hexagon logo
Register Sign in
  • Replies 15 replies
  • Subscribers 221 subscribers
  • Views 157 views
  • Users 0 members are here
Options
Related

Change in actual position vs theoretical

msteffen99
Just had our machine calibrated on 12/6 and I ran an automated probe calibration program late last week. Started crashing on programs that have a tight target window prior to alignment and notice that we have quite a bit of variance between actual hit points versus theoretical targets. X is +.0081, Y is + .00285 and Z is -.0304.
The technician did comment that our calibration sphere was loose and that it was ok to super glue it back into place. I did that this morning and reran the probe calibration probe with no change, still crashing.

Machine Brown & Sharpe one running PCDMIS 2020 R1

any ideas?
Thanks,
Mike
Parents
  • Is it ALL probes, or just 1 probefile for this issue? That can change your approach to finding a solution.

    Do you have a loose probe tip in a build? I've seen loose tips calibrate and pass for StdDev then give errored results. (How it passed I'll not know, but we did get seemingly crazy actuals by comparison to the theo targets)
    I realize that your description says you used the glue after, did any get on the qual sphere?
    Last thing to check, if you are using an L or other Starcluster configuration, have they rotated from their original setup? Calibration can correct to a point, but the mechanical application could result in shanking where you wouldn't have before.

    ^Just my thoughts when reading about your issue, hope you find your solution quick and easy!
Reply
  • Is it ALL probes, or just 1 probefile for this issue? That can change your approach to finding a solution.

    Do you have a loose probe tip in a build? I've seen loose tips calibrate and pass for StdDev then give errored results. (How it passed I'll not know, but we did get seemingly crazy actuals by comparison to the theo targets)
    I realize that your description says you used the glue after, did any get on the qual sphere?
    Last thing to check, if you are using an L or other Starcluster configuration, have they rotated from their original setup? Calibration can correct to a point, but the mechanical application could result in shanking where you wouldn't have before.

    ^Just my thoughts when reading about your issue, hope you find your solution quick and easy!
Children
  • in reply to R2ah1ze1l
    I checked the probes. We only use 2 and they don't have any extensions.
    The machine was off prior to gluing the ball in place, I used minimal glue because my first attempt had too much and I couldn't get it seated.

    I just didn't know if the calibration routine would affect anything with the home position
  • in reply to R2ah1ze1l
    only the "Z" would be affected the most in this case. adjust your prehit/retract distances in probe calibration utility box and see if that helps it from tripping up.
    or
    In the Measure Probe dialog box you have an option for 4 different modes; Manual, Man+DCC, DCC and DCC+DCC.
    We have it set to DCC+DCC. This causes the machine to probe the single point (just like you answered "yes" to the question "Has it moved?") and then it does the 3 hits to establish where the sphere is.
    After I switched to this method all of the problems we were having with calibration went away.
    This takes a bit longer since it does the 3-point thing on every tip, but it feels better knowing that we have a "good" calibration every time.
  • in reply to R2ah1ze1l
    That's the weird part. It calibrated fine. We didn't notice a problem until we ran a program that had to measure a 0.125" hole at the start of the program and hit the side of the part instead of the hole.
    I could rewrite the program to accommodate for this spatial error that has appeared, but then I would have to do that for a lot of programs. I'm wondering if during the annual calibration that the technician altered something or maybe cleaned out some garbage in our ways that altered the home. Just doesn't make sense to have all 3 axes move .03 or less out of nowhere.
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement