This is typically as close as I can get when trying to my points after running the built in calibration tool in Roboguide. I did not use UFrames for this test program. But Ive done several of our robots now and this is usually as close as I can get. This makes for alot of touchups to the program in the real world after transfering from roboguide. I would like to do less touchups if possible. Any suggestions on how to get an even more accurate calibration?
WeldPRo (Roboguide) to real world Calibration
-
ORDEP81 -
February 1, 2019 at 5:09 PM -
Thread is Resolved
-
-
Anyone?
-
I have never done any robotic welding and don't know if user frames are typically used for welding. I would guess they are... When setting up material handling cells, I make user frames and touch them up in the real world and this works just fine to get the positions right. Is this something you can do?
I use Roboguide extensively but have never used the calibration tool. If I want Roboguide to match the real cell, I serialize the robot form a real world backup and move fixtures around to match the program. It's pretty rare I do this though. Hope this helps...
-
I have never done any robotic welding and don't know if user frames are typically used for welding. I would guess they are... When setting up material handling cells, I make user frames and touch them up in the real world and this works just fine to get the positions right. Is this something you can do?I use Roboguide extensively but have never used the calibration tool. If I want Roboguide to match the real cell, I serialize the robot form a real world backup and move fixtures around to match the program. It's pretty rare I do this though. Hope this helps...
Yes, we use user frames. I typically use the calibration tool exclusively for no other reason than just what was told to me to use in an email from Fanuc support. I also serialize robot from backup the model up the cradles fixtures, get them placed close and then run the calibration tool, but as in the picture shows that as close as I usually get.
-
Is the tool offset in one direction (X,Y,Z) or off in all directions? Are the tool center point and mastering accurate?
-
To limit touchups you need to get the real world and OPL matching. 1st thing we did was Perform enhanced mastering of the robot, then set up a new TCP. Once this is complete start picking points around the table and bring them into the OPL. Move the OPL table or positioner to match along with the corrected TCP numbers. When performing the above I use a 5/8" Teach tip instead of the wire. Also when you perform a new TCP use the 6 point method it is more accurate in the end. Now when I send a program from Octopuz to my robot I am within a 1.5mm or better.
All of above starts WITH GOOD MASTERING!!! it has taken me 3 months to get this dialed in, and now all that's left is dialing weld schedules in.
-
Do you mean that we shouldn't take for granted the factory mastering?
-
Do you mean that we shouldn't take for granted the factory mastering?
FActory mastering is BS. you're lucky to be within a pencil eraser diameter if checking TCP. The perfect world would be to perform the enhanced Mastering method then keep reteaching the TCP until you are where you want to be and that is 0 walk from the set pointer you are using on the table. -
Is this story finished as happyend story?