J6 rolling/twisting while making linear moves

    We have had multiple m-20ia robots with this issue. While in JGFRM or USER coordinate system and moving in the linear X, Y, or Z directions the wrist (J6) twists or rolls. This makes teaching linear moves extremely difficult.


    One robot started this after getting a new wiring harness. Remastering fixed this one. Another started acting up after replacing the J6 harmonic drive. Remastering worked here as well. A week later this same robot needed a new J6 motor and another new harmonic drive. It started doing the "roll" again. Remastering did not work this time. The 3rd robot started this with no prompting. Again remastering did not fix the issue.


    I have checked the frames on all 3 robots and all are still at 0,0,0. The core programs also appear to be correct on all 3 (I have compared the data from these robots to a couple others without the problem in the plant).


    This is driving me nuts! Please help!

    From the screen that you posted the picture of, press the previous key, that will bring you to a list of all of your jog frames. Then press F3 for other and select tool frame:



    Then use the arrow keys to go down to tool frame 10 and then press F2 for Detail:


    A picture of all ten tool frames before you press F2 for detail would also be nice, but is not necessary.

    Uggh. No Frames at all. Thats ugly. The fact that you don't have a UTOOL defined would make jogging all but the simplest of tools a nightmare. What is this robot doing, and what does it's end effector look like?


    First of all, why bother using the user or jog frame coordinate system if you don't have any frames?


    Second, what do you mean by rolling? When jogging X, Y, or Z in a Cartesian coordinate system (any other mode than joint), the robot will maintain the orientation of the tool center point, which in this case is the middle of the faceplate. This means that the wrist axes, J4, J5, and J6, will all move to maintain the orientation. This is normal.


    Could you share a picture of your overall workcell and maybe a short video of your issue? A few pictures of your tool along the path that you are jogging might help instead of a video if you don't have an extra pair of hands.

    Attached is the video of the robot moving in JGFRM coordinate, just moving +/- X. The gripper starts with the opening facing up. As it moves along X to the right, the gripper opening begins to face right. As it moves along X to the left, the gripper opening rolls back to vertical and begins to face left. It's almost as if the hand rotation that should keep the gripper at vertical is backwards.


    I checked a few of the other robots that have NOT had this problem and all of the tool frames look exactly the same as this one.

    The other robots have the same gripper, but on those, the gripper stays vertical while making +/- X movements.


    I haven't been able to find anything different between the ones that roll and the ones that don't.. But I'm still relatively new to the programming side of robots so I'm sure I must be missing SOMETHING! lol

    Myself and 2 other engineers have each triple checked the mastering on these. That doesn't appear to be the issue...

    Could it be that the gripper was installed at the wrong orientation? Between the hand and the J6 on the robot there is a wedge block that is set at a different position on all of the robots, but I don't see that having this affect. especially when the robots that work normally are all different as well.

    Is there a link or something you can share with me about how to correctly master? I'm sure there are a few ways based on personal preference, but I'm willing to try all the ways if it will solve this problem!

    If you move it by program (L PR[x], L PR[x], offset PR[y]) does it move in a straight line? If it does, it ain't mechanical/electrical/mastering related. Do you have wrist jog enabled when in teach? It shows up as W/TOOL, W/USER, W/JGFRM on the pendant.

Advertising from our partners