Strange KR5 malfunction

    Hello,

    I am experiencing a very strange issue with a robot I have. The robot is a KR5 with a KRC4 controller running KSS 8.5.7.

    When I command movement on the 4th axis, it rotates but then starts making a whistling sound. After rotating for a certain period, the motor stops with a "Gear torque" error. Along with the sound, there is a high-frequency vibration. Even when the motor is not rotating, this vibration persists as soon as power is applied.

    At first, I suspected a mechanical issue, but when I removed the wrist, I realized that the same situation occurs on axes 4, 5, and 6. However, the whistling sound is only present on the 4th axis motor, while the others only exhibit high-frequency vibration.

    All motor brakes function correctly, and the measured cable values appear normal.

    The KSP module for these axes works fine in other robots.

    I checked the connectors, motor cables, and encoder cables one by one but not found abnormality.

    Motors have been tested and are in good condition.

    The RDW is fine.

    The wrist is fine, yet the motors vibrate when idle.


    Despite all values appearing normal, I suspect internal cables, but honestly, I'm stuck at this point. The probability of all three motor cables failing at the same time doesn’t seem make sense.

    Does anyone have any ideas?

    people make mistakes...

    are you sure that is the correct robot? and are you sure that MADA is also correct?when wrong MADA is used things could be worse (wrong loop gains, inertia etc.)

    in fact i was just commissioning cell with new robot that was delivered to us with incorrect configuration. actual robot arm was KR16R1610-2 but in WorkVisual projects (base and working project) someone entered KR16R2010-2. under Project Info serial numbers of robot and cabinet are correct but the arm model was also wrong. that is a 400mm difference. Focus center sticker on the main cabinet was showing correct arm model. Presence of that sticker suggests that system was inspected by KUKA in North America before being delivered, but again wrong MADA was not caught.

    robot was still moving, mastering check was fine but of course linear motions were not linear, and measuring TCP using XYZ 4-point method produced huge errors. after correct robot was configured in WoV project, deployed, etc. everything was back to normal.

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

    Hello panic mode


    Thanks for clarification.

    The robot is not newly installed; it has been operating on the production line for about three years. So, there is no software issue. What makes the situation interesting is that three motors are vibrating at a similar rate, and the issue is not caused by the KSP or RDW

    I would normally suspect the A4-A6 servo amp, because that's the main place all three of those axes "converge". But it sounds like you already tried swapping that out.

    A4-A6 are usually mechanically interlinked as part of the wrist assembly, as well. So it's possible that, if one of the motors were having an issue, it might cause a vibration on all 3. A4 would be my prime suspect, because it's the one at the start of that kinematic chain. Just spitballing, but if the A4 motor is starting to fail (say, the junction between the drive shaft and the resolver is coming a bit loose), it might cause the motor to "hunt" at high frequency, because it thinks it's moving when it's not. That should generate errors, but if the hunting amplitude was small enough to stay below the error detection threshold, it could cause something like what you describe. A5&A6 might simply be counter-rotating against the side effects of A4's motion.

    On a larger robot, I would start swapping the motors around, but I'm not sure that's so easy on a KR5.

    Have you tried taking a Trace of A4-A6?

    SkyeFire, thank you so much for reasoning through the issue and trying to help. The problem has become a bit more complex. I've already disengaged all the motors, and they have no mechanical connections between them. I tested the motors on another robot, and they run smoothly. I'm also testing the other components separately. Right now, the only thing left is the internal cable bundle, which I will remove tomorrow, but I can't evaluate this within a logical framework.

    The situation has evolved into a rather strange point. The KPP and KSP have been replaced in the control panel. The RDW has also been replaced. However, the whistling sound and vibration still persist.

    What's odd is that I'm only experiencing issues with the 4th and 5th axes. When I replace the 4th and 5th axis motors with larger motors, all problems disappear. Specifically, when I replace the original 1FK7032 motors with the 4th and 5th axis motors from a KR210, the vibrations stop. The motors rotate normally.

    This could lead to the conclusion that the original motors are faulty, but those same motors work perfectly fine on other axes. I also replaced the cables between the control panel and the robot, but the issue remains unresolved for some reason. Of course I measure normal values on PE and phase lines.

    Logically, there are no other components left to replace. The ballast resistors show normal values. The MADA is in its original state, and I have tried several different software versions. However, there is still no solution.

    Quote from panic mode

    Since it does not squeal when larger motors are used, it must be wrong MADA or wrong motor data. Delete robot from wov project then add It again from correct catalog


    Hi panic mode and SkyeFire:


    I know that the motor data is incorrect. However, I only installed a different motor for the cable test, and interestingly these motors are working while the original ones are vibrating. We tested with motors of the original code but the result did not change they are also vibrating.

    The robot has a history of continuously operating in mass production, performing simple transportation tasks. I did the installation myself 3 years ago and until now there have been no issues. The customer reported that there was noise coming from the motor and that they encountered high current faults on the A4 axis. Based on this we intervened but due to the limited space we had to remove it and it is currently in our workshop.

    There is no untested component left in the cabinet that could be related, except for the Mains Filter which we haven't tested. However, logically I don't think this could be the cause.

    Quote from M.Ozkan

    There is no untested component left in the cabinet that could be related, except for the Mains Filter which we haven't tested. However, logically I don't think this could be the cause.

    ...wow. This one's a stumper. The "obvious" answer is to just replace the arm entirely, but that's a brute force/deep pockets solution.

    My gut feeling is that is has to be something mechanical. Electrical failures are usually 0% or 100%, no middles. And you've already eliminated most of the potential electrical causes. It can't be environmental, since you've relocated the robot entirely. Errors that slowly creep in over time in-use are almost always mechanical.

    Process of elimination suggests it almost has to be the motors, or at least one of them. But the troubleshooting data certainly isn't providing a clear "smoking gun."

    Yes, we mostly use similar logic to troubleshoot failures, and I wouldn’t consider myself inexperienced in such situations. However, the actions taken so far are not leading to a solution.

    By the way, the motors are not attached to the mechanics; I am testing them separately. This is because when they are connected to the mechanics, the exact situation cannot be fully understood.

    The strange thing is that there are no issues with the large motors at all. This makes me think that the difference in dynamics and response between the small and large motors might be the cause. However, I would still expect a sensitive component like a servo motor to respond similarly with vibrations.

    I ran the motor idle for about an hour, and it didn’t even heat up at all.

Advertising from our partners