S4C+ M2000 IRB2400 not in zero position after Rev. Counter Update

    I have a problem with updating the rev counter of all axes. I set the robot arm to the calibration position and then clicked Service>View>Calibration>Calib>Rev. Counter Update. Unfortunately after this operation all axes counters are not in zero position(biggest deviation on 1st axis ~3degree). Anyone know what steps to follow to have a well calibrated arm?

    PS:The robot has a new system, previously it was mounted on the ceiling, now it is mounted on the floor. Offsets from the label on the arm are entered into the system.

    Thanks and best regards,

    Radosław

    If the values are different, then the markers line up differently at zero. Fine calibration is usually only used when a motor is replaced. Maybe one has been replaced. It seems to me that there are no programs yet which would be ruined by changing things now. You will need to make sure that gravity Beta parameter is set correctly for floor mount, rather than inverted.

    Quote from Enio

    Hello Lemster, in this case, can be done a fine calibration in order to bring all axes in 0° . It will change the offset values of course. what does it change if the offsets are not the same as those written in the cabinet door

    Original factory calibration numbers are the result of a mechanical/electrical setup far more precise than lining up notches. That precision gives better TCP, Linear & Circle moves. As Lemster68 says, any mechanical changes done on an axis will invalidate the factory cal offset number.

    Quote from Lemster68

    Is there also a sticker inside the cabinet with calibration offsets? If so, do they match? Double check that you did enter the numbers correctly.

    No, i don't see sticker inside cabinet, only on arm. The values on the sticker are the same as in the controller memory . I used original calib,cfg file during preparation of new system.

    Photo of sticker and data from controller to each axes: sticker & teach pendant

    Quote from Lemster68

    Next, ensure that the robot type in the new system matches the robot,

    Do you know how to check that? I prepare system by own maybe i did something wrong.

    Photos of system preparation: System preparation

    Quote from Lemster68

    Did you make a moveAbsJ with all zeros in the jointtarget and tell it to go there?

    I did calibration routine based on this video couple months ago

    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.

    but after that robot is not into calibration marks. Tomorrow i will check it again.

    Here i put main page of teach pendant and gravity beta Teach pendant

    Today i updated rev. counters and i have some "progress"

    axis1: 0.0deg

    axis2: -0.3deg

    axis3: -0.6deg

    axis4: 0.0deg

    axis5: -1.0deg

    axis6: -1.8deg

    What should I do? Make some offsets during program preparation or call ABB service to prepare fine calibration for robot arm? Any ideas?


    Thanks!

    Radosław

    Edited once, last by crazyferajna (November 9, 2023 at 5:02 PM).

    It looks like the sticker says robot type is 2400F/10, and the system you made says only 2400. Not sure what difference is but I would go back to check if the other type is available to create the system. Should indicate Foundry option.

    The degrees that you are listing I assume to be what you see when you line the markers up "perfectly".

    It really does not matter if it is only a degree or two different. It only does matter if it is more than one half? or one full motor revolution or more. I had one that was about 3.5 degrees off the marker and when I did the rev counter update, it had the desired effect. Smaller values than that usually have no effect.

    Quote from Lemster68

    It looks like the sticker says robot type is 2400F/10, and the system you made says only 2400. Not sure what difference is but I would go back to check if the other type is available to create the system. Should indicate Foundry option.

    take a look on attachment, i used 2400/10 M2000 as a floor mount. And in my opinion, this is the most suitable option

    Quote from Lemster68

    The degrees that you are listing I assume to be what you see when you line the markers up "perfectly".

    I agree, listing when the position is more or less perfect - I tried my best :smiling_face:

    Quote from Lemster68

    It really does not matter if it is only a degree or two different. It only does matter if it is more than one half? or one full motor revolution or more. I had one that was about 3.5 degrees off the marker and when I did the rev counter update, it had the desired effect. Smaller values than that usually have no effect.

    Thanks for the tip! Tomorrow i will try prepare "calibration procedure" once again a will back with results.

    After you reset rev counters, you need to run the 'CALIB' routine you wrote to according to video. The axis position value for all 6 axes should now be 0.0. If they are not at 0.0, then the p10 position somehow did not have all axes set to be at 0.0. Follow the video again and make sure to edit p10 for all 6 axes. If you are not sure, make a backup and post a zipped copy of the backup. Someone here can help find the issue.

    Quote from Lemster68

    If you make another system with the hanging selected, you will have to correct the gravity beta again, and, I just remembered the base frame orient. Just select florr.

    I used option 2400/10 M2000 (i think default is floor)

    Quote from Skooter

    After you reset rev counters, you need to run the 'CALIB' routine you wrote to according to video. The axis position value for all 6 axes should now be 0.0. If they are not at 0.0, then the p10 position somehow did not have all axes set to be at 0.0. Follow the video again and make sure to edit p10 for all 6 axes. If you are not sure, make a backup and post a zipped copy of the backup. Someone here can help find the issue.

    I repeated the "CALIB" routine again and now it WORKS! :smiling_face: All arm joints are in 0.0 in controller and also in position marks! I don't know why that was not working couple months ago.

    One last question in this topis: I need to replace the battery inside the controller(18NiCd batteries), because after 3 days since I last turned it on, I get the error like "Battery discharged, lost rev. counter data." Should it be replaced when controller is power on.

    Should it fit the S4C+ controller?

    https://www.batterien-und-akkus.com/akkupack-nimh-…fuer-21-6-v-abb ?

    Quote from crazyferajna

    One last question in this topic: I need to replace the battery inside the controller(18NiCd batteries), because after 3 days since I last turned it on, I get the error like "Battery discharged, lost rev. counter data." Should it be replaced when controller is power on.

    Should it fit the S4C+ controller?

    https://www.batterien-und-akkus.com/akkupack-nimh-…fuer-21-6-v-abb ?

    You can replace with the power on if you are very careful.

    If you want to do it with power off, make sure to do manual shutdown found in the Service menu under the 'File' dropdown. After selecting 'Shutdown', wait for the controller to finish imaging and the LEDs on the CPU go out (about 8 - 18 seconds) before turning off the disconnect. This will ensure you have a known good image when you power the controller back on.

    The battery pack in the link looks correct but I've always used the 3HAC5105-1 pack made with NiCads, not NiMH. Not sure how the charging would be affected.
    The NiCad packs require a 24hr 'ON' period to be assumed fully charged.

    Quote from Skooter

    You can replace with the power on if you are very careful.

    If you want to do it with power off, make sure to do manual shutdown found in the Service menu under the 'File' dropdown. After selecting 'Shutdown', wait for the controller to finish imaging and the LEDs on the CPU go out (about 8 - 18 seconds) before turning off the disconnect. This will ensure you have a known good image when you power the controller back on.

    The battery pack in the link looks correct but I've always used the 3HAC5105-1 pack made with NiCads, not NiMH. Not sure how the charging would be affected.
    The NiCad packs require a 24hr 'ON' period to be assumed fully charged.

    Okey, I'll try to find a NiCad battery instead of a Ni MH battery.

    For this topic all i clear!

    In addition, I have one problem with the connection to the FTP on controller, but I will prepare a new topic.

    Thank You Skooter & Lemster68 for your help!

    Quote from Lemster68

    If the values are different, then the markers line up differently at zero. Fine calibration is usually only used when a motor is replaced. Maybe one has been replaced. It seems to me that there are no programs yet which would be ruined by changing things now. You will need to make sure that gravity Beta parameter is set correctly for floor mount, rather than inverted.

    Yes!! Cannot stress this one enough!

Advertising from our partners