PID Sampling Interval
There something I don't understand with PID control systems and sampling intervals.
Everybody say the sampling interval should be as small as possible.
But I would say that: "If the sampling is too fast the derivative (d-term) as no influence".
Why do I say this:
Let take a given servo motor system:
The motor turns max 6000 rpm.
The tacho is a 4096 pulse per rev.
Let the sampling time be 2MHz.
This will give you steps per sampling at full speed: 0.2 steps/sampling
Every 5 sampling you get 1 step.
Most of the time you will get nothing out of the derivative calculation.
But the integral (I-term) will wind-up all the time due to the fast sampling time and computation interval.
Or have I misunderstood someting here?
Please clarify it for me.