hexagon logo
Register Sign in
  • State Not Answered
  • Replies 4 replies
  • Subscribers 227 subscribers
  • Views 1968 views
  • Users 0 members are here
Options
Related

Auto Calibration Issues

Vulcan

Hi All,

I am currently attempting to roll out Auto Cal at the start of all of our programs to save time calibrating probes - we used to auto calibrate over a weekend (would take around 40hrs to fully calibrate) but we found that after a few days we would see issues coming from our measurements (error varied between 0.005 up to 0.15) - all measurements where manually verified and then re-calibrated and re-measured and wold then come in and agree with the manual check.

Please see below a snippet of how our autocal works currently

 

C9         =COMMENT/YESNO,NO,FULL SCREEN=NO,AUTO-CONTINUE=NO,OVC=NO,
            Do Probes Require Calibration?
            IF_GOTO/C9.INPUT=="YES",GOTO = PROBES
            GOTO/SKIP_PROBES
PROBES     =LABEL/
            LOADPROBE/LEITZ TREE 01 1X50 P&P
            TIP/T1A0B0, SHANKIJK=0, 0, 1, ANGLE=0
            AUTOCALIBRATE/PROBE, PARAMETER_SET=AUTO_CAL, QUALTOOL_MOVED=NO, 
                          CHECK COLLISION=NO, SHOW_SUMMARY=NO, OVERWRITE_RESULTSFILE=NO
            GOTO/SKIP_PROBES

The issue we are seeing is that with this they seem to autocalibrate over eachother overwriting results (overwrite is turned off and results file is not updated) however if we do a manual calibration then the results come into real life values as oer manual checks

does anyone have any suggestions or any help as to why this could be happening.

  • 0

    Rename_all_your_probes_to_remove_spaces_and_take_out_any_special_symbols_like_&_then_see_if_it_makes_any_difference.

    Your issue is probes with different names are over wiring each other though?

  • 0

    The results file, which is referenced in the command, is NOT the same as the calibration results, which are stored in the PRB file.  They would have the same values but they are not the same instance.  Whether or not you over-write the results using the AutoCalibrate command, they will ALWAYS be updated in the PRB file to the current values and prior values are destroyed.  I hope I am not confusing you, but if yo don't overwrite the results file they are merely appended, so you will have a series of calibration results.  Only the last one will be used in the PRB file.

  • 0

    When I started at this new job, there were about 50 or more probe builds to choose from and some of those had pretty much every tip angle possible. Funny thing was that there were only about 6-8 actually built probes. One of the first things I did is delete all of them. Created the ones that were in service with a handful of common tip angles. Most programs were fine with what I started with but when I needed to add here and there I calibrated the "primary" probe and the probe I added angles to. I just calibrated them all at the same time because it didn't take long and I always felt better about the calibration that way. 

    I'm wondering why it would take about 40 hours to do a full calibration. I have 8 probe builds and probably about 75-100 tip angles amongst all of them total. Takes about an hour and fifteen minutes. Have you thought about purging unused tip angles? If you're going to do a calibration at the beginning of each program anyway, might be worth it to purge and after a few months your full calibration might be shortened significantly.

  • 0 in reply to ML

    We don't know his situation, while it does seem extreme. My full calibration process takes 10-12 hours, but it has to support almost 15 years of programs and a very large program library at a moment's notice. 

© 2025 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement