Performance of the controllers during load changes

I'm interested in buying Roboteq controller, either GBL26xx or RGHL18xx but I haven't found any performance test documentation about them. By performance I mean how well the controller can stay in its reference speed when load changes. The controller is used to drive the permanent magnet motor attached to the hydraulic pump. The dynamic characteristics of the motor are adequate for my needs and I would like to know before buying if the controller is sufficient enough for it. So does there exist any performance test for the controller with changing load?

The second question: is it possible to add some sort of load compenstation to the control logic via digital inputs from pressure measurements across the pump? For instance with feedforward loop to current control.

Unfortunately, we have not completed any definite tests regarding our controller's response due to the large variance in applications that are determined on a case by case basis. However, our control loop operates at 1KHz. Therefore, the maximum bandwidth we recommend for applications across the board is 200Hz so that there are no errors due to sampling.

The best way to implement a feed forward control on our controllers is though a MicroBasic script.

Thanks for the answer! I went through datasheets and the parameter "Motor Acceleration/Deceleration range" came up with the minimum value of 100 ms (Time in ms for power/current to go from 0 to 100%). I didn't find any parameter regarding rate of current change. The motor that I'm using has ~6 ms electric time constant with 180 Arms rated current. So that 100 ms sound highly limiting factor with respect to a rapid load change performance. The only affecting parameter that I found was "Set Acceleration" which have maximum value of 500000 which equals to 1000 rpm acceleration in 20 ms which is adequate for my hydraulic application.

So where does that 100 ms minimum for rate of current change comes from, is it a fixed internal safety measure or can it be bypassed somehow?

There might be a risk of damage to the controller risk if you try to deccelerate faster than our default ramp down speed. But I will have to consult with one of our electronics engineers.