Probe qualifications

[[To...]]

We have a customer who checks the validity their probe qualifications by running program that measures a ring gauge with all of their -Z probes and find the X and Y results can be off a .0001/.0002" Each of the probes measure the same circle on the ring, using same strategy and filter/outlier settings. Alignment is using cylinder in ring for spatial and X/Y origin. They edit the X/Y geometry in the Stylus System Management window until the program produces adequate results.

Zeiss has been in to check the machine and they cannot find anything wrong.

This is on a Contura using VAST-XT Gold in a relatively temperature controlled environment.

I can recall one time, a similar problem traced back to a faulty adapter plate. Didn't seem possible but the problem went away after they replace the adapter plate.

Any thoughts?

[[Me...]]

I would say the Customer needs a better understanding of Zeiss's calibration results compared to a ring gauge.

[[Ow...]]

Some probably already thought of questions.
How many probes are they checking in -Z and is it the same probes that continuously need adjusted or is it random?
Are they using the Master probe to set up alignment?
Does changing probe tips make a difference?
Is ring gauge secured properly?
Is each probe measuring the gauge in the exact same location in Z?
Are they measuring the cylinder for spatial with at least 3 circles or two circles recalled into a 3D line?
Would using the face of the ring gauge for spatial improve location repeatability even though you'd need to report the X&Y for each circle measurement?

[[Ri...]]

I've been down this road before.

I don't want to go down it again. Lol.

Make a measurement plan that uses the MasterProbe to measure the Reference Sphere (use the sphere macro - 6 points), and use that sphere in the Base Alignment as the X,Y,Z origins. Then take a qualified stylus system in the -Z direction and measure it using the same strategy, and report the difference. The error should be less than the allowable error of the machine.

If it isn't, it's most likely temperature changes, but it could be something else.

If you do see more than the allowable error, add a qualification of the stylus system (geometry requalification is fine) before it measures the sphere for deviation. Does your error go down?

Good luck.