Hello,
I am working on a servo-drive, and my company tries to promote some characteristics of its product. One of them is the bandwidth of the control loops. The motors are synchronous motors (6 to 10 poles) arround 1kW.

Some previous measurements show a bandwidth higher than 1kHz for the speed loop with the best motor, without load. I try to show in which cases this "high-bandwidth" could be useful. I am currently thinking about a Cartesian robot with spindle axes, to show that the tracking error is relatively small.

Is it a good idea, do you have better ones ?

 

Such 1kHz bandwidth seems very high even without load. What's bandwidth of inner current loop?

With application you'll need to decrease bandwidth due to low stiffness between motor and load especially with robot arm usually. In any case high bandwidth of speed loop allows reaching shorter move and settle time for point-to-point movements also.

 

The bandwidth of current loop haven't been measured but is in theory around 3 kHz.

For a spindle axis, I am able to tune the controller in order that the complet system has a bandwidth of 30 Hz.

The short settling time is the only advantage ?

 

Bandwidth of current loop can be measured at stall rotor location. You may use current monitor existing for each amp for this goal.

Position accuracy, repeatability and "in position" stability (position jitter) except scanning position error and move & settle time can be improved with higher bandwidth.

 

Ok, thank you for your help. I will do some tests in this direction.
I did not understand your protocol to measure the current bandwidth.

 

The measuring of any loop bandwidth is well-known procedure. I hope that you use one of standard servo amp. Than you need apply harmonic signal with various frequency without current limitation and test current sense signal.

You may ask me more with [email protected]

 

Ok, it is exactly what has been done for the speed loop. I believed that you had a special method for the current. Thank you for your fast response.