Hi all, a simple PI(D) control loop question. (I?m going to ignore D, for now) When doing PI loops I use the ?standard? equation. Out= Gain*(e + 1/Ti(int(e))) Where e is the error, Ti is the integral time, and int(e) is the integral of e for a time Ti. In analog land, for long times this means big resistors and capacitors. I then noticed that people do not always keep the gain in the P and I branches the same. So I can reduce Ti, and then attenuate the integral term by the same amount, and keep the response the same. (Say I reduce Ti by ten,by cutting down integrating resistor) Out= Gain*(e + 10/Ti(int(e))*1/10)
What effect does this choose have on the control loop? The only thing I see is a dynamic range change in the integrating term.
(I assume I can do the same sort of thing with the D term.)
Thanks, George H.