[NMIS]

Hello,

I used 'teach targets on surface' function to do some path planning simulation and then exported the programme to UR5e for testing. The actual motion of UR5e was different from the RoboDK simulation (the end-effector moved a few more cm than the simulation). I am wondering if that is a issue which simply can be sorted out by setting the same TCP in UR5e and RoboDK tool (if so, is it enough to change TCP in RoboDK?) or an issue need some serious robot calibration?

Thank you very much!

[Sam]

Are you using the post-processor or the driver?

In any cases, the TCP in RoboDK should be defined and accurate.

[NMIS]

Are you using the post-processor or the driver?

In any cases, the TCP in RoboDK should be defined and accurate.

Hi Sam,

Thank you very much for your reply.

I thought that I used the post-processor. What I did was to create the urp file in RoboDK and copied and loaded to UR pendant to run.

I found that UR has the function to re-define the TCP. But I am not sure if it is helpful. I suppose there should be some way to compare datum in both sides and then synchronise from RoboDK side. But I am not sure how exactly to do so.

[Sam]

You can define the TCP from RoboDK directly by using our driver (or copying the values manually).
If you already defined the tool on the UR, you can also copy the defined tool into RoboDK.
Once RoboDK is the owner of the TCP, the post processor will set the tool accordingly.

Refer to our documentation for more information and videos on how to define a TCP:
//www.sinclairbody.com/doc/en/General.html#CalibrateTCP

[NMIS]

Hi Sam,

Thank you very much for you advice. That was really helpful. I got the solution for this problem.