Re: quiet list and EMC

Dan,

I went to the Gibson site and saw the Maestro with the lyre tailpiece.
It looks beautiful. Is this something you did for Gibson? How did you
program it?

Regarding probing, I did this several years ago using a Delta Tau PMAC
motion control board. I would like to duplicate this for the EMC motion
software. I haven't done it yet but I will eventually get to it. It
requires:

1. modifying the motion control application programming interface (API)
command set to call out for probing (EMCMOT_COMMAND in
emc/src/emcmot/emcmot.h).
2. Modifying the motion control status API with a field for the results
of probing, and a flag that the probing is in progress, finished, or
failed.
3. Modifying the motion control software to handle probing and reporting
the results, similar to how homing is done (emc/src/emcmot/emcmot.c).
I'm in the process of cleaning up the split between Cartesian XYZ moves
and motor moves, which are the same for mills but are different for
hexapod machines, so this will impact how this is done.
4. Modifying the external interface (emc/src/emcnml/extintf.h,
emc/src/emcmot/extshvintf.c and equivalents) to bring in a trigger from
a touch probe that will cause a latch of motor positions and conversion
to XYZ.

I did this with the PMAC using a Renishaw touch trigger probe. What did
you plan on using? To get any accuracy the probe needs to be compensated
for the eccentricity of the ball, and probe lobing error. If you are
only interested in digitizing to a few thousandths then this is not
necessary.

Also, without hardware support for latching encoder counts with an
external trigger, the latch needs to be done in software and this
introduces a latency on the order of the servo cycle time. The
uncertainty is the probing speed times the servo cycle time. So, to
minimize error, you need to probe slowly. Digitizing can take quite
while.

--Fred