This might be a little drawn out, but I'm looking for some help with a performance problem with our new CMM.

I set up 30 probing points with direction vectors on the Masterball (all driving numbers computed in Excel first and keyed in). I had just calibrated the same probe (Tree#1 - our main probe), only 3 hours earlier; machine was otherwise idle.

Further I ran the program with an iteration so that it was sure to be properly "finding" the sphere. (Using stored coordinate systems.) Also all points are probed with one styli angle from top of sphere down to latitude 12° below equator.

I reported out the 30 individual readings and found a problem: 10 readings were beyond the envelope created by the volumetric probing error specified for our machine. Also the volumetric probing error itself was huge 0.018mm, and even after removing three apparent flyers, it was 0.0036mm, significantly over our P tolerance (0.0027mm).