I've started up a gantry machine with a dual Y axis and found that the master axis showed a 30% higher motor load displayed on the drive than the slave axis.  Almost as if the master was dragging more of the gantry load than the slave was.  The servo drives are set up in torque mode so all the gains are in the K flop and the gains are identical for both axes. 
This brought up the question of how the master - slave axis works.
Do they operate as a parallel axes set where they both receive the same motion commands from the processor or does the master receive a motion command and then the master passes it's resultant motions on to the slave?  In the second scenario I can see where there could be a phase lag between the master and the slave therefore creating a master that is pulling more of the load. If the latter is the case, can I set up a single "phantom" master to which I can slave two parallel axes to eliminate the potential phase lag?

AZ
 

Hi AZ,

There shouldn't be any lag between the trajectories of the Master and Slave Axes.  The trajectory of the Master is computed and the motion delta of that axis is passed on to the Slave before any motion is acted upon by either axis.

Where do you see the higher load?  During Acceleration?  Constant speeds?  Both directions?

A gantry is inherently an over constrained system that can have a lot of "binding" if there isn't any built in compliance to allow for mechanical imperfections.

You certainly could configure a "phantom" master with no output that both real axes could then treat as their master to test your theory.

Or you might swap the Master and Slave roles to test your hypothesis.

Regards

TK

 

I've started up a gantry machine with a dual Y axis and found that the master axis showed a 30% higher motor load displayed on the drive than the slave axis.  Almost as if the master was dragging more of the gantry load than the slave was.  The servo drives are set up in torque mode so all the gains are in the K flop and the gains are identical for both axes. 
This brought up the question of how the master - slave axis works.
Do they operate as a parallel axes set where they both receive the same motion commands from the processor or does the master receive a motion command and then the master passes it's resultant motions on to the slave?  In the second scenario I can see where there could be a phase lag between the master and the slave therefore creating a master that is pulling more of the load. If the latter is the case, can I set up a single "phantom" master to which I can slave two parallel axes to eliminate the potential phase lag?

AZ
 

Tom:

Thank you for the explanation. That being the case, it indicates that there may just be more friction in the Master axis or the motor may have a bit lower Kt even though they are the same model motors.  I will try switching the master and slave assignments to see if anything changes.
The load difference is seen at constant vel during both high and low speed jogs in both directions.

Best Regards,
AZ