Hello,
We are working with the KR6 Agilus 6 axis robot. When the robot first came in March this year we didnt master the robot and we worked with the factory mastering. In time we realized that the robot was not accurate enough, and also we couldnt calibrate a tool better than 1 mm of error (4 point method).
The Agilus robot is mastered with the MEMD device and axis 6 has scratched markes that need to be aligned. I have checked the mastering with the MEMD and axis 1-3 were ok(or I think these results are OK):
Axis position difference for A1 was -0.002deg (motor angle difference 0.2deg)
Axis position difference for A2 was 0.04deg (motor angle difference -4.7deg)
Axis position difference for A3 was 0.029deg (motor angle difference -2.99)
But A4 and A5 were way off:
Axis position difference for A4 was -0.8deg (motor angle difference -75.91)
Axis position difference for A5 was -0.155deg (motor angle difference -14.24)
So I have mastered the robot and thought that this would fix all our problems regarding the poor robot accuracy (TCP calibration errors, rotation around the TCP).
So as axis A4,A5 and A6 are mechanicaly coupled I needed to remaster all three in this order. I firstly remastered A1-A3 and then A4,A5 and finally A6. The errors of axis position difference after mastering for each axes were under 0.002deg. I have done the mastering without any load on the robot.
Afterwards I have tried to calibrate the robot TCP (a very light aluminum flange and some other mechanical elements were put on the robot flange - not heavier than 0.7 kg and with the centre of mass around 50mm in z axis). The TCP was around 250mm in Z direction, 50 in X and 100 in Y. I tried calibrating the TCP few times but I have got errors around 3-5mm each time! I tried calibrating some other TCP's but the errors were present again.
My conclusion is that my mastering introudced these errors as the robot worked more accurate before this mastering!
These are my questions and I would be very greatfull if someone could help:
1. Is it possible to restore the old mastering? This is most very important as we cannot work with the robot at this moment as it is very unacurate. (I tried restoring the old robot archive with the old mastering but no luck - I have got errors again when calibrating the TCP. I used the MEMD to check the mastering and all axes were in the range of 0.002deg)
2. How is it possible that the new mastering is less accurate than the factory one?
3.Is it possible that the robot has some mechanical failure inside the arm where motors for axis A4-A6 are?
Kind regards,
Marko