I'm sure the question has been asked but here goes. When I'm measuring a point I am turning the probe comp off. For some reason my upper management (who do not run the equipment) are telling me to NOT shoot them as an uncomp point. Other than me telling them that a comp point is giving me a false reading how do I explain o them this is the proper way?
I measure all my "points" with the "compensation off". You just need to be aware what type of measurements you are doing. I mostly measure my points as "auto vector points" with the "comp off". This works great for measuring tooling ball pads, or using hidden bar measurements where you need the measurement to be at the center of the reflector. I've been doing it this way for the last 10 years with PCDMIS and it has worked flawlessly, plus we also have Spatial Analyzer and when you measure a point in that software there is no radial compensation applied to the measured point....which means its "comp off". Hope this helps.
ok, to clarify what we are actually shooting, we are shooting drilled hole patterns on basically a flat surface. Diameter really isn't a factor only a TP pattern to itself. I shoot a plane-Level then I turn comp off measure points and construct a line from those points to stop rotation and use a measured point for my x,y axis. From that alignment I measure the hole pattern using uncomped points. I never seem to have a problem but as the story goes when you get bad numbers its not the part that's wrong it the way we shot it or the equipment. I understand that we can measure these as circles rather than points but like a said we are not looking for diameter. My QE thinks they should be done as single point circles. Thoughts
?
You have a couple of options that I would try. The first would be to do the original alignment that you stated above, and continue to measure the holes as "comp off" points. The second would be to try the single point circle and see if your measurements repeat the first set of data. The third that you might try is to measure your plane, then level your "z" axis and then if you know the nominal for all the holes create however many "auto vector points" you need and enter the nominal data for each hole. Make sure the "comp" is turned off and measure all the auto vector points that you created. Then do a new alignment; do a Best Fit (type 2D, 2D plane Z plus). This will only best fit the X and Y nominals of the auto vector points in your pattern. This normally works better for a pattern because you are taking out the error of your rotation through two points that may have a lot of included error. Say if your hole 1 is at -.002" on the X side and your other hole is +.005 on the X side, then you are including that degree of error into your coordinate system. Sometimes it is better to try a variety of methods to see if they repeat themselves, and to see which one works best for your application. If one of the holes is your origin then after the best fit alignment origin your coordinate system on that point and see what your pattern of holes deviate to there origin point. Hope this helps.
