Hi Johannes,
How are you?
I have a question related to the robot's wrist.
As you probably know, A5 and A6 rotate undesirably when you move A4.
The encoder position of A5 and A6 remains unchanged but the real position is different because of the belt and gear transmissions. The same phenomenon exists for A6 if you only make A5 move.
Do you perhaps know the angular rotation of A5 and A6 when A4 is rotated 360° and angular rotation of A6 when A5 is rotated 360°?
The robot is a KR200/1A and the wrist of the robot is ZH200-2 (0319)
Thanks in advance!!
Best regards,
Dimko
Hello Dimko,
While I'm aware that they can do 3 axes with 2 motors, I have never seen A5 and A6 move when you move A4, so that sounds like a hardware problem.
Unfortunately I don't have any information on the gearing / relation between those axes. Would you like to counter the rotation via software?
Best,
Johannes
Hello Johannes,
Thank you for your reply!
I did some testing today and i come up with these numbers.
When you ONLY rotate axis A4 with 360° then axis A5 wil move 2° and axis A6 wil move 4°
When you ONLY rotate axis A5 with 360° then axis A6 wil move 4°
Here is a youtube video for who's interested.
https://www.youtube.com/watch?v=iRKDfknqtbc
So yes i need to counter the rotation via software.
In the attachment i uploaded a foto of the flange of the robot.
There where some numbers on it for axis A4 A5 and A6, do you now what they stand for?
Big thanks!!
Dimko
Hello Dimko,
Personally I feel that is something that would be better to fix in hardware, than offset in software.
But I made a quick prototype how that could work in KUKA|prc. Of course there are several issues, e.g. that you are now using PTP interpolation instead of LIN, so you need pretty dense toolpaths. The multiplication is of course just a placeholder for your factor.
Maybe you could also work with the planes from Analysis instead.
But it is a bit hacky and can cause issues, so be careful.
Unfortunately, I haven't seen these numbers anywhere yet.
Best,
Johannes