Designing a PID Controller


Thread Starter

Leon Chetty


I'm a third year Electrical Engineering student from the University of Durban-Westville
Department of Electrical Engineering. I require
assistance to design and construct an Analogue Electronic PID Controller with adjustable parameters and optional switch with hysteresis. The actuator is to be a relay. I would appreciate any information that you may possess to assist me in the design and implementation of such a controller. The PID controller is a
temperature controller.

Thank you in advance


Pravin Fatnani

PID control would not be possible with relay as the actuator element. It would be only ON/OFF type control. To implement PID control, the heating element should be continuously (analog output) controlled which would not be possible using a relay as the actuator Hope , I have got your point correctly. In the simplest case, you could design a PID controller using few OPAMPs. Use variable resistors to control the coefficients and hysteresis.

CAT/ Indore/ India

Anthony Kerstens

PID loop control is possible with a relay in the form of time proportioning control. That is, the PID loop controls the on and off timing of the relay. The PID loop output could simply go to a 555 timing circuit.

To do this with a heating element, I would recommend using a solid state relay instead of a mechanical contact relay because of contact pitting and other damage due to the inductive load being switched frequently.

Also, instead of opamps, which could get a little messy for the separate P, I parts of the loop (I wouldn't bother with D for heating control) , I would use one of the cheap programmable experiment boards. Yes, it's more expensive
than opamps, but what's the cost of time and pain playing with opamps compaired to something on the order of US$100.

Better yet, I recall seeing a special integrated temperature control chip in a catalogue I no longer have. You should be able to pick-up a catalogue from an electronics mail-order
house and find something like that (if you're interested in that sort of short-cut :).

Anthony Kerstens P.Eng.


I would have to disagree on this. It is possible to do PID control with a relay output. Honeywell UDC3000, among *many* others, have this option.
The relay basically operates your heater element in a psuedo-PWM mode. The output comes on as the temp begins to drift down, as the temp is rising
again, the controller shuts the output off so that your temp ramp "coasts" up to your preset. As the controller learns the responsiveness of your loop, the hunting decreases until the loop is able to maintain a fairly tight control on the temp. (Less than 1 degree F drift at 325 deg. F)

I used several of these (both Honeywell and other brands) to control steam and electric heaters, gas fired sterilization chambers, etc. Worked very well...

I would suggest the original author look at some of the stand alone loop controllers (Honeywell or Love controllers, for example) to get ideas as to
operation. Google searching would probably be helpful.

--Joe Jansen

Johan Bengtsson

If you want any such shortcut yo could perhaps use a microchip PIC circuit. They have models that can handle analog inputs and PWM (pulse widht modultation) output directly.

But perhaps such solutions would be out of scope for the task you got.

Well however, you will of course need a pot for the setpoint. Take the setpoint - actual value
Multiply that by the gain (another pot) Take that value + the integrated value + the derivated value
(two more pots for integration time and derivation time) The sum will then need to be feed to the PWM-circuit giving you pulses for the relay. (another pot? for adjusting the period)

To make it good you should put in low pass filters (first order is probably enough) for the actual value and preferably also for the derivative part.

/Johan Bengtsson

P&L, Innovation in training
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]