Auto Calibration

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

  • Auto Calibration

    For "Auto Probe Calibration" . so i inserted below command. but while execution it calibrate angles, which are all used in this probe. i need to run only angles which are used in this program. how to set "Parameter Set' to calibrate angles used in this program?
    i am using PC-DMIS 2019R2.

    STARTUP =ALIGNMENT/START,RECALL:USE_PART_SETUP,LIST=YES
    ALIGNMENT/END
    MODE/MANUAL

    PREHIT/0.1
    RETRACT/0.1
    MOVESPEED/ 100
    SCANSPEED/100
    FORMAT/TEXT,OPTIONS, ,HEADINGS,SYMBOLS, ;NOM,TOL,MEAS,DEV,OUTTOL, ,
    LOADPROBE/PROBE_4
    TIP/T1A0B0, SHANKIJK=0, 0, 1, ANGLE=0
    AUTOCALIBRATE/PROBE, PARAMETER_SET=ALL-TIPS-WITH-DEFAULTS, QUALTOOL_MOVED=NO,
    CHECK COLLISION=NO, SHOW_SUMMARY=NO, OVERWRITE_RESULTSFILE=NO






    Calibration.PNG
    Last edited by Satheeshkumar; 11-12-2019, 07:43 PM.

  • #2
    To create a parameter set, open the probe dialog box of you the probe you wish to create the set for. Select the probe tips or all the probe angles. Click the measure button. In the Measure Probe dialog box, on the lower right of the window, there is a section called Parameter sets that has a name with a drop down box and save and delete buttons.

    So lets say we are going to make a parameter set for our master probe. I have have a 4x10 on tp20 and I have no angles, only A0B0. So I'll just click on the name box and type Master_Probe and click save. You should see a prompt that the parameter set has been created. So now go ahead a measure the actual probe. Some might say you don't have to actually measure it, but sometimes my parameter set wont appear in my code it I don't so I always do.

    Now go back to your program and F9 on the autocalibrate code and you'll see a little dialog box open. This is my master probe and my sphere never moves so I'm going to select Yes_DCC and selct my parameter set of Master_Probe, and show summary results.

    So if you run all that code, it will automatically select the master probe, run that parameter set and display calibration results.

    Comment


    • #3
      if you have another probe like a 4x20 with 10 different angles on it, do the same thing to create a parameter set. You can either select certain angles or all of them. If you are making a calibration program, you might as well calibrate all angles for each probe. i don't think many people make parameter sets for a hand full of angles but idk. Open the probe dialong box, measure button, create and save a name for the parameter set. add another line of code for Load 4x20 probe, tip 1 a0b0, autocalibrate tab and select the parameter set. For this parameter set, you'll select No to the sphere has move question since we already located the sphere with the master probe. Yes manual hit and yes dcc are only used on the master probe and for only locating the sphere.

      Say you created the next parameter set and you need to add an angle to the 4x20. open the probe dialog box, add the angle, measure button, select the parameter set in the measure window and click save to update the new angle to the set. Now when you run your program, the newly added angle will be calibrated like your other angles.

      Comment


      • #4
        Now if you want to get fancy, you can automatically print out a pdf with all the calibration results and even use variables and assignments to check the standard deviation of the results and have pcdmis re run a parameter set if the results are out of you criteria. I use if the standard deviation is greater than 0.0005", go back and repeat once, and if the results are still out, go to an error message stating help is needed.

        this way operators just run the cal program without having to manually calibrate anything and we have a record when calibration is done.

        Let me know if you have any other questions.

        Comment


        • #5
          Here are a few rules about "master probes"

          1. if you are using a tp20 (touch trigger), use a large diameter ruby with a short stem like 4x10 or 5x10. If you don't have something these size just use what you have but just make that probe the master probe.
          2. If you are suing a LSP scanning head, you can use a 5x20. That is only size that is supposed to be used for the lower matrix calibration so might as well make it the master probe for the probe calibrations.
          3. I don't use my master probe for any measurements, I only use it for locating the sphere and calibration. My parameter set just has angle A0B0. Some people have to use their master probe for measurements and that fine. Some people have multiple angles with their master probe and that fine too. You gotta do what you gotta do sometimes.
          4. On some of my machines, the cal sphere physically doesn't move from the granite. So when I add an angle to a probe, I update my parameter set for my cal program and calibrate that 1 new angle by answering no, sphere has not moved and everything is good. My other machines however, the sphere has to be moved in and out for parts. So when I need to add a new angle, I have to first place my sphere back where I originally had it, run my master probe with YES DCC, load my probe with new angle, add my angle to the parameter set for my cal program, and then run that new angle. Awesome.
          5. You only need 1 master probe. I don't think people have more than one master probe.
          6. This doesn't apply to me but some people have to use 2 calibration spheres. I know the procedure is somewhere on the forum, just search for it if you need it or maybe someone will chime in on this thread.

          Comment


          • #6
            Thank you so much acgarcia

            Comment


            • #7
              Originally posted by acgarcia View Post
              Now if you want to get fancy, you can automatically print out a pdf with all the calibration results and even use variables and assignments to check the standard deviation of the results and have pcdmis re run a parameter set if the results are out of you criteria. I use if the standard deviation is greater than 0.0005", go back and repeat once, and if the results are still out, go to an error message stating help is needed.

              this way operators just run the cal program without having to manually calibrate anything and we have a record when calibration is done.

              Let me know if you have any other questions.
              acgarcia - I've been looking for this for awhile. We have out master calibration program and one other program with a calibration in the middle. What I would like to do is have the program stop and put out a comment stating that that the calibration has failed and asked if they would like to re-run it. I can't seem to figure out how to get the cal results into a variable. Any thoughts?


              Thanks in advance.

              Comment


              • #8
                acgarcia I fully understand parameter sets and have made programs that run with autocalibrate, the problem I have is the actual parameter sets seem to randomly disappear, deleted...poof they are gone. I have given up and do all calibrations without autocalibrate. To be honest now that I have the whole calibration thing pretty much figured on pcdmis I don't need programs, but operators do.

                Any idea what is causing that? something I am doing or was there a known bug in earlier versions? I am on 2013mr1.

                Comment


                • acgarcia
                  acgarcia commented
                  Editing a comment
                  I have no idea Douglas, earliest version I've used is 2015. In my experience which is different for some other users,when ever I created and saved a parameter set, I had to measure it was well so the parameter set would stick. So open the probe dialog box, select your probes, click measure, make your parameter set, save your parameter set, you'll see a prompt saying the set was saved, then click measure and actually calibrate the probes. Someone on this forum told be I didn't' have to do that but again, my experience is that I had to so then I could select it for my autocalibrate program.

              • #9
                Originally posted by John Kugler View Post

                acgarcia - I've been looking for this for awhile. We have out master calibration program and one other program with a calibration in the middle. What I would like to do is have the program stop and put out a comment stating that that the calibration has failed and asked if they would like to re-run it. I can't seem to figure out how to get the cal results into a variable. Any thoughts?


                Thanks in advance.
                Thoughts? Yes.

                All I do is use a variable for each probe/angle to check the standard deviation of the tip. then I use logic to see if the results are larger than or equal 0.0005" to repeat which usually fixes things. Here is one of my star probes. See below.

                Code:
                AUTOCALIBRATE=LABEL/
                AUTOCALIBRATE/PROBE, PARAMETER_SET=Star_Calibrate, QUALTOOL_MOVED=YES_DCC,
                SHOW_SUMMARY=YES, OVERWRITE_RESULTSFILE=NO
                BEGIN AUTOCALIBRATE RESULTS FOR PROBE VC0022 STAR USING SET Star_Calibrate
                Probe file=VC0022 STAR Date=10/3/2017 Time=6:02:09 PM
                Dixie Sphere CENT X 21.0250 Y 9.8345 Z -12.3495 D 0.7500
                T1A0B0 THEO X 0.0000 Y 0.0000 Z 8.2262 D 0.1969
                T1A0B0 MEAS X -0.0296 Y -0.0223 Z 8.1695 D 0.1948 StdDev 0.0002
                T2A0B0 THEO X -0.9252 Y 0.0000 Z 7.1927 D 0.0787
                T2A0B0 MEAS X -0.9491 Y -0.0197 Z 7.1290 D 0.0770 StdDev 0.0002
                T3A0B0 THEO X 0.0000 Y -0.9252 Z 7.1927 D 0.0787
                T3A0B0 MEAS X -0.0243 Y -0.9461 Z 7.1275 D 0.0765 StdDev 0.0003
                T4A0B0 THEO X 0.9252 Y 0.0000 Z 7.1927 D 0.0787
                T4A0B0 MEAS X 0.9035 Y -0.0166 Z 7.1416 D 0.0769 StdDev 0.0002
                T5A0B0 THEO X 0.0000 Y 0.9252 Z 7.1927 D 0.0787
                T5A0B0 MEAS X -0.0207 Y 0.9022 Z 7.1400 D 0.0769 StdDev 0.0002
                END AUTOCALIBRATE RESULTS FOR PROBE VC0022 STAR USING SET Star_Calibrate
                LOADPROBE/VC0022 STAR
                ASSIGN/STD1=PROBEDATA("STANDARDDEVIATION","T1A0B0")
                IF_GOTO/STD1>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD2=PROBEDATA("STANDARDDEVIATION","T2A0B0")
                IF_GOTO/STD2>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD3=PROBEDATA("STANDARDDEVIATION","T3A0B0")
                IF_GOTO/STD3>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD4=PROBEDATA("STANDARDDEVIATION","T4A0B0")
                IF_GOTO/STD4>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD5=PROBEDATA("STANDARDDEVIATION","T5A0B0")
                IF_GOTO/STD5>=0.0005,GOTO = AUTOCALIBRATE
                I also have a comment in the beginning of the program to have the operator take off the star probe and clean it with the gum stuff to clean the module and cmm head contacts.

                So if everything works out well, it goes to a timed operator comment saying the cal was successful and then the probe moves to the back of the machine ready for parts.

                Then you can put in a counter like CALCOUNT=CALCOUNT+1 so like if it automatically repeats more than once, (meaning you've calibrated twice and the results are still not good.), then human intervention is needed so then it will go to not timed operator comment say to seek help. Maybe a probe or 5way is loose or dirty or bent or something. I also have a backup star probe under the CMM so they can just replace it, calibrate the new star probe and continue running until I can asses the probe causing problems.

                1 time I was called in because the machine would not calibrate and a ruby had fell off one of the probes.

                Let me know if this helps.
                Last edited by acgarcia; 01-15-2020, 03:22 PM.

                Comment


                • acgarcia
                  acgarcia commented
                  Editing a comment
                  This is the only probe on this machine so I use YES_DCC which makes tip 1 A0B0 the master tip. It will repeat automatically without operator input if it fails.

                • acgarcia
                  acgarcia commented
                  Editing a comment
                  So I also print a report to our network for record keeping. policy is that the operator must calibrate before each shift. once the calibration is done, the probe moves to the back which is an indicator that the calibration had good results. Otherwise they'd walk up to the CMM and see the operator comment saying the cal was unsuccessful and to seek help.

              • #10
                Originally posted by acgarcia View Post

                Thoughts? Yes.

                All I do is use a variable for each probe/angle to check the standard deviation of the tip. the I use logic to see if the results are larger than or equal 0.0005" to repeat which usually fixes things. Here is one of my star probes. See below.

                Code:
                AUTOCALIBRATE=LABEL/
                AUTOCALIBRATE/PROBE, PARAMETER_SET=Star_Calibrate, QUALTOOL_MOVED=YES_DCC,
                SHOW_SUMMARY=YES, OVERWRITE_RESULTSFILE=NO
                BEGIN AUTOCALIBRATE RESULTS FOR PROBE VC0022 STAR USING SET Star_Calibrate
                Probe file=VC0022 STAR Date=10/3/2017 Time=6:02:09 PM
                Dixie Sphere CENT X 21.0250 Y 9.8345 Z -12.3495 D 0.7500
                T1A0B0 THEO X 0.0000 Y 0.0000 Z 8.2262 D 0.1969
                T1A0B0 MEAS X -0.0296 Y -0.0223 Z 8.1695 D 0.1948 StdDev 0.0002
                T2A0B0 THEO X -0.9252 Y 0.0000 Z 7.1927 D 0.0787
                T2A0B0 MEAS X -0.9491 Y -0.0197 Z 7.1290 D 0.0770 StdDev 0.0002
                T3A0B0 THEO X 0.0000 Y -0.9252 Z 7.1927 D 0.0787
                T3A0B0 MEAS X -0.0243 Y -0.9461 Z 7.1275 D 0.0765 StdDev 0.0003
                T4A0B0 THEO X 0.9252 Y 0.0000 Z 7.1927 D 0.0787
                T4A0B0 MEAS X 0.9035 Y -0.0166 Z 7.1416 D 0.0769 StdDev 0.0002
                T5A0B0 THEO X 0.0000 Y 0.9252 Z 7.1927 D 0.0787
                T5A0B0 MEAS X -0.0207 Y 0.9022 Z 7.1400 D 0.0769 StdDev 0.0002
                END AUTOCALIBRATE RESULTS FOR PROBE VC0022 STAR USING SET Star_Calibrate
                LOADPROBE/VC0022 STAR
                ASSIGN/STD1=PROBEDATA("STANDARDDEVIATION","T1A0B0")
                IF_GOTO/STD1>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD2=PROBEDATA("STANDARDDEVIATION","T2A0B0")
                IF_GOTO/STD2>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD3=PROBEDATA("STANDARDDEVIATION","T3A0B0")
                IF_GOTO/STD3>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD4=PROBEDATA("STANDARDDEVIATION","T4A0B0")
                IF_GOTO/STD4>=0.0005,GOTO = AUTOCALIBRATE
                ASSIGN/STD5=PROBEDATA("STANDARDDEVIATION","T5A0B0")
                IF_GOTO/STD5>=0.0005,GOTO = AUTOCALIBRATE
                I also have a comment in the beginning of the program to have the operator take off the star probe and clean it with the gum stuff to clean the module and cmm head contacts.

                So if everything works out well, it goes to a timed operator comment saying the cal was successful and then the probe moves to the back of the machine ready for parts.

                Then you can put in a counter like CALCOUNT=CALCOUNT+1 so like if it automatically repeats more than once, (meaning you've calibrated twice and the results are still not good.), then human intervention is needed so then it will go to not timed operator comment say to seek help. Maybe a probe or 5way is loose or dirty or bent or something. I also have a backup star probe under the CMM so they can just replace it, calibrate the new star probe and continue running until I can asses the probe causing problems.

                1 time I was called in because the machine would not calibrate and a ruby had fell off one of the probes.

                Let me know if this helps.
                Perfect. That is EXACTLY what I was looking for. I was playing with it a bit, and I was in the right neighborhood. Thanks again.

                Comment


                • acgarcia
                  acgarcia commented
                  Editing a comment
                  Glad I could help.

              Related Topics

              Collapse

              Working...
              X