Hey there,
I ran into a problem in regards to the force_mode at my work. I have a UR3 (Polyscope 3.5) and a UR5 (Polyscope 3.9) and the issue happens for both of them. When I start the force_mode with 0 N applied, the robots are starting to move in the positive direction of the compliant axis. I was expecting there wouldn't be any movement with 0 N. When I apply a force of -100 N, for example, he first moves forwards, then stops and moves back, as expected for -100 N. Does somebody know why this is happening?

Best regards
Sebastian

Hello @Sebastian

Did you set the payload properly in the TCP setting? 
I tested it here and it's work fine as long as there is no force applied on the tool. However, once you apply a force and release it, the behavior is not what I expected as it starts moving in the opposite direction.

The force mode of the CB series isn't precise enough to handle such low forces. 
I would recommend adding a Robotiq FT 300 to get a better force control using Force Copilot software

Best regards

Hello David,
thank you for your answer. Yes I did set the payload in advance. I am using an FT300 on both of them. I configured the FT300 in the Force Copilot mode. It gave me a Payload of 1.3 kg, which I wrote into the Payload field of the TCP Configuration, as well as the suggested center of gravity from the Force Copilot configuration. And I also checked the box for using Force torque sensor data in force control mode in the Force Copilot Configuration. After I restarted the system, the robot has the same behaviour as you explain. It does not move until I apply a force. In my case it always moves into the positive Z-direction, no matter from which direction my external force is applied on the TCP.
Is there a place inside the code from the UR5 I can look into, in order to find the problem? Do you have any ideas?

Best regards
Sebastian