hexagon logo
Register Sign in
  • Replies 18 replies
  • Subscribers 223 subscribers
  • Views 263 views
  • Users 0 members are here
Options
Related

SP25 calibration issue

JEFMAN
Please note that this thread is not in the OFF TOPICS section, so only clever answers are welcome !!!

I greet you all, colleagues.
I have a small problem that tends to come and go when calibrating angles. I work with many styli and use a 9-slot magazine for this. Each stylus has about 10 different angle positions. When I did the calibration, two angles on different probes were out of tolerance. One probe has a CM25-2 probe and the other CM25-3. And at the same time, PSDMIS accepts the calibration result without asking the operator. The next day I did a similar calibration and the result was OK.And always out-of-tolerance angles equal to A90 or more than A90 It is very strange . I have been working in this mode for 3 years and there were no such problems. And this year it started. I check the probes visually and do the cleaning before each calibration the same way always. What could be the problem ?
  • I think the rotation head is missaligned, because of more than 1 mm difference between theo and meas on 2 axes at A90.
    In addition, you're using a 10 mm ball, so, function of the sphere stem, maybe some hits shank the stem (even if the std dev seems to be right.
    I don't know enough about SP25 to help you more...


    Could you change decimal place to 4 and post some results ?
    You could also change the sphere axis (for example I 0.707,J 0, K 0.707) and look at the results.
    Then, measure the sphere withe the same number of hits and the same number of rows, and check the form.
    Have a nice day, all.
  • Thx to Cris_C Slight smile



    I see what you mean. Those are some odd results. I see that the standard deviation looks really good for all angles, but somehow the probe radius deviation varies at different angles. It is like the results say that the CMM is very certain that the probe radius is different at different angles. That makes no sense.

    If the standard deviation were larger I would suspect that you have worn spots on your stylus or something is loose, but that is not the case.

    My next guess is that your SM25 module is wearing out, but based on the names of your probes, it looks like you are having the same problems with both a SM25-2 and SM25-3 module. If that is so, it would rule out the module as a problem.

    Maybe there is a problem with the SP25M probe body itself. Make sure the optical window is clean and use some cleaning putty to clean all the magnets and kinematic contacts. Something may be causing the modules to rock a bit when the probe is at an A90+ angle.

    It is also possible that your probe head is getting worn out at some angles. I don't have any suggestions to troubleshoot that other than to try a different probe head, but you probably just have one. Maybe you can at least check for loose fasteners.

    While you are at it, check that the qualification sphere has not come loose on it's mount. Also check all the probe/styli connections. Probably not the problem, but doesn't hurt to check.

    It might also give you clues to create a quick measuring routine to measure the qualification sphere with lots of hits using different probe angles. You can review the diameter results to check if you get different diameter measurements with different angles. Also, you can graph the roundness to see if the problem is just with certain hit vectors. That might reveal a problem.

    Good luck. I hope the problem is something easy/simple.

  • If using an 'L' tip or a starcluster it appears that the tip itself may not be squared to an axis. But Chris has a very comprehensive list of items to help.
    Try partial calibration instead of the full scan calibration, see if your results change significantly. While performing partial calibration you can utilize the 'user defined' area of calibration mode to ensure the probe isn't able to 'accidentally' touch the stem of your qualification sphere.

    Last thing I think you could have is an incorrect component selected in your probe builds. The selection is close enough for you to utilize but could be why you see deviations that appear so drastic.
  • in reply to R2ah1ze1l
    Thank you very much, I will take your recommendation to check how I assembled the dipstick
  • : Hi Sergiy, what's new about your calibration issue ? (the other post is too far from metrology, partly because of me... (sorry for this !)
  • Hello Jeff. Excuse me for being away for such a long time (production issues forced me to work on a tough problem for two days).


    I calibrated in automatic mode and set the marker to the reset of probes. For the probe master, I told the machine that the sphere changed position and at startup I selected YES (take a point on the sphere manually). As a result, for the A0B0 angle, the probe master THEO and MEAS the measurements became the same. All probes were calibrated OK. But on the probe SP25M-2_L40_D1-D1, the set of angles did not pass the calibration. I observed visually ---- there was a feeling that the machine does not see zero(center of spher) for this probe. I checked the probe assembly --- everything matches. As a result, for each angle of the probe SP25M-2_L40_D1-D1, without changing the position of the sphere, I manually calibrated oll angals (each time pressing YES to take the point of the sphere manually) After that, all the results became ok. After that, I measured all the standards of our production parts ---- Everything is ok. That is, PC-DMIS has accepted all the positions of the probes. The problem is solved, but now I think where this problem came from. I have to find out why this started so that I can have an understanding of the problem in the future.
    Attaching photo
  • in reply to JEFMAN
    Hello Jeff. Excuse me for being away for such a long time (production issues forced me to work on a tough problem for two days).
  • Looks like the KGB came in & cleaned the place up Rolling eyesRolling eyesRolling eyesRolling eyesRolling eyesRolling eyes​​​​​​​
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement