Re: [CAD_CAM_EDM_DRO] Re: high-performance EDM power supply - software

Hi All:
Are you sure that you need to be able to reverse the toolpath on a wire
EDM at the level of the G code?
Seems to me, that if you were data logging the pulse stream just enough so
that you could reverse the sense of the pulses for a few pulses, you could
back up enough to clear any arcs, then reverse the pulse stream again to
where you were and resume reading the code.
That way the controller wouldn't need the capability to calculate the
reverse path from the G code...it would just be temporarily interrupted
while this little independent auxiliary program was running.
I'd expect to trigger the interrupt sequence by a low voltage condition.
I know nothing of the motion control requrements for this, and I'm dangerous
around sparks and programming, but it seems conceptually sound to me.

Another consideration with reference to sinker EDM:
The capability to do movements in x and y is very useful...so is the ability
to do 3 axis full interpolation.
Again, the difference between this problem and that posed by 3 axis milling
is in the need to interrupt and reverse the pulse stream for only a few
pulses in response to an external triggering condition, and then resume
where the motion left off.

In both instances, is it easier to recalculate the g code twice (once to
back up along the reverse of the original vector, then again to go back to
the interrupt point and beyond), or is it easier to do a local pulse stream
override quickly and frequently??
Don't forget that the G code language specifies a trajectory from where you
were to where you're going in discrete and large chunks, so it's more than
just reversing the sign of a pair of axes and splicing in the existing
coordinates to go backwards.
Seems to me, you'd need to establish a new "where you are" and a "where
you're going", statement, along with a feedrate (or rapid rate for linear
moves only) together with a new statement as to how you want the controller
to get there. (ie a G02 will need to become a G03 F max ...a G01 could
become a G00).

So...should we recalculate the code...or should we locally override and
reverse the pulse stream???
I await all responses with interest.
Cheers

Marcus

----- Original Message -----
From: "turbulatordude" <davemucha@...>
To: <CAD_CAM_EDM_DRO@yahoogroups.com>
Sent: Friday, July 18, 2003 5:46 AM
Subject: [CAD_CAM_EDM_DRO] Re: high-performance EDM power supply - software


> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "eexgs" <eexgs@n...> wrote:
> > Well I'm in!
> >
> > I offer my limited electronics knowledge e.g. a modern electronics
> > degree :) as well as the knowledge of some friends who do power
> > electronics and spend a lot of time shoving big volts and amps at
> > massive motors and design the controllers to do so.
> >
> > As for software, once you register TurboCNC then you get the source
> > code which might be modified for wire edm, reversing the tool path
> > would involve just flipping the directions of z and y plus other
> > stuff. For a sink machine then you could start from scratch
> assuming
> > single axis.
> >
> > I say don't worry about the software and get the design underway.
> >
> > Graham
>
>
> I am not sure what software will do such a tool path reversal, but
> one good thing is that it only needs X and Y.
>
> I see that a look ahead design, software that reads the next 5 or so
> codes and blends them from a smoother cut around courners
> (eleminatess stops and starts) would be part of the software, and
> although it would need to be able to reverse entirely out of a
> toolpath, it would only need to react quickly into the last few spots
> it was in, probably less than 5 of the last moves.
>
> Another option, and this one may be much harder, it so has the
> software know where an open area might be and be able to move into
> that.
>
> Dave