All,
Some background:
- Third-party positioner integrated as an external axis system on an ABB robot cell (welding application)
- Two mechanical units defined in the controller: STN1 (RobMechUnitType_MultiAxes, 2 axes; 7 and 9 driving tilt, Axis 8 driving rotation) and Slave_Motor (1 axis - Axis 9), which is electronically linked to Axis 7 to jointly drive the positioner’s tilt motion
- Axis 7 = ELM master, Axis 9 = ELM slave
- Confirmed working: jogging STN1 from the FlexPendant/RobotStudio moves Axis 7, and Axis 9 correctly follows and tilts the positioner as expected. RAPID-programmed moves also work correctly. So the ELM link itself (system parameters / MOC config) is functioning properly under normal motion.
When attempting to tune Axis 7 in TuneMaster, the tuning routine appears to command Axis 7 alone. Axis 9 does not follow, and since the two are mechanically/rigidly coupled through the tilt mechanism, this produces an immediate torque error and aborts the test.
Both mechanical units (STN1 and Slave_Motor) remain activated throughout and deactivation isn’t an option in this application, so it’s not an activation issue.
Does TuneMaster’s excitation/test-signal injection bypass the normal motion-planner path that ELM tracking relies on? Is there a supported way to keep the slave axis synchronized during a TuneMaster test if so?
IRC5
RobotWare version: 6.16.02.00
TuneMaster version: 25.1