Re: Why is the push to such high step rates (Grex and Arc100)?

--- In CAD_CAM_EDM_DRO@yahoogroups.com, Jon Elson <elson@...> wrote:

>
> turbulatordude wrote:
>
> >--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Mariss Freimanis"
> ><mariss92705@> wrote:
> >
> >
> >>256 microstep drive at 3,000 RPM equals a 2.56 MHz step pulse
> >>frequency.
> >>
> >>
> >>
> >
> >256 microsteps ?
> >
> >I had thought that at 10x the microsteps were not really dead-on balls
> >accurate (an indusry term) so just how useful is a 256 microstep when
> >it comes to motion control of our mills and routers ?
> >
> >
> >
> It isn't. I believe Mariss is absolutely correct that any finer
> division of the motor's
> inherent steps below about 10 is pointless, unless the motor has an

encoder.

>
> There definitely are closed-loop microstepping systems, where they

need fine

> microstep resolution to keep the system from dithering under mechanical
> load.
>
> Without an encoder, there is no way to control motor position even to a
> single
> full step. Friction and torque applied to the motor will cause the
> position to
> be offset from the nominal step position. And yet, they keep making

these

> systems.
>
> Jon

It is that last sentance. If sounds like they are not all that good.

When I have done the math, it often comes up that it really doen't matter.

Considder that a common screw is 5 TPI for a small machine ball-screw
or even ACME. Some are finer, but 5 is easy to work with.

a direct drive of a 200 SPR stepper on a 5 tpi screw yeilds 1,000
steps per inch. a 10x microstepper would have a theoretical 1 tenth
resolution, but the 1 thou resolution is for full step.

On a rough cut, the machine might vibrate and everything will flex
somewhat. the finish will be less than perfect as will the accuracy.
A finish cut is designed to be less demanding and *should* be MUCH
closer to the actual machine theoretical resolutions. Be that as it
may, we can assume that the motors will 'fail' to the nearest full
step. for small shops and home projects, that is more than enough
resolution.

So, the serious modelers find they want a tad more precsion and they
would often figure the motors are not strong enough, or that the axis
is not tight enough, or that they do not have enough step resolution.

So, the direct drive gets a 2:1 set of belts. that doubles the SPI
and solves a lot of problems. it also opens the door to a 3:1 if need
be. But at 3:1 the driven pulley is huge and that is about the limit.

So, the machine if then fitted with either larger motors or re-vamped
to better screws, or re-vamped to servo's with encoders.

From watching the list over the years, this seems to play out over and
over. The user keeps tweeking until his needs are met.

I can see that any motor with an encoder would offer much higher
positional accuracy, but since there are so many machines without it,
I don't think it matters for many of them.

I have a professional PCB machine that has steppers. since the
product is circuit boards, or rather engraved copper traces on them,
it is simple to say that a 1 thou error would be of no consequence.
And, further that the loading on the machine is so light that I
probably get darn close to the theoretical positional accuracy.

I also have a Roland CAMM-3 machine. table top mini-engraver/modeler.
not a metal cutting machine. it can do aluminum with a 1/8 or 3/6
end mill, but not deep cuts. for plastic or synthetics, or wax, I can
imagine it is wonderfully accurate, but on metal, the feeds and speeds
have to be just right, light finish cut and climb milling, to get
anything even near a decent finish.

My point is that there is a huge segment that would work wonderfully
with steppers.

Dave

>