Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Problem when defining reference frame
#1
Hello,
I got some problems when I try to define a reference frame with the 3 point (p3 crosses y +) method.
I am using a Kuka KR 6 and the kuka driver.
Everytime i try to do the process, my workpiece is not in the exact position and off for about 200 millimeters.

I have some questions about it:

Should i choose joints or points? What is the difference?

I calibrated the TCP first, do I have to enter these values to the Kuka? Or will I leave the tool and the frame to zero?

I also made sure that my frame is a parent of the robot frame, is there something else I should care about?
#2
It is better to use joint values to define your coordinate systems and calibrate the tools because they do not depend on the active reference or tool frame.

If you use points or poses to define your coordinate system or calibrate your TCP you should make sure your tool (TCP) and coordinate systems exactly match (in RoboDK and in your robot controller).

Did you properly select your robot? This could be the reason why you see these strange errors.

Can you share your RoboDK project file and/or the data you used to define your reference frame? We can help you better.
#3
Thank you for your reply,
So i teached my TCP once again and transfered it to the Kuka. I could move the angles A B C and the TCP stayed in place, so everything should be fine.

After this i did a new definition of the reference frame. This time i did it with points and the frage was of for about 63 mm.

In the project the robot is at the last position of the frame definition. So this should be on the Y-axis.

When I go to the p1 of the calibration, the robot is in the root of the Frame 3. The point i used on the real part was not in the center.


Attached Files Thumbnail(s)
   

.rdk   FrameError.rdk (Size: 1,004.75 KB / Downloads: 15)
#4
We realized the image you see is incorrect as it does not properly describe the points you should take. The current calibration method you are using requires you to take the first point at the origin.
   

We just updated RoboDK to correct this and properly reflect the calibration method you selected.

Alternatively, if you used the points depicted in the image you shared, you can select the second calibration method (point 1 doesn't need to be measured at the origin of the reference frame).
  




Users browsing this thread:
1 Guest(s)