Re: PID tuning

[Fred Proctor <proctor-at-cme.nist.gov>]

EMC PID tuners,

Ray Henry wrote:

> What amps and kinds of signals did you use.  Did you use
> the same amps for both halves of the test?  If so, how did you defeat the
> current loop or did you simulate the tach when you didn't use one?

Originally we put the amp in velocity mode and measured the velocity
feedforward gain by running at 1, 2, 3 volts for a given time and
computing FF1 as the average of the volts per elapsed distance per time.
The P gain was tuned by raising it until oscillation.

We then put the amp in current mode by rejumpering it and removing the
tach wires. We ran Kees' system ID programs using voltage steps, and
followed the procedures outlined in the aforementioned references to get
P, I, and D gains for a given design factor that depends on the voltage
and current output the amp is capable of.

--Fred