these currents are steady (unchanging), so the motor position does
> > not dither.
>
> No, I was referring to steppers systems with encoder feedback. Unless
> the stepper positions are aligned well with the encoder transitions, I
> would expect some hunting in a closed loop stepper system, but
> not in an open loop stepper. This was just an aside at the end of
> my comment.
>
> > Accuracy depends on the ratio between the motor's holding torque and
> > the mechanical system's friction. As an example, if your motor has a
> > holding torque of 150 oz-in and the system friction is less than 15
> > oz-in, you can pretty well be guaranteed an accuracy of 1/10 of a
> > full step.
> >
> > There is nothing special about a half-step versus a microstep. Half-
> > stepping is the equivalent of 2 microsteps per step. Going to a finer
> > resolution of 10 microsteps per step simply means your resolution now
> > equals the motor's native accuracy.
> >
> > Anything less than that and you are not getting the full potential
> > from the motor. Anything much more than that buys you "empty"
> > resolution since accuracy pretty much 1 part out of 2,000 (1/10 step)
> > on an average step motor.
>
> Right. This is really what I was trying to get at, that a X1000 microstepping
> drive will not give you positioning of 200,000 steps/revolution. A lot
> of people apparently think they really can move the motor reliably and
> repeatably in micro-arcseconds, just because the computer can command
> such a microstep.
>
> >
> > In practical terms this accuracy can be depended on. As the load on
> > the motor relaxes, it moves into the correct (accurate) position.
> > When you are cutting metal, the inital load may be high (>15 oz-in).
> > As the metal is removed, the load lessens and the motor responds by
> > moving into its accurate location. All this takes milliseconds.
> >
> > You mention "may move 3 or 4 steps, erratically", "no guarantee of
> > any movement at all" and "cause really annoying vibration". This is a
> > phenomena I have never observed. Can you please elucidate? None of my
> > instrumentation has ever even hinted at this "effect" when using step
> > motors. Perhaps you have taken data that you would consider sharing
> > to shed light on this behavior? It would help everyone.
>
> What I meant, there, was 3 or 4 MICRO-steps, erratically, in a microstepping
> drive with, say x10 or x16 microstepping. Every single microstep may not
> cause actual movement of the motor, depending on load.
>
> The really annoying vibration was ONLY in reference to closed-loop
> stepper drive systems with insufficient deadband. There are lots of
> servo systems that jitter, too.
>
> Jon