[will707]

Hello!

I am attempting to update the modified DH parameters from a calibrated robot to our R30iB plus FANUC controller. I followed the directions found here https://robodk.com/doc/en/Robots-Fanuc-U...eters.html. I updated the joint values for "a" and "d" in the respective system variables as shown below:

   

   

I was then noticing some very odd motion in some of the linear moves. Looking back at the instructions, it seems like I may be missing some offset and/or negative signs. It looks like there may be some pattern I am missing? The values in the example exported table don't quite align with what is on the teach pendant images. I've included screenshots of the images that I am referencing from the instructions below:

   

   

Looking for advice on how to proceed, thanks!

[Albert]

Once you have calibrated a robot with RoboDK and you use the calibrated robot with a program, RoboDK automatically filters robot programs when you generate them. This means RoboDK will alter the robot programs to correct robot errors. This means you should not modify your robot controller and this is also how you'll obtain the best accuracy results.

On the other hand, if you still prefer to export the robot kinematics to your robot controller (in case you are not planning to use RoboDK for simulation and offline programming for example), I recommend you to export only the Geometry simplified subset of parameters. These are mostly the link lengths that are different from zero.

[will707]

I am trying to debug some joint trajectory differences seen along a linear move between two joint targets in RoboDK and ROBOGUIDE (ROBOGUIDE has shown to be fairly representative of our physical cell). I was thinking the difference in DH parameters could be contributing to this discrepancy as it affects kinematics. This specific program plotted below (joints values plotted over the course of a single linear move) show J5 achieving a value much closer to zero in ROBOGUIDE (and online) than in RoboDK, resulting in an uncaught singularity during offline programming. For this specific move we have verified the joint targets are the same at the start/finish of the program for both RoboDK and ROBOGUIDE, and the tool frame pose relative to the robot flange is the same for RoboDK and ROBOGUIDE. If you don't recommend updating DH parameters when using a calibrated robot in RoboDK, could there be another source of this difference? Thanks!