CNC edgefinding/touch probe update

I ordered and received the edgefinding touch probe that I spoke of last
week. Specifically, we purchased an Acu-rite 38-80-00-155, which is an
electrical contact type probe (ie between the probe stylus and the
workpiece). The probe is self centering, with a 0.5" shank (ie works
with R8 collet mills...) and a 0.2" diameter tip. The stylus has about
1/2" of movement in each direction in case of over run. THe probe cost
$250, plus shipping. I have no affiliation with Acu-rite.

There are less expensive touch probes on the market, but:

a) they are usually battery powered and visual, ie they light an LED.
b) they don't readily tie into a CNC controller, although they could be
made to...
c) most have a 1" shank, ie they won't fit an R8 mill.
d) most have a 0.4" sensing ball, which is too large for a lot of holes.
e) most don't do Z axis touches. The Acu-rite does.
f) most have limited over travel capability.

We mounted the the touch probe in our CNC mill, which is running
TurboCNC as the controller. TurboCNC had a G50 command for center
finding a hole. The G50 routine had a few bugs, which I fixed. I also
added a G51 routine for center finding on a protruding shaft and a G52
routine for edge finding on an ... edge !

I've got mixed feelings about the performance of this setup. First, the
Acu-rite touch probe itself works well, ie it centers itself well, has
good over travel capabilities, etc. No problems there.

The issue I have with this setup is that it utilizes an electrical
connection between the workpiece and the probe to sense when the edge is
"touched". In my (limited) experience, one must have perfectly clean
parts in order to get repeatable results. One can get Total Indicated
Runout measurements of 0.0004" on a hole, but it might also take 3 or 4
centering efforts to achieve that.

I'll continue to refine our edge finding efforts. If anyone has any
advice, I'm all ears.

--
Kim Lux <lux@...>