Here's something I plan to try some time this year. It's a motor driver with a difference - it can simulate negative resistance!
Voltage driven DC motors already have inherent speed feedback: if the speed of a motor drops then the back EMF falls which leaves more of the supplied voltage developing across the motor's ohmic resistance.
This results in a higher current, higher torque and some degree of restoration of the set speed. You can prove this to yourself: get a low voltage toy motor and try to brake the output shaft with your fingers. It fights you!
Okay, now that we've established that principle, how can we improve it? Answer: make the winding reistance small so that the current rises sharply if the speed drops, and so applies more torque to restore the correct speed more agressively.
The attached circuit might permit this. The driver shown has the equivalent of the circuit below it, i.e. a voltage source with a negative resistance in series. This negative resistance can be set to almost cancel the motor's ohmic resistance, thereby making the motor look like it is wound with superconducting wire!
In theory this should make the motor change speed to order much more quickly than with a simple voltage driver (e.g. power opamp). That should translate to much improved damping when a phase control loop is wrapped around it for synchronisation.
I haven't done the PLL part of the circuit. I might not get around to this for another year so I thought I'd publish the principle in case others might want to try it.
Hint: I would not cancel all of a motor's winding resistance as thermal drift might lead to an unstable situation!!!