hexagon logo

Im lost with this one. I need help.

Hello, so I have been using pc-dmis for about a year now. I know how to align a part so I know its not that. This is how my probe looks after A1 manual first alignment. I tried to reset product as well and I still get this problem. Starting fresh, I load a model, and the xyz are all flat meaning if I use vector points to interrogate it, the x, y z never move if I were to click on other locations, so I know I don't have to flatten the model, because as we all know sometimes its origins are not square with the model, but this is not the case.

Here is the problem though, when I do 3 2 1 level rotate origin alignment and then after A1 this is how the probe looks on the screen, before alignment A1 if I were to click on CTRL+F for auto feature vector point the probe looks as it should look, straight down, but after a1 code if I press CTRL+F and put a vector point, this is how the screen looks. After A1 if I click on x y z surfaces with vector points, they are all 0, meaning my alignment is how I want it, but again this is how my probe looks, and im unable to actually align the part. What is this? I tried rebooting, powering the machine on and off and whatever I do, after A1 my probe goes like this on the screen / as you can see from the screenshot.

After A1 I would turn on dcc mode and adjust my clear plane using f10, but I cant even get passed A1 which is odd because usually the machine knows where the part is, but if I do auto feature after dcc it will show on the screen going with this angle in the screenshot below and it ends up missing the part. Guy who is running the part says its not the part and that he checked with indicator. Im using flat surfaces for the A1 manual alignment, but again after I do ctrl+alt+a for alignment utilities and do my 321 and press ok, and press ctrl+f after A1 code, my probe goes from straight down to how it looks in the picture below.

Any help would be greatly appreciated.

Attached Files
  • Your on the right track. Using a new model on an old rev program can be a smart move to save a ton of time! I do it all the time

    However if you want a mostly seamless transition then you need to import a new model with its origin point (trihedron) exactly as it is for your startup alignment. If you have access to cad software it makes the very easy, spend a few minutes adjusting things before you import and replace and you should be good to go.

    You can even do this in PC-DMIS if you don't have cad software, start a new program, import the new model, perform all your translations and rotations, then save the program. Open your old program and import the .cad file from the program you just created and replace, then you should be off to the races!
  • Hello again, thank for all your feedbacks! So I was programming the other dash for this part and the same thing happened, but when I created a vector point after physically aligning it went to the right place, however everything else that I had programmed new (without first physically aligning), it did not know where it was, it only knew after physically aligning. Is this because I'm trying to program under machine mode and not offline? I have been programming it remotely when someone is physically there so we align first and then I program as we go, but lately I have been trying to program it under machine mode without first physically aligning it, I would create alignment 1, but not actually align, and just continue programming it while not running anything even though its under machine mode. Could this be why? This was a fresh program no copy paste and the same thing happened.

    For example: Its a new program, new model, I create alignment(without physically aligning first)and then continue to add vector points, when I run the program it does not know where the vectors are. It only knows if I physically align first and then add new vector points and circles, etc, it knows where it is, but if I try to do it without aligning physically it wont know where it is. However, if I physically align first, and then create vector points and program as I go, it knows where it is.
  • Are you using a Read Point so that the program knows where the origin is when you start?
    Are you taking manual hits off the part and using that data for your program?
    You are going off the theoreticals from 0,0,0. Read Point tells it where your origin is. That needs to be translated into machine axis data somehow if you aren't programming under an alignment using machine axis data.
  • Is that ctrl+alt+g? I don't think I have ever used that when aligning the part. I have been doing it basically the same way I have done it with the previous 30 or 40 programs I have made and never ran into this unless something was lost?

    I have been doing most of my alignment like this, its how they taught in the class I went to.

    Load model, check to see if xyz are flat to the model, and then throw 3 vector points then create a plane, two vector points to create a line and then a single vector point for datum c for example. I would then ctrl+alt+a, level plane 1 to z, rotate line about z to y+plus in this case and then origin plane to z, origin line to x and then origin the pnt for y(for this program), double check to see if xyz is zero and then to continue programming. A1 would be my manual alignment, so im taking it the rest of the vector points dont update if they were created after alignment 1 without first physically aligning it first, right? and that would be the reason why I would need to use read out in this case?
  • Code please. Based on what I saw, you could use readpoint to speed up manual alignment.
  • Next time it happens Ill copy the code here. Finished the dash 2 part is good.
  • If you are programming online you need to physically take manual hits off the part and create some sort of quick alignment for the machine to know where it is in space in some capacity. I'm not entirely sure what had happened with your program, but I'm glad you got it figured out.
  • It's strange, I don't know why I would think it was possible to create a program from start to finish in online mode without ever starting it and then run it for the first time once finished by pressing CTRL+Q and then it having me physically take the hits that made up A1 once CTRL+Q was pressed. Guess this is not the case right? Because once I was finished with the program I pressed CTRL+Q to start it for the first time, it would then ask me to take hits on the model where I put my points that made up A1, but right after this is where the problem starts, its as if my vectors for everything I had programmed were wrong, however if I create a new vector point after aligning it knows where it is under the same program, or is this working as intended? Like I said if it happens again ill paste the code here, thanks again everyone. Slight smile