Display MoreOkay... that's bad. As Panic said, unlike "classic" KUKAbots, there are no user-serviceable parts inside the LWR (or iiWA). In fact, what I was told (back when the LWR was new) was that just opening the shell was a good way to ruin the robot, b/c the innards are packed so tightly and require such extreme precision. The phrasing was something like "the moment you take out the last screw, the entire insides kind of explode into your face".
Still, something seems off about this. Does the Mastering position agree with the marks on the axes, particularly A2? After you Master the robot, does it perform Linear moves correctly? IF A2 is mastering 90deg off, then LIN motions should be very non-linear. Also, the Cartesian position monitor should also be way, way off.
If the robot moves correctly, despite A2 showing 90deg at Mastering position, then I might be completely wrong and this LWR is okay, just unusual.
If not... I hesitate to suggest this, but it might be worth trying to set MAMES[2] to either +90 or -90, and try Mastering again, and see if that helps. As long as you're careful, that shouldn't break the robot any worse than it already is.
Just watch out for Gravity Compensation Mode. If the LWR is torque-mastered incorrectly, or if the gravity vector setting is incorrect, enabling the drives will cause the robot to start moving, potentially in very unexpected ways. So test in clear airspace, and be ready to stop the robot quickly.
Thank you for your comment!
The mastering position does agree with the marks on all joints. What's weird is that the Cartesian position of the mounting flange at the Candle position seems to be right.
This doesn't make sense to me with MAMES[2] = 0 if A2 is mastered 90 degrees off.