Re: stepper-servo system

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Graham Stabler" <eexgs@...>
wrote:

>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "caudlet" <thom@> wrote:
>
> > Actually it's pretty easy. Your define a target voltage. You compare
> > that with the scaled gap voltage. If it's higher you set the
> > processor to flip the DIR pin the correct dir and then output a
> ramped STEP rate from another pin.
>
> But if it were a dc servo system there would be some gain factor to
> take the error voltage and determine the rate of that pulse train.
>
> >The gap voltage becomes the feedback. The coupling is loose rather
> >than like in a postional loop.
>
> The coupling is based on the gap voltage of course but this should
> respond quite quickly to electrode position. I don't care at all
> about the actual servo position just maintaining the gap. In terms of
> a dead band, the steady state should have the head moving forwards
> slowly with the feedrate adjusted to acheive this rather than a
> bang-bang approach which will always cause oscillation or a pulsing
> forward motion with a deadband.
>
> Graham
>

So if voltage of the gap = height (or depth or whatever you call it)
and the wire cuts the "servo" (reading the voltage would force the
motor to move down to correct. You are not talking about inches here
but fractions. Things like acceleration, move rate etc would be close
to meaningless. The target voltage would determine the speed at which
it cut and it would move in the correct direction to maintain the
balance. Maybe I still don't understand, but it appears to me if you
used a motor to adjust postition and the gap voltages to measure it,
it would just simply move down as the material was cut "looking" for
more material to balance the voltage. I don't think complex velocity
or encoder posititoning is of use here with the tiny amount of moves
and velocity you are talking about. It sounds like it doesn't matter
what the actual position really is in the material just that it is
cutting at the right rate. The feedrate would be determined by the
voltage change. The more the DC is filtered the slower the response.
Either it's cutting or not and either is finished or not. Some way to
sense extremes so it would fault out if the wire broke or the cut was
finished.

The unknown is the amount of voltage varience you need to detect a
change. In plasma we keep the tip volts within 1.5 volts (out of
140VDC) or 1%. It will hold it within about .025 gap and if the
material is not level or dips it would follow it to the center of the
earth if your Z had that much stroke trying always to maintain the gap.

The THC is integrated with MACH3 so the THC motor and moves are just
part of the Z axis and it listens to the THC circuit and just works to
maintain voltage equilibrium.

TIG works on somewhat the same principle with a constant current arc.
We are currently getting ready to ship a THC that has been
re-programmed to work with the lower voltages and tighter gaps needed
in TIG.