long calibrationan examples & questions...

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

  • long calibrationan examples & questions...

    hi frends and famili
    lets say:


    Monday: I auto calibrate all tip 1, 2, 3, 4 & 5 with a tp200 and rack in all angles. All calibrations start with A0B0, and then calibrates the angles all sorted by default . (Tip 1 A0B0 being master tip)
    Tuesday: The ruby tip crashes during a routine at 2.5mm / 0.1'' measure speed with tip 2 in A90B180.
    I re calibrate master tip, then tip 2 A0B0, A90B180.


    Will all other angles in tip 2, 3, 4, & 5 still relate to master tip & tip 2 with A0B0, then A90B180?
    Is it even necessary to re calibrate tip 2 A0B0 & not just master tip > A90B180 or will it break the chain?



    Monday: I auto calibrate all tip 1, 2, 3, 4 & 5 in rack and all angles. All calibrations start with A0B0, and then calibrates the angles all sorted by deafult . (Tip 1 A0B0 being mastertip)
    Thuesday: The rubytip crashes during a routine at 2.5mm / 0.1'' measure speed with tip 1 in A90B180.
    I calibrate master tip & then tip 1 in A90B180

    Will all other angles after from master tip and tip 1 A90B180 relate?
    If master tip (tip 1 a0b0) crashes, can i just calibrate that one?
    Is it even necessary to re calibrate if it crashes in that meas speed, will it do that much damage to the tip?
    Is there a difference between crashing perpendicular to a surface to with ex. a90b180 against a surface pointing z+ in machine or a0b0 against a surface point z+ in machine?
    Does the damage to tip depend much of what kind of length/size it is?
    Should you start the calibration with a0b0 > the angles you use the most? Or does it mater if it go a0b0 > a25b0 > a55b0...etc..




    Do you guys re-calibrate or qualify/check your tip(s) if it crashes like that? (we usually don't)

  • #2
    when you say recalibrate master tip, you are reseting it to THEO first? you should never touch your master tip to the sphere without resetting to THEO and telling pcdmis the sphere has moved. As soon as THEO does not match MEAS for x,y,z on your master you have error start there and rule that out, that will have an effect on correlations

    when you say all 5 you mean the master included? there are no possible settings that are good for both the master and the other tips, if you do the master along with any other tip stop doing that. Set the sphere location first with only the master tip, then it is done only do the other 4 saying the sphere has not moved do not re-do the master on a sphere that has not moved... ever

    edit: if you do want to recal only the master without physically moving the sphere, lie and tell pcdmis you moved it. if your master tip bent then you need the sphere location to follow it to keep correlations, telling it the sphere did not move will give the master an offset instead and a level of error
    Last edited by Douglas; 08-20-2020, 01:44 PM.

    Comment


    • #3
      These are all really good questions pcdmisstudent I too am curious about a few questions you have. I know what do to to avoid potential errors caused by crashes a calibration steps being wrong though I would love to hear what others have to say.
      PcDmis 2015.1 SP10 CAD++
      Global 7-10-7 DC800S

      Comment


      • pcdmisstudent
        pcdmisstudent commented
        Editing a comment
        @Kp61dude!
        What would you do if a tip crashes during a routine? =)
        How do you avoid errors cause

    • #4
      Originally posted by pcdmisstudent View Post
      hi frends and famili
      lets say:


      Monday: I auto calibrate all tip 1, 2, 3, 4 & 5 with a tp200 and rack in all angles. All calibrations start with A0B0, and then calibrates the angles all sorted by default . (Tip 1 A0B0 being master tip)
      Tuesday: The ruby tip crashes during a routine at 2.5mm / 0.1'' measure speed with tip 2 in A90B180.
      I re calibrate master tip, then tip 2 A0B0, A90B180.


      Will all other angles in tip 2, 3, 4, & 5 still relate to master tip & tip 2 with A0B0, then A90B180?
      Is it even necessary to re calibrate tip 2 A0B0 & not just master tip > A90B180 or will it break the chain?



      Monday: I auto calibrate all tip 1, 2, 3, 4 & 5 in rack and all angles. All calibrations start with A0B0, and then calibrates the angles all sorted by deafult . (Tip 1 A0B0 being mastertip)
      Thuesday: The rubytip crashes during a routine at 2.5mm / 0.1'' measure speed with tip 1 in A90B180.
      I calibrate master tip & then tip 1 in A90B180

      Will all other angles after from master tip and tip 1 A90B180 relate?
      If master tip (tip 1 a0b0) crashes, can i just calibrate that one?
      Is it even necessary to re calibrate if it crashes in that meas speed, will it do that much damage to the tip?
      Is there a difference between crashing perpendicular to a surface to with ex. a90b180 against a surface pointing z+ in machine or a0b0 against a surface point z+ in machine?
      Does the damage to tip depend much of what kind of length/size it is?
      Should you start the calibration with a0b0 > the angles you use the most? Or does it mater if it go a0b0 > a25b0 > a55b0...etc..




      Do you guys re-calibrate or qualify/check your tip(s) if it crashes like that? (we usually don't)
      Obviously the best course of action would be to not crash the machine. I am still very new to this kind of work (just a little over a year). I personally do not recalibrate if I have a small crash UNLESS the tolerances I am measuring too are super tight. Previously I stated that I work with a lot of big weldments. Weldments very a lot from part to part so It's not uncommon to have a small crash from probing limits or something of that sort. The tolerances on these weldments are usually around ±.030 which is really big for some people. I would not personally recalibrate for something like this. However we also do machined and cast parts with tolerances at ±.05mm which is MUCH tighter than our weldments. In this case, yes, I would recalibrate if i was to crash.

      This is all just my personal and slight professional opinion. Let me know what you guys think. Great topic pcdmisstudent !

      Comment


      • pcdmisstudent
        pcdmisstudent commented
        Editing a comment
        ok =)
        we work with a tolerance width of 0.01mm (so +-05mm)
        How would you recalibrate if you were workin with tight tolerances?
        Master > tip that crashed or just master > angle that crashed ? Or whole calibration routine?

      • ThePudds
        ThePudds commented
        Editing a comment
        pcdmisstudent Me personally, I would pull the probe back, do whatever fixens needs done to the program or part, and run the program. Is your part passing? Is it close? Were engineers or superiors expecting it to pass? I would personally recal the angle on the probe that "crashed". Rerun the program, if you get the same result, you most likely have a bad part. It also depends on how critical the part itself is. Are you working with aerospace? Could the part you are making save someones life? Recal everything in that case!

    • #5
      There's a great tutorial video on this subject made by one of my badger friends https://www.automettech.com/master-probe-calibration
      CMMXYZ Applications Specialist: PC-DMIS Support and Training. Check out my PC-DMIS tutorial videos... https://blog.cmmxyz.com/blog/tag/cmm-tech-tips

      Comment


      • #6
        If you're using a star probe and you crash any tip at any angle, you should recalibrate the entire star probe. While it's possible for just the stylus that was crashed to be affected, it's just as likely for the module or the extension to suffer. If either are bent, then everything below it will be off.
        PC-DMIS 2016.0 SP8

        Jeff

        Comment


        • pcdmisstudent
          pcdmisstudent commented
          Editing a comment
          @Schrocknroll
          I have a star probe (last tip 5 in rack).
          Why just star probe, and not a straight styli"?

          And by below it will be off, you mean the order in calibration?

          So if we have master tip, and tip 2 a0b0, a90b0, a90b180, a90b90 and then a tip 3 with same angles as tip 2.
          (and it calibrates in that order) if we crash with tip 2 a90b0 all angles/tips after a90b0 must also be calibrated?
          So if master tip is tip1 in rack a0b0, and tip 1 a90b180 crashes, all other 5 tips in rack will be calibrated because the styli can get bent?

      • #7
        define 'crash'. a 'little crash', does that mean you touched something you weren't supposed to while a routine was running? If you have a 'little crash' that triggers the probe, but DOES NOT move the probe to the limit of the hardware you are using, that is not (IMO) a crash, but a unexpected trigger. If your 'little crash' push the probe to the limit of the hardware, you have no way of knowing if it was really beyond the limit of the hardware and you could have bent the tip. If any hardware becomes disconnected (TP20 module unseats) that is a crash, again, did it bend the tip? Can't know without checking it, and just because you think you are currently working on something with enough tolerance to disregard a possible micon-bend in the probe, doesn't mean the next part isn't going to require it.

        sigpic
        Originally posted by AndersI
        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

        Comment


        • #8
          Originally posted by Matthew D. Hoedeman View Post
          Can't know without checking it, and just because you think you are currently working on something with enough tolerance to disregard a possible micon-bend in the probe, doesn't mean the next part isn't going to require it.
          Well, maybe that is how it works with you. In MY opinion (opinions are like butts) if you needed that micron-bend in the probe to pass the part with ±.03, it should be closer to nominal in my opinion. If you are using up that much tolerance, whose to say the next part WILL pass? Would be better to re-target if able and resubmit.

          Comment


          • Matthew D. Hoedeman
            Matthew D. Hoedeman commented
            Editing a comment
            not what I was saying. Just because the CURRENT part has enough tolerance that you can ignore a possible micro-bend in the probe doesn't mean the next part will also have enough tolerance to ignore a micro-bend in the probe.

            In other words, NEVER risk running if you MIGHT have mucked up your probing system in any way, be it a micro-bend or a fish hook.

          • ThePudds
            ThePudds commented
            Editing a comment
            Matthew D. Hoedeman thats pretty much what I was saying above. If a part has ±.05mm vs ±.03in, it would be a no brainer to re-cal. I have personally never had a major crash. I take a lot of time with my programs for now and make sure they are right before i let others run them. I don't consider a false trigger a crash. especially when i only probe at 2mm/sec.

          • pcdmisstudent
            pcdmisstudent commented
            Editing a comment
            So you guys always calibrate or qualify after you suspect( crash & have micron-bended tip)?

        • #9
          Here, i sometimes crash my Ø1by30mm probe, this tip experienced many crashes in his life (not at high speed), but we calibrate it every week and the stddev is still pretty the same. The thing is, how to see if the extension is slightly bent ? we are only checking the "ball form" when calibrating no ?

          Comment


          • #10
            Originally posted by pab39 View Post
            Here, i sometimes crash my Ø1by30mm probe, this tip experienced many crashes in his life (not at high speed), but we calibrate it every week and the stddev is still pretty the same. The thing is, how to see if the extension is slightly bent ? we are only checking the "ball form" when calibrating no ?
            You get data for the form of the tip, which shows that it is round, and PROBABLY has no cracks or chips (but not for certain, you might not touch where it is chipped!). But the calibration gives you the calculated center of the probe ball to the master tip (when using a rack) or the relationship of that probe angle to all the other probe angles of that probe file. Your probe can be bent, your extension can be bent, but once you calibrate it, the machine knows where it REALLY is and uses that data. As long as you don't shank out on anything, all the data is good. There have been times when I HAD to make a fish-hook probe to check a part. Calibration is what keeps the data good.
            sigpic
            Originally posted by AndersI
            I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

            Comment


            • #11
              Originally posted by Douglas View Post
              when you say recalibrate master tip, you are reseting it to THEO first? you should never touch your master tip to the sphere without resetting to THEO and telling pcdmis the sphere has moved. As soon as THEO does not match MEAS for x,y,z on your master you have error start there and rule that out, that will have an effect on correlations

              when you say all 5 you mean the master included? there are no possible settings that are good for both the master and the other tips, if you do the master along with any other tip stop doing that. Set the sphere location first with only the master tip, then it is done only do the other 4 saying the sphere has not moved do not re-do the master on a sphere that has not moved... ever

              edit: if you do want to recal only the master without physically moving the sphere, lie and tell pcdmis you moved it. if your master tip bent then you need the sphere location to follow it to keep correlations, telling it the sphere did not move will give the master an offset instead and a level of error

              Yes by recalibrate i reset it with theo (parameter set tip 1 in a0b0 with reset to theo, yes has moves, so its always XYZ-perfect.
              and then "no hasnt moved" with no reset to theo for all other angles/tips. And tip1a0b0 is in tip 1 in the rack.
              Everytime we calibrate or recalibrate a tip we always start with the master tip (autocal with parameter set).
              Then we remove the sphere when its complete. We dont have it mounted to a fixture or something.

              So calibrate, remove sphere, "crash happens", put back sphere in approx same place, locate the tip 1 a0b0 with a manual point and "no hasnt moved" for the other tips.
              Why shouldnt we start with master tip? Its the one that starts the chain right, so everytime we calibrate we have to use t1 a0b0, and since the sphere has moved we take manual point, we creates a correlation.

              Why shouldnt we recalibrate the master ? Its the one that starts the chain?

              Comment


              • Douglas
                Douglas commented
                Editing a comment
                recal on the master tip saying the ball has not moved... what that will do is any slight measurement error will show up in the form of an offset on your master probe. over time these small errors could compound, next time you say the sphere moved if you were to not reset to theo it would use that error offset to locate the sphere. If you do reset to theo when saying the sphere moved you mitigate that

            • #12
              Originally posted by Matthew D. Hoedeman View Post
              define 'crash'. a 'little crash', does that mean you touched something you weren't supposed to while a routine was running? If you have a 'little crash' that triggers the probe, but DOES NOT move the probe to the limit of the hardware you are using, that is not (IMO) a crash, but a unexpected trigger. If your 'little crash' push the probe to the limit of the hardware, you have no way of knowing if it was really beyond the limit of the hardware and you could have bent the tip. If any hardware becomes disconnected (TP20 module unseats) that is a crash, again, did it bend the tip? Can't know without checking it, and just because you think you are currently working on something with enough tolerance to disregard a possible micon-bend in the probe, doesn't mean the next part isn't going to require it.
              What do you mean by push the probe to limit of the hardware?
              By crash i ment ur description: it hits something it shouldnt do (maybe part was badly/fixtured and error message probe deflected show up.
              So since you consider that an unexcpected trigger, i assume you dont recalibrate/verify it?

              I guess what im after is how the calibration chain works: We locate sphere with man point mastertip, and "no hasnt moved to all else tips", thus creating relation, from it (i.e. all angles/tips after that first "yes has moved"-tip will change its xyz value in according to calibrating master, right?

              lets say we crash tip 2 a0b0 on stem in high speed (so it gets slightly bent), we have to calibrate master with man pnt > tip 2 a0b0 > all other tips/angles to relate. Or can we just calibrate master > tip 2 a0b0. (because angle a90b9, tip 3, tip 4 etc..relates to tip 2 a0b0 (which relates to previously angle at tip 1?)

              Comment


              • #13
                thanks for discssion guys.
                so what im curious is about the chain/relation between the tips. Does all the angles relate to the previous one calibrated or just the master?
                Examples:

                so we have tip 1, tip 2, 3, 4 and tip 5 (only having one angle, a0b0 for simplicity)
                We locate the sphere with man point with tip 1, and dcc with "no sphere hasnt moved" for the rest tips.
                if we crash tip 2. Do we just have to calibrate tip 1 and 2 (and since we calibrate the first in the chain, all the other tips will follow the chain).


                And what if we for example 10 angles for each tip.
                If we crash tip 2, with angle 8th (i mean the 8th angle it calibrates after a0b0, lets say a90b0)
                Do we have to calibrate just master tip & tip 2 a90b0 (Cause other will follow)
                Or do we have to recalibrate master and tip 2, angle 1-8.
                OR do we have to recalibrate the master > all other angles that have been calibrated previosuly up until tip 2 a90b0 (all other 9 angles for tip 1, and then angle 1-8th for tip 2)
                Last edited by pcdmisstudent; 08-21-2020, 09:29 PM.

                Comment


                • Douglas
                  Douglas commented
                  Editing a comment
                  If it was enough to be concerned about any one tip then I would recal all tips on that same star, odds are pretty good if it was hit hard enough to change anything there is at least potential the rest moved, or the whole star spun.

                  I use a separate master tip, standard force module with 1 4x10mm tip and it does nothing but locate my sphere and always within .00002 STD dev, never touches a part. That way all I ever have to do is re-locate my sphere and recal the affected probe, the master correlation to other probe sets is unaffected.

                  When your master is on the star that 'crashed' then the only safe answer is double check for physical damage, re-locate your sphere and recal your whole rack. A physical change in the master breaks correlation to any tip calibrated prior

                • Douglas
                  Douglas commented
                  Editing a comment
                  also yes for all ten angles, if there was any physical change in a tip in one angle it will be there in all wrist angles

                • pcdmisstudent
                  pcdmisstudent commented
                  Editing a comment
                  Ok thanks for reply: unfortunately our "master probe" is problably the most used probe in our rack. (But we rarely use A0B0), might be wise to get a seperate master tip.
                  "also yes for all ten angles, if there was any physical change in a tip in one angle it will be there in all wrist angles"
                  This is correct, but if we only need that A90B0-angle for tip 2, we could just calibrate mastertip > tip 2 a90b0, (but it wouldnt relate to the other angles for tip2, because all the angles for that tip would be affected if the styli got bent):

                  Or are you saying that if a phsyical change in one tip in one angle would happen, it would affect ALL wrist angles for ALL tips?

              • #14
                Originally posted by pcdmisstudent View Post
                What do you mean by push the probe to limit of the hardware?
                By crash i ment ur description: it hits something it shouldnt do (maybe part was badly/fixtured and error message probe deflected show up.
                So since you consider that an unexcpected trigger, i assume you dont recalibrate/verify it?

                I guess what im after is how the calibration chain works: We locate sphere with man point mastertip, and "no hasnt moved to all else tips", thus creating relation, from it (i.e. all angles/tips after that first "yes has moved"-tip will change its xyz value in according to calibrating master, right?

                lets say we crash tip 2 a0b0 on stem in high speed (so it gets slightly bent), we have to calibrate master with man pnt > tip 2 a0b0 > all other tips/angles to relate. Or can we just calibrate master > tip 2 a0b0. (because angle a90b9, tip 3, tip 4 etc..relates to tip 2 a0b0 (which relates to previously angle at tip 1?)
                hardware limit is just that, the limit of the hardware. Look at a TP20. Push the probe, it moves. Keep pushing it until it won't push anymore. That's the hardware limit. If your 'crash' didn't push the probe to that limit, that isn't (IMO) a crash, but an unexpected probe contact. For that, nope, I wouldn't recalibrate.
                sigpic
                Originally posted by AndersI
                I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                Comment

                Related Topics

                Collapse

                Working...
                X