No announcement yet.

Recalculating PID variables mid process.

  • Filter
  • Time
  • Show
Clear All
new posts

  • Recalculating PID variables mid process.

    As I understand it, the Autotune process works by determining the Time constant and Deadtime of a system (in addition to gain). I have a system where both the TC and DT fluctuate significantly with output, but in a very predictable fashion. I wanted to use the math function to calculate new values as output changes and write them into the PID instruction.
    But I don't know what math the closed loop autotune uses in order to come up with its values. When I tune, the result work well, as long as the output stays in a certain range, once forced outside of that range, it either responds too slowly, or oscillates.
    Is the domore autotune procedure description available? I thought it was a ZN method, but that (if I understand it correctly) should result in I values that are exactly 4x the D value. Domore tune result in I values that are typically 6 times the D value, except occasionally when its 5 (I simply looked at various already tuned loops to get those).
    Thanks in advance

  • #2
    Do-more does use ZN.


    • #3
      So when I do a closed loop auto tune, where it performs 3 jumps from 0 to 100, it uses that to find delay time (L) and time constant (T), and then P=1.2(T/L), I=2L, D=0.5L? Or are there there other values?

      It looks like what I'm really trying to do is re-invent the dead time compensator. Is there a block or command for this?