New toy... economical 3D probe

Martin, you should raise that issue with the supplier.

Have done. He wants me to do a few tests at different settings.

Martin C

Would be great if you could share the results please Martin. I have downloaded and installed the wizard and had a play, including the tip calibration where it didn't seem to do anything alarming, and probing an external edge. Haven't tried a hole yet, or paid anything!

I had another play with ProbeIt, this time before doing anything else such as probing with my usual macros or running any code. The issue seems to be that ProbeIt does not respond as quickly as my macros triggered by clicking a button. This may be due to the simplicity/complexity of the different probing options or macros. I put a spring over the tip of my rigid probe to try ProbeIt and it is clearly over-running after contact. With my macros I can do a fast approach at 330mm/min and get what looks like an instant stop on contact however ProbeIt bends the spring at both a 330mm/min and a 33mm/min approach. This will not be an issue with a probe such as the probe being discussed in this thread which has a 4mm leeway in the X and Y axes.

I could buy one of these probes but the simplicity and low cost of a home made rigid probe seems hard to beat as long as you have a system that responds rapidly on contact and you don't want to use it on non-conductive materials (I pull out the edge finder for such times).

Martin C

That's a bit puzzling Martin! I assume that ProbeIt uses the standard G31 command, which is executed at a very low level in Mach3 - either in the PC with a parallel port or in the actual motion controller - so the wizard shouldn't be able to affect the timing at all. Did you also raise this with Eric?

Based on my communications with Eric, the ProbeIt originator, both my macros and ProbeIt use G31. The operation takes place in the Smooth Stepper I use. The only thing I can imagine is that there is either something in or not in one of the macros that is causing the issue. My macros do a "while ismoving() wend" loop directly after the G31 is called so that nothing else happens until the G31 finishes. Without this it is possible the lookahead is doing something in ProbeIt that delays the response to a G31 move and so causes an over-run.

My probe is a Ø16 piece of ground steel with a thread on the end to suit an ER collet or a Posilock collet. The other end is drilled out Ø10 and a piece of acrylic is fixed in place with anaerobic adhesive. This is then drilled and reamed Ø6 and a Ø6 hardened and ground dowel is pushed in. Nice and cheap and accurate but not immune to over-run damage.

Martin C

Thanks Martin, I'll have a play. We had a long exchange over on Mach3 Support about how Mach does probing, and it isn't to my mind very sensible. Basically as I understand it the G31 movement starts decelerating when the probe triggers and the axis value it stops at is registered as the trigger point, so will always be slightly over the correct value by an amount which depends on your acceleration profile. I wonder if ProbeIt modifies the profile in some way?

In UCCNC though the axis value is registered immediately on trigger as well as deceleration starting, which seems the right way to do it to me.