[brice6678]

I recently ran a ball bar (150mm) on a Mitsubishi MPA H50A with a Fanuc 11M control on the X-Y plane. The positional error was just under .007" with x-axis scaling 248 ppm, with the majority of the error from scaling and scaling mismatch. The machine is setup for inch input with the NC side in metric. Looking at the x-axis (+/- 12" range) the ball screw was compensated +3 every 1/2 inch. After making a global change in the pitch error compensation table the positional tolerance was .0013" and the circularity was .0010".

Question 1: The ball screw has a 10mm pitch, should the pitch error interval be in multiples of 10mm, not inches?
Question 2: Is there another parameter that I should be using instead of pitch error for this type of scaling error?
Question 3: How can I compensate for reversal spikes, on he fanuc 11m control?

[underthetire]

You can make pitch comp anything you like, no set rule there.
Pitch comp works for this, but yes you can tweak other parameters but it can get involved.
Parameter 1851 may help for reversal spikes.

You really need to set up a laser for positional checks. The ballbar is a good tool, but not so much for what your trying to do.

[brice6678]

Anyone else out there that may have an opinion?

I understand "underthetire's" statement about pitch compensation error and that the interval can be anything you like. I thought their would be a best practice for compensating pitch errors.

I'd also like to know the specifics of the other parameters involve with pitch error compensation.

I also understand backlash and reversal spikes, but don't see anything on the control except backlash. I've read about acceleration/deceleration on other controls to correct for reversal spike but I haven't seen anything in the fanuc M11 manual or a procedure to correct the problem.

Any help would be greatly appreciated.

[underthetire]

Problem is each control can be set up differently with different parameter sets. I can look tomorrow for reversal spike corrections on the newer controls, but 1851 usually helps for that. Its called rapid backlash I think.
And I'll set up comps every inch on standard tenth machines, might go as low as .100 if I'm comping below a micron.