No announcement yet.

Temp control

  • Filter
  • Time
  • Show
Clear All
new posts

  • Temp control

    I am using a Do-More (H2 series) PLC and looking to program a simple temperature control algorithm and not sure if i should use PID temperature control or simpler logic. My inputs are read in through RTD sensors via the F2-04RTD module. The output is to turn on a glycol pump and open a motor controlled valve... depending upon which of 8 different tanks requires the cooling load. I'm told that the algorithm should use a 3 degree variance as such:

    if the set point is 68 F, then open MC valve and turn on pump when the temp input is 71 F or greater and keep the outputs on until the input temperature has reached 65 F.

    The three degree variance reduces the cyclic loading on the pump and valves.

    Would PID temperature control be different? and where can i find some examples of the control logic for the program?


  • #2
    Unless your pump and valve are variable I'm not sure that PID would give you anything more than could be accomplished by tightening the On/Off levels, which will just increase the number of cycles. If a 3 degree variance is good enough, simple is good.

    If the valve is variable, you could tighten the temp control up considerable with PID, but the pump would have to run pretty much continuously.

    Standard disclaimer here: I make PLC products for a living, but I don't don't actually use my application thoughts must be viewed skeptically.


    • #3
      The logic I use for heat trace cable on piping is just simple greater than less than compares with a debounce coil to control chatter close to the setpoints. If you're always driving the temperature one direction and you process temp doesn't need to be exact, this will probably work fine.


      • #4
        Liquid tanks often have enough thermal mass relative to the controlled variable that On-Off control (with debounce) is perfectly adequate.

        I do prefer to do the debounce with time rather than a temperature differential, because then it automatically optimizes for ratio of load/ouput capacity while still maintaining whatever you define as acceptable levels of chatter in your control device. Further, I do the debounce as a minimum on-time, so that you may in fact be able to change state immediately as soon as desired, so long as you're not causing unacceptable chatter.

        Finally, if you process allows, you might leave the last open valve open and just turn off the pump when you no longer need cooling. Don't close the valve unless some other tank wants cooling while the original one is still satisfied. That way, if the original tank IS the next one to call for cooling, the valve is already open, and you'll have saved two changes of state for the valve.


        • #5
          Did you ever receive a sample program for this application??
          I have a similar application.