Re: EMC-programming request

Dan Falck wrote:

> From: Dan Falck <dfalck@...>
>
> I have a request for the programmers who are working on EMC (Fred, Will,
> etc...):
> Would it be possible to make a probe (CMM type) software interface with
> XEMC? It would be great to be able to digitize a part in XEMC and then
> output a basic x,y,z coordinate file. It would be nice to make a teach
> mode for it at the very least.
> I would like to be able to write a G-code program that would move in .100"
> increments over a part, slowly lower the spindle, touch the part with a
> crude microswitch rigged probe, stop spindle movement, record the position,
> and move on.

Yes, I'd like this, too.

Also, in a similar vein, I have a VERY quick and dirty EDM, which I have
to conrol manually. If EMC could monitor the EDM current, and adjust
Z feed to keep that current within specified limits, and also back up the
Z if the EDM electrode shorts to the work, that would be neat.

Actually, I know how o do BOTH of these things, in a general sense.
You make a replacement for the user interface, and set it up to spoon-feed
RS-274 to the interpreter. Based on some reading either from an ADC
channel on the servo-to-go card for EDM, or an extra encoder channel
for the touch probe, it decides what to tell the interpreter to do next.
(Actually, both of these could be done with jog commands, rather than
RS-274.)

Oh, and the microswitch is a bit crude, you want something that can
detect touch with a precision of at least .001". I think a tiny linear
optical encoder can be cobbled up pretty cheaply out of various
bits and pieces. Or, you could get a dial test indicator, pull off the
dial pointer and attach a small encoder disk to the shaft. That was
what I was going to do.

Jon