I have used the SetToolRot procedures succesfully in the past (S4C and IRC5), but recently experienced a problem at a customer site with a slightly modified implementation (non-Pickmaster). In the past, I remember that a mis-calculated configuration value would result in an error. For example, using a configuration of 2 with an angle of 10 degrees which should have configuration 0, 4, 8 etc. would give an error. It appears now that there is no error even if the configuration does not make sense given the angle.
Has there been a change in the robot in the way configuration is handled on an IRB360, or is my memory faulty? How does the robot respond if the configuration value is not in a quadrant adjacent to the actual target angle? I think I may have a sign error in my equation which causes the wrong configuration to be used when the angle is very near the boundary of a quadrant.
Conflon is definitely set. I can force the tool to make extra rotations by setting the configuration value to a high number. I confirmed my field observations with a simple test on our demo robot by commanding the robot to a simple robtarget and varying the configuration value manually through the teach pendant.
This is an interesting point and we have also experienced it.
It seems that there are some side effects in using SingArea /Wrist on Delta robots.
Although it does not really makes sense to have this instruction, it is needed if you are looking for a more than 90A? rotation between 2 tracking points as otherwise MoveL won’t let you do this rotation.
We have seen that there are some side effects in this case since RW5.12 (as i remember), some of them improved with 5.14.
I guess it is under evaluation at ABB as we have raised the question since then.