Cal Check +Calibration in single program.

  • Filter
  • Time
  • Show
Clear All
new posts

  • Cal Check +Calibration in single program.

    I recently succeeded in combining my Cal Check program with my main Calibration program by using marked sets. The only problem is it requires 2 points at the beginning, one manual point used for the Cal Check alignment after full calibration, then the point to start the full calibration. Once the Full Cal is done, it remembers the Manual point and goes on it's way without need for operator interaction. I'm ok with it for now, but I'd like it to just have one point so it isn't as confusing for the operator. I tried searching the forums and I haven't found anything like this. Some things I have tried...
    First I tried using the qualtooldata("xyz") function, assigning a variable to it, assigning the same variable to the xyz of the Manual point and marking out the relevant assignments depending on which marked set is chosen. Once the Cal Check started, it looked like it went to the probe but when it read the sphere code, it drifted until alarming out of the axis. It looked like it automatically changed the target of the sphere to the same distance as the variables.
    I tried inputting qualtool variables into a generic feature, aligning to it, saving an external alignment, recalling startup, calibrate, manual point, align to it, save it as an external alignment with the same name as the generic feature so it overwrites it when needed, then adding if/goto statements to bypass certain external alignment saves depending on what marked set was chosen. Got the same result.
    I tried a few other things along these lines and got the same result each time. I feel this should be easier than I'm making it seem. Should i be inputting the variable into the theos of the manual point, aligning to it? Or should I be doing it for the initial pickup of the sphere?
    I thought about removing the master probe from the Cal Check section completely and just using the auto calibrate of the master probe, pulling the qualtooldata, and inputting it into the theos of every probe but I'm not sure how that would affect calibration results with other probes. I keep reading that it shouldn't as long as you always say Yes to the sphere moving on the master, which the program does. Or should I use the qualtooldata variable and input it into a sphere the master will pick up for the cal check, align to it, then go to the other proves?
    I've been racking by brain all week. Any and all help is very much appreciated.

  • #2
    I have a calibration program, but mine is probably different than yours.
    My "master probe" is a camera.

    What you're trying to do is verify your calibration before you run it, and, if it fails verification, check it?

    That seems fairly straightforward, if you use Parameter Sets.

    Create a DCC+DCC Qualification Parameter Set to determine the deviation; if the tool hasn't moved, the operator doesn't need to take a hit.
    If this Qualification fails, then perform a DCC+DCC Calibration Parameter set, where YES the tool has moved. This will be the one hit the operator takes.
    Everything else would be DCC+DCC, no the tool hasn't moved.
    You shouldn't need any special variables...


    • Douglas
      Douglas commented
      Editing a comment
      I have trouble with parameter sets on version 2103mr1, must be something I do but at times I will go to use one and all parameter sets are gone, ever seen that?

    • SingularitY
      SingularitY commented
      Editing a comment
      I used 2013 MR1 before moving to 2017 R2 and I personally never saw any parameter sets disappearing. I believe, correct me if i'm wrong, that parameter sets are tied to the probe files. Does anyone have access to the probe files who can accidentally delete them? (But your angles would disappear.) Do you run PC-Dmis online as administrator? Did IT not give the computer modify permissions for the user account PC-Dmis is ran from?

  • #3
    It really does seem straight forward, but everything I've been doing fails. Let me explain a little more..
    The company I work for has 2 programs. An auto-calibrate program and an in-house written program that mimics a qual check. Manual point on sphere, align, pick up the sphere twice and origin both times, then check the sphere with all of the probes. They use this as a way to check that the calibration did it's job or to do checks per shift to see if a calibration is required and also so the info exports to stats so we can keep an eye on things.
    I want to makes things easier so they don't have to open 2 programs, enter info twice, and touch off the probe twice. So I copied the cal-check program to the end of the auto-calibrate routine and I use marked sets to either bypass the calibration and only cal-check or do a full calibration and automatically cal-check. The program is, even if I use qualtooldata as a variable and plug it I to the first sphere probing dcc, it will automatically move the target values of the second sphere pickup by exactly the distance of the sphere from machine home and alarms out. I can plug in the qualtool data for all sphere pickups and bypass all alignments and it works, but I really don't want to go that route. (Nominals won't be zero anymore and will most likely confuse the CMM Operator.)
    So now I have it to where they take one hit on the sphere, which is the programmed manual point which aligns to this later after the calibration code, then it immediately autocals the master after they take a second hit to start the program. This works very well but I was just hoping I could reduce it to just one point...
    Still much much less time consuming than the previous method.


    Related Topics