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

One place where high encoder rates would be useful is if an
interferometer was being used, with a typical step of 0.3 µm (1/2
wavelength) a 1 MHz rate would be 300 mm/s which is pretty fast, but
not too extreme for robotics.

I might use an interferometer here and there because they are
sometimes much easier and cheaper to set up than encoders of high
precision.

For example an interferometer can be set up to obtain up to 6 axes of
motion from an object from any distance, no contact.

There are used to measure the speed of deformation of exploding
objects, for one example. This gives tremendously high fringe rates.

They do have disadvantages, which is why you probably aren't using
them.

regards-
Elliot Burke
--- In CAD_CAM_EDM_DRO@yahoogroups.com,
"lcdpublishing" <lcdpublishing@...> wrote:

>
> Thanks for piping in Patrick.
>
> I can understand the desire for a smooth pulse stream and certainly
> can appreciate the effects of a bad pulse stream on steppers, but
> does that have any effect with servo driven systems?
>
> Also, do you know of any applications that would use such a high
> frequency for the pulse stream? I am obsessing about that
> specification - sorry.
>
> I just think about the industrial CNC machines I am familiar with
> (which also transcends into robotics). 4000 IPM at .0001"
> resolution is the fastest I have seen and that was on a FANUC ROBO
> Drill machine (basically a 3 axis machining center for drilling).
> At 4000 IPM and .0001" resolution, works out to about 660,000 SPS.
> It's just mind boggling to watch something move that fast.
>
> Chris
>
>
>
>
> >
> > It is not only a matter of how high a pulse rate a device can
> produce.
> > It is also how much jitter the pulse train has. While a PC may be
> able
> > to reach 45KHZ. The pusle train can have severe jitter if you

try

> to
> > go just a bit higher and that causes difficulty in building

systems

> > that push the PC's limits and run in production environments even
> at
> > bellow 100KHZ.
> >
> > Having a stable pulse train allows running the same
> motor/driver/load
> > combo at higher pulse frequencies without missed steps or stalls
> in
> > long production runs. That also applies to hobby use if you make a
> > part that takes days of machine use like it often happens when
> using
> > software such as DeskProto for example.
> >
> > That is where products such as te ARCNC100 come in. And once you
> have
> > solved the jitter issue, it doesn't cost any more to make it go

up

> to
> > 4 MHZ or even 20 MHZ for that matter. Any solution that solves the
> > jitter issue involves hardware and for hardware, 4 MHZ or 20 MHZ

is

> > pretty much the same ball park.
> >
> > Patrick Robin
> >
>