Passive probe manual calibration and accuracy [ITTP]

[Snecx]

Recently I have purposely made the tip of the ITTP stylus non-concentric, to get an “effective concentricity” upon probing, if it can even be called that. It has been working well and it is highly repeatable. Now I am not being affected by the pre-travel variations of the stylus tip after doing this "manual calibration".

I have taken a video to illustrate how the ITTP is working for me now, repeatability and accuracy of better than 0.005mm/0.0002" is achievable.



Rotating the probe exactly 180° in the spindle gives me a reading that’s only off by ~0.0025mm one of the probed direction (X+). I suspect it is just a negligible error at this point. It also measures my fixture center hole quite precisely too.

To have this works reliably, I mark both my spindle case and spindle nose, with my ITTP inserted to the spindle in the exact same orientation. My probing speed is 1000mm/min and this is carefully matched to the "effective diameter" of the stylus tip. It does not have to be slow, only has to be the same speed every single time.

In order to get consistent readings on both X and Y axis with a stylus that works on 3 contact points, I had to be very careful with the orientation as well. With the ITTP, I managed to get this with one of the screw facing me in the Y- direction. If it is off by even a few degrees in the spindle, the measurements will be totally off, back to the "default accuracy" achievable by these tools.

If the Tormach mill is properly compensated for backlash/lost motion, the positioning can be very accurate. For example, using it as a poor man's CMM to take measurements of known gaugeblocks, gives extremely precise dimensions, to much better than 0.01mm precision. My ballscrew/ballnut has been quite accurate when I tried to map it, so I guess I got lucky as well.

And of course there are thermal expansion and all other factors involved with such a limited machine. When properly compensated (manually), they can still do high precision work. And I use the ITTP primarily as the tool to compensate for thermal changes.

I just wanted to share this to show what's possible on a Tormach mill with some careful manual calibration. Maybe some hobbyists like me who wants to achieve better precision from a hobby grade machine, can benefit from knowing the possibilities.

Similar Threads:

• Need Help!- Passive probe calibration in PathPilot
• Need Help!- Passive Probe Calibration
• Passive Probe Accuracy
• Need Help!- Tormach 1100 Passive Probe Calibration Issue.

[jwatte]

Fascinating!

If it is off by even a few degrees in the spindle

I line up my ITTP with a mark I've made on the spindle, but I do it by eye, so I'm sure I'm off by a few degrees.
That being said, I believe I'm getting 1-thou or better out of the ITTP, and tool deflection and other parts of my setup contribute likely just as much (or more) than the probe.
After all, I mainly do structural parts, where, if a screw hole is off by a thou or two, that's not a problem, and if a bearing race is slightly undersize, well, aluminum is softer than the bearing steel, and I have a swing press ...

[kstrauss]

Perhaps I misunderstand but I think that he is referring to the orientation of the probe relative to the X/Y axis in order to minimize the error resulting from lobing.

[Snecx]

Perhaps I misunderstand but I think that he is referring to the orientation of the probe relative to the X/Y axis in order to minimize the error resulting from lobing.

You are right, keeping the orientation just right for accurate X/Y axis probing with just 3-point-contact. One can keep it at a specific orientation and it would repeat, but there would be a slightly different reading for both X and Y.

Also purposely making the stylus tip non-concentric, I rotate the probe 180º and adjust the "effective concentricity" from the readings in PathPilot. After this adjustment, if I put a DTI to the stylus tip, it is not mechanically concentric to spindle rotation. But what matters is the effective concentricity than the actual stylus tip's concentricity. It can be trickier to get right, and whether it is worth the hassle or not depends on the individuals and their required precision.

Fascinating!
I line up my ITTP with a mark I've made on the spindle, but I do it by eye, so I'm sure I'm off by a few degrees.
That being said, I believe I'm getting 1-thou or better out of the ITTP, and tool deflection and other parts of my setup contribute likely just as much (or more) than the probe.
After all, I mainly do structural parts, where, if a screw hole is off by a thou or two, that's not a problem, and if a bearing race is slightly undersize, well, aluminum is softer than the bearing steel, and I have a swing press ...

This is an extreme use case as most parts don't need such precision. However I find reducing all the errors to a minimum allows for a much more accurate part due to tolerance stack. Tool deflection, tool wear, thermal expansion and more, all play a role. Being able to reduce errors in probing allows for a much more accurate final part.

I only hope to share what's possible, even when it is impractical. As a hobby, it sure is fun to know how much we can squeeze out of these mills.