7th Axis mounted robot square

    Hello


    I am working on a R2000iB 150U robot with R30iA controller, the robot is hung from a rail upside down, the rail is axis 7 in group one of the motion groups. My issue is how can I correct for the robot not being mounted perfectly square to the 7th axis rail. When I move the robot in world frame y almost as far as it can reach both ways and make a mark on the floor at both points, I then make sure the TCP lines up with the first mark, and jog the 7th axis to the second point(instead of with the robots y in world frame) it is out by about 8mm. My assumption is this is caused by the robot thinking its perfectly square with the linear motion of the 7th axis, but I think it is out slightly. I have a attached a picture to help show what I mean, I have greatly exaggerated the issue to make it clear.

    any help is appreciated. thanks

    I believe the only way to fix this is to single master J1 slightly off to match the y axis to the rail. You would then need to scribe a new witness mark.


    After doing this you would need to reteach any points. Also be aware that your axis limit will be slightly off.

    Following HawkME's advice for J1, I would check the mastering on the other axes too. Since it's an R-30iA, there could have been a poor/rushed remastering in the past.


    Kudos on the great drawing!

    HawkME is correct. You need to remaster J1 because for a 7th axis to work correctly it must line up exactly with the X or Y axis.


    PS. make sure the robot knows it is upside down. if you didn't do it in initial setup, it is in $PARAM_GRP.$MOUNT_ANG. if you don't the motor torques are reversely affected by gravity. I've seen this cause the robot to stop altogether.

    This is for an R30iB, so I am not sure it applies to a R30iA.


    During your axis 7 setup, you can enter a custom angle into the traverse direction.


    Figure out what that angle is, and enter it. J1 custom mastering would work too, but feels a bit hacky.

    Check out the Fanuc position converter I wrote here! Now open source!

    Check out my example Fanuc Ethernet/IP Explicit Messaging program here!

    That guys for the replies. Very helpful.


    I find it very interesting that adjusting the J1 master will affect this. The robots mastering was incorrect, so I vision mastered it, which apparently doesn't do anything to axis 1 and axis 6. So I don't feel bad about changing the J1 mastering. I am also not worried about messing up taught positions because I am still setting it up, and not much has been programmed yet. We did get it used.


    I believe this specific robot 150U is specifically designed to be hung upside down, there is no balancer on it. But I will double check the parameter.


    again thanks for help.

    Just saw your post now Nation, I was wondering if there was an adjustable angle like that. That's exactly the way to do it. It makes sense because what if you mounted it at lets say 45 degrees off on purpose. I will check if I have that setting. Thank you very much.

    If you don't need to, do not "destroy" mastering data - you'll loose the linearity of the robot 6 axes, which in not good. Then making linear motion in the robot program (L), it'll not move straight, but making a little arc - depends how much mastering will be missed.

    Just checked the axis 7 configurations and there is no setting for angle, only direction :face_with_rolling_eyes:. So I guess adjusting J1 axis master it is.



    Neighbour1, I believe if I only adjust axis 1 master it will not effect the linearity of the robot, in my head it will only turn the whole frame of the robot which is exactly what I want to do. I believe that is why vision mastering does not effect axis 1 and axis 6.


    thanks

    Quote from jariuscs

    Neighbour1, I believe if I only adjust axis 1 master it will not effect the linearity of the robot, in my head it will only turn the whole frame of the robot which is exactly what I want to do. I believe that is why vision mastering does not effect axis 1 and axis 6.

    Partly yes, you're right with the vision mastering, but I wouldn't be so sure about the linearity. Also it always depends how much you'd change the mastering, how many degrees - so "a little" change makes only "a little" linearity problem.

    But anyway, in my opinion, linearity will be affected by changing mastering data on J1.

    Changing J1 or J6 mastering does not affect linearity at all. That is why vision mastering can't calculate it ... Because it has no effect.


    Kwakisaki is right, if you can square it up mechanically, that is the right thing to do. But I also understand it may not be practical in every situation.


    Also, the hacky J1 mastering solution comes directly from a Fanuc engineer.

    A hack from the vendor is still a hack. When the next time it has to be mastered comes around, if the stickers weren't moved, or the new mastering position wasn't marked or documented in some way, I feel sorry for the poor bastard that has to deal with it. Probably why Fanuc still uses stickers for mastering position.


    Kwakisaki's solution is really the correct one here.


    Modifying the J1 mastering effectively rotates the world frame about Z relative to the base casting. Don't forget to adjust your J1 soft limits.

    Check out the Fanuc position converter I wrote here! Now open source!

    Check out my example Fanuc Ethernet/IP Explicit Messaging program here!

    I like this thread, very useful information. Thanks to all who have contributed. The reason squaring it in my case doesn't make much sense is that the J1 stickers were gone when I got the robot, and the mastering was definitely done "by eye" therefore incorrect in my opinion. So basically I had set J1 mastering by making the robot move fairly squarely with the axis 7 rail. Maybe Ideally I would square it up and then remaster J1 the way that I did now by squaring up the world frame with axis 7. I adjusted the J1 mastering by less than 0.05degrees so I think everything will be fine. I could record that J1 axis is mastered simply by squaring the world frame to the rail, for the next guy who has to master it.


    thanks again. I am happy with the solution.

    Quote from Nation

    When the next time it has to be mastered comes around, if the stickers weren't moved, or the new mastering position wasn't marked or documented in some way, I feel sorry for the poor bastard that has to deal with it.

    Precisely.


    I would pen mark the robot mounting bolts and also document what you have done, so that 'the next poor bastard' called to troubleshoot/maintain it is aware that JT1 mechanical zero is not true mechanical zero.

    Especially if you ever sell on the robot arm.


    0.05 degrees and 8mm........????......just what distance lies between these 2 points of reference?

Advertising from our partners