Manual Probe Hits Read Very Differently Than DCC Hits

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Manual Probe Hits Read Very Differently Than DCC Hits

    I've found something quite perplexing, hoping it is something silly. When I take manual hits on the granite of the CMM, I get an error value of around .012 (inches). When I copy and paste that plane and run it in DCC mode, I get an error value of .0002. I am positive that I'm taking the points appropriately. Does anyone have any ideas? Figured I'd give this a shot before calling in.

    Thanks

  • #2
    what hardware are you using? TP2, TP20, TP200, SM-25?
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

    Comment


    • #3
      As Matt is alluding to some probes will do this better than others. Also I believe some machine controllers just aren't designed for this.
      Automettech - Automated Metrology Technology

      Comment


      • #4
        Originally posted by NinjaBadger View Post
        As Matt is alluding to some probes will do this better than others. Also I believe some machine controllers just aren't designed for this.
        well, what I've seen is that when taking manual hits with the SP25 stuff, the vectors are crap compared to the TP20.
        sigpic
        Originally posted by AndersI
        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

        Comment


        • #5
          Using a TP20. The vectors are also a lot worse with manual hits. I've tried 3 different probes, getting the same results with all of them.

          Comment


          • RIDER
            RIDER commented
            Editing a comment
            Do you get the same results when you disable the x and y movements on the jogbox?

          • RanDawgg
            RanDawgg commented
            Editing a comment
            Yes, the results were pretty much the exact same. What worries me is that my manual hit readings are VERY consistent. I'd be more understanding if it varied a bit between attempts, but it isn't.

        • #6
          touch speed, twitches on the joystick, all can effect manual measurements. Which is why final part/fixture alignments should always be done in DCC.
          sigpic
          Originally posted by AndersI
          I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

          Comment


          • #7
            If this is Hexagon hardware the slow button must be enabled for manual measurements.
            Systems Integrator
            Hexagon Manufacturing Intelligence

            Comment


            • RanDawgg
              RanDawgg commented
              Editing a comment
              I was not aware of that. I've read a lot about PC-DMIS over the last few weeks, and I haven't seen that mentioned at all.

          • #8
            Manually measured 40 points across the granite, in turtle mode the entire time. Results are comparable to the DCC runs. I saw the setup technician taking points way harder than that of turtle mode during calibration, but I suppose turtle mode will now be the only way I take manual points. Really strange.

            Comment


            • #9
              Nothing strange here, slow mode exists just for it !!!!!
              If you measure a manual alignment without it, it will be inaccurate, but a dcc align after will help to find more accuracy.
              When the probe touche the part, there's in side a electrical circuit that opens. Then, you retract the probe, and when the circuit is closed, the hit is taken. If you do it very fast,you can't obtain accuracy, because of the dynamics of the cmm.
              In addition, the ball radius is compensated along the closer cmm axis, that's why a BF circle constructed from manual hits is often inaccurate.

              Comment


              • #10
                Originally posted by JEFMAN View Post
                Nothing strange here, slow mode exists just for it !!!!!
                If you measure a manual alignment without it, it will be inaccurate, but a dcc align after will help to find more accuracy.
                When the probe touche the part, there's in side a electrical circuit that opens. Then, you retract the probe, and when the circuit is closed, the hit is taken. If you do it very fast,you can't obtain accuracy, because of the dynamics of the cmm.
                In addition, the ball radius is compensated along the closer cmm axis, that's why a BF circle constructed from manual hits is often inaccurate.
                That is for the SM25, the TP20 is 'done' when the circuit is broken.
                sigpic
                Originally posted by AndersI
                I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                Comment


                • JEFMAN
                  JEFMAN commented
                  Editing a comment
                  Right !
                  I did a mistake with a special cmm (spheritical coordinates - the tp2 was uni axial) which worked in the other way... !

              • #11
                I just skimmed through the training book for PC-DMIS for CMM 101 and it never once mentions to take manual points in turtle mode. I have scoured the internet and all help files. How can something so important never once be mentioned?

                Comment


                • NinjaBadger
                  NinjaBadger commented
                  Editing a comment
                  Because people very rarely measure anything manually.

                  If your controller supports it then manual hits in normal (not slow) mode are most likely fine for a manual alignment.

              • #12
                FWIW I have 2 B&S Globals with TP20s and I never take manual hits in slow mode and I have never seen anything like this. Manual and DCC readings are consistently within <.001 of each other

                Comment


                • #13
                  Interesting.. I changed the speed of the CMM to resemble the speed of turtle mode, and I still get a bad form error. Also, whenever I take 2-3 points on the granite, it asks if I want to change my work plane. How on Earth could it request that when I'm taking points with a very clear vector direction?

                  Mike- that is what all of my contacts have said as well. Not one of them uses turtle mode, nor were they ever told they had to for accurate readings. They've all been to the certified training.

                  This CMM is a month old for what its worth. 7.10.7 SF.

                  Comment


                  • #14
                    Originally posted by RanDawgg View Post
                    Interesting.. I changed the speed of the CMM to resemble the speed of turtle mode, and I still get a bad form error. Also, whenever I take 2-3 points on the granite, it asks if I want to change my work plane. How on Earth could it request that when I'm taking points with a very clear vector direction?

                    Mike- that is what all of my contacts have said as well. Not one of them uses turtle mode, nor were they ever told they had to for accurate readings. They've all been to the certified training.

                    This CMM is a month old for what its worth. 7.10.7 SF.
                    This is because this is a mode setting, it's not about how fast the CMM is moving.
                    Systems Integrator
                    Hexagon Manufacturing Intelligence

                    Comment


                    • RanDawgg
                      RanDawgg commented
                      Editing a comment
                      Wasn't aware of that either! In the help center, the only document that refers to toggling turtle mode just says: indicates that the UJB is in slow mode, or fast mode.

                  • #15
                    From Hexagon Support (search keywords: Jogbox slow)
                    Manual diameters are not measuring correctly - When measuring a ring gage it is deviating by 0.010 - 0.15", yet measuring this ring gage in DCC is within expected tolerance.




                    26-May-2018•Technical Support TopicsAttachment
                    Details

                    Note the following general statements:
                    1. Manual measurements are always less accurate than DCC measurements.
                    2. Some machines must be in a certain mode to take accurate manual hits. In particular with a Leitz protocol machine it is necessary to have the Jogbox Slow button on when taking manual hits. This changes the state of the controller so that it is "expecting" hits.
                    3. Manual measurement requires some practice and more points than DCC measurements. Practicing on the qualification sphere will let you know how you are doing and give you 3D practice. Measure the sphere, set your origin on it and measure it 5 times in a row. Evaluate your results. When you get an acceptable range, then qualify the tips. Then, measure your parts. In DCC you always qualify the tips at the same parameters used in the measurements. Manually, practice will help you maintain your touch speed technique, helping your accuracy.

                    (Topic ID 16101-347)

                    Comment


                    • louisd
                      louisd commented
                      Editing a comment
                      I do agree however, it should be stated in more places than just a technical document that's buried deep.

                    • RanDawgg
                      RanDawgg commented
                      Editing a comment
                      Great find! To think i thought I was thorough..​​​​​​​

                      At least I know now.

                  Related Topics

                  Collapse

                  Working...
                  X