A colleague of mine was trying to tighten couple of bolts on EOAT and in doing so I noticed Wrist 2 rotated few degrees. The robot was Powered ON during this process i.e. in Normal Mode. We then ran a cycle in auto with no faults. However, on the next couple of cycles the robot was off its waypoint on the last application point by 3 mm. I ran the robot to a known TCP check fixture, and it was also off by almost the same measurement.
Is it possible the joint encoder readings are off? I reboot the robot and ran the program with same offset issue as before.
Any recommendations on how I can fix this issue without reteaching the waypoints?
Are you sure wrist 2 rotated, and not whatever potential bracket the tool is mounted to? I’d wager your EOAT is in a different place, if your known TCP is off by 3mm. You could opt to just edit/reteach your TCP to pick up on this new 3 mm degree of rotation if that’s the case.
You can move the robot while it’s completely powered off via the backdrive feature and it still knows where it’s at when it boots up. So I think it unlikely that the encoders are wrong.
1 Like
It was Wrist 2 that rotated. It could be the tool was slightly off. For now, I have just retaught the points. Didn’t want to adjust TCP value since it affects other sub programs. Thanks Eric