AncientBrit wrote:I've often wondered whether dynamic braking would improve the performance of the servo...
Likewise. When and if I ever get around to a mechanical monitor that's what I plan to do assuming the use of a conventional brushed DC motor.
Replace the power MOSFET with a small signal version (say a 2N7000), provide it with some form of resistive load to replicate the gain arrangement of the existing motor drive circuit then follow this with a power op-amp. A TDA2040 springs to mind, it's billed as an audio amplifier, but a power op-amp is all it is. Without referring to the datasheet, I'm not sure it's unity-gain stable though.
Failing that, a pair of complimentary trannies, say TIP29/30s or Darlingtons (TIP140/145) plus suitable bias.
A worthwhile experiment that those who have operational mechanical monitors would be to run it up and achieve sync-lock, disconnect one lead of the motor and time how long it takes to come to a complete stop. Repeat the exercise but this time disconnect the motor and short out its winding(s), note the difference (if any).
If there is any difference in the times then the above push-pull drive may be a valid improvement.
Or perhaps others have tried this before?
Steve A.
P.S. I've just checked the TDA2040 datasheet, it isn't stable at unity gain, but that could be got around...