Post by thread:Digital PID

I've been on this topic this week. I don't have a good routine, but here are a bunch of good resources: The best tutorial so far: http://members.aol.com/JohnShaw/pid/index.html John Shaws basic tutorial on PID with a Basic (pun intended) program example of PD (not PID) Here is a good link for PID loops. http://rclsgi.eng.ohio-state.edu/matlab/PID/PID.html. > Controller Tuning and Control Loop Performance" by David W. St. Clair. > http://members.aol.com/pidcontrol/booklet.html > I'm designing an oven controller with a nasty, nonlinear thermistor for feedback. This week I'll be testing a routine that I've put together in C. Questions: 1. Are you using C or ASM 2. Whatr are you trying to control? Motors? Relays? 3/. What type of feedback do you have? Thermistors? Tachs? -----Original Message----- From: Tiziano Tommasin <spam_OUTTiziano.TommasinTakeThisOuTBRAHMA.IT> To: .....PICLISTKILLspam@spam@MITVMA.MIT.EDU <PICLISTKILLspamMITVMA.MIT.EDU> Date: Monday, April 19, 1999 5:39 AM Subject: Digital PID >I'm searching a digital PID. >Does anyone have any PID routine avaiable I can take a look at. >Tnk in advance. >Tiziano

<snip> Lawrence, My first PIC application was a PID temperature controller using just such a nasty nonlinear thermistor for highly accurate control of a laser. In fact we started a thread on filtering techniques (using your median filter routines) if you remember! The problem with nasty, nonlinear thermistors (TM) is that, by definition, their response is different at different temperatures, i.e the slope changes. This means the gain of the entire system changes with temperature which can spell big trouble with instability, at least, it did for me. You could get around this by either calulating your system gain (P term etc) at the point where the slope of the thermistor is highest and accepting the lower gain at other temperatures OR you can linearise the thermistor. The other thing to watch with thermistors is that the ADC circuit dosen't put too much current through them which can lead to self heating effects. You can do this via analogue means but I went for digital linearisation with an interpolated lookup table. Anyway, the thing worked pretty darn well. Regards Mike Rigby-Jones {Quote hidden}> I'm designing an oven controller with a nasty, nonlinear thermistor for > feedback. This week I'll be testing a routine that I've put together in > C. > Questions: > > 1. Are you using C or ASM > > 2. Whatr are you trying to control? Motors? Relays? > > 3/. What type of feedback do you have? Thermistors? Tachs? > > > > > > > > -----Original Message----- > From: Tiziano Tommasin <.....Tiziano.TommasinKILLspam.....BRAHMA.IT> > To: EraseMEPICLISTspam_OUTTakeThisOuTMITVMA.MIT.EDU <PICLISTspam_OUTMITVMA.MIT.EDU> > Date: Monday, April 19, 1999 5:39 AM > Subject: Digital PID > > > >I'm searching a digital PID. > >Does anyone have any PID routine avaiable I can take a look at. > >Tnk in advance. > >Tiziano