11-04-2019, 12:00 PM
Hi,
I have discovered a potential problem with UR TCP point and its RoboDK.
The problem is that although reference frame is (0,0,0) both in robot and RDK the TCP points have different coordinates and angle rotations. The problem is that this difference is not linear: for different points and angle rotations, the total difference between coordinates and angle is different. For instance, TCP can be (100,100,100) on the robot controller and (110, 105,100.5) in RDK. On the same time, robot joints are same both in robot controller and RDK. Probably, the problem is that inverse kinematics is computed the wrong way? I've tried to change something in settings but it didn't work. Also, tool calibration is not helping to solve the problem.
I am using RDK 4.0 on Windows and Ubuntu and have this problem with two URs. So, this problem is independent of platform and/or robot unit.
I have discovered a potential problem with UR TCP point and its RoboDK.
The problem is that although reference frame is (0,0,0) both in robot and RDK the TCP points have different coordinates and angle rotations. The problem is that this difference is not linear: for different points and angle rotations, the total difference between coordinates and angle is different. For instance, TCP can be (100,100,100) on the robot controller and (110, 105,100.5) in RDK. On the same time, robot joints are same both in robot controller and RDK. Probably, the problem is that inverse kinematics is computed the wrong way? I've tried to change something in settings but it didn't work. Also, tool calibration is not helping to solve the problem.
I am using RDK 4.0 on Windows and Ubuntu and have this problem with two URs. So, this problem is independent of platform and/or robot unit.
