Re: [CAD_CAM_EDM_DRO] Re: Software for wire EDM control?

Hi Ray,

You certainly COULD cut tapers and transitions, as you suggest w/ arcs.
The key is the Kn.n and Ln.n parms. This implies, to me anyway, that a
PAIR of "circles" must be solved for. Not in your typical hobby
controller program. And probably not (I can't speak from experience) in
many industrial controllers not designed for EDM. Yeah, I can see how
to expand my controller to handle a PAIR of circle calculations. But if
three popular CAD/CAM packages just generate line segments, then that's
probably the way to go. No special controller needed, other then the
arrangements for current sensing and "backing up".

When I looked at the Bobcad and the Vector CAD/CAM EDM demos, and
portions of their manuals, it was obvious to me that they intend to
generate "LOTS OF LITTLE ARCS" (cords), and thus use G01s to do the
moves. Other CAD/CAM programs (even these) might also allow arc moves
(G02/G03) in EDM calculations.

I found it quite interesting to "learn how they do it", and contemplate
how I might make additions to my controller program (or someone might
make changes to TurboCNC, for that matter).

If the two planes were slaved, then you'd have something like a gantry
(?) router. One plane would SIMPLY FOLLOW the other. While for axis of
LINEAR moves can all be coordinated, If you have a taper or transition
could require TWO arcs, if both top and bottom had radiuses in their
profiles. These arcs would be of different sizes, and thus require
SEPARATE computations.

BUT YES, both tapers and transitions can be handled quite simply if line
segments are used. Which seems to be the case, with the limited
sampling of CAD/CAM solutions I studied.

Alan KM6VV
(NOT an EDM expert)


Raymond Heckert wrote:

>
> I don't see why you couldn't cut tapers, AND transitions. If the U-V is
> 'slaved' to the X-Y, and, for example, you ask the top axes (X-Y) to cut a
> 1/8" radius while the bottom (U-V) axes are cutting a 1/4" radius, then
> you're cutting a taper. You'd have the Xn.n, Yn.n, In.n, Jn.n, as well as
> the Un.n, Vn.n, Kn.n Ln.n code in the same line. You should also be able to
> even do transitions, i.e. circle on the top, and square on the bottom.
> Your cutout part would reflect the rays generated by the wire. By slaving
> U-V to X-Y, you'd have the necessary co-ordinated movement. Besides the
> technicalities, am I missing something? You'd, of course, need to calculate
> the exact toolpath(s) required for the part suspended somewhere between the
> top and bottom planes.
>
> (Aside:) At a place I'm currently working, we order Punches & Dies from
> Wilson Tool Co, in Minnesota, and I'm sure they use EDM to make their dies.
> They make a product called a 'Slug-Hugger" (might be TM'd) that surely is
> generated via EDM. There are little vertical 'lands' uniformly spaced
> around the die opening, and they restrain the slug from sticking to the
> bottom of the punch & being pulled part-way back through the stock.
> DISASTER strikes when this happens, as either the servo's drop out, or
> worse yet, we have a tooling crash, when the die holder is pulled part-way
> out of it's housing.
>
> RayHex
>
> ----------
> > From: Jon Elson <elson@...>
> >
> >
> >
> > Raymond Heckert wrote:
> >
> > >So, why can't you just have an ordinary X-Y axis pair, 'slaved' to a
> X'-Y'
> > >set of axes? i.e. make one set sub-ordinate to the other?
> > >
> > >
> > Because you then lose the ability of cutting tapered walls on the
> > holes. That is something you can't do without a LOT of trouble
> > with milling cutters.
> >