I was trying to explain to a new person to controls what a PID does. First thing would be gain... But the DeltaV PID doesn't act like I am expecting.
I set up the PID block with the out going back to the input, having a scan of 1 sec, gain of 0.5 with 99999 reset. Made a 20% step change, got an initial 10% out change, then did some jumping around and finally landed 3% shy of where I though it should be.
In fact there were some small integration going back wards from what the action of the PID controller is set for before it settled down.
I don't see any other options in the controller that would do this.
Can someone explain to me what's going on?
The trend hopefully helps with what I am describing.
Thanks in advance
Gary
The ringing is because you have a control loop with no lag, and the proportional action is the result of Error. As the PV moved, the Error decreased and Out was subsequently decreased, which decreased PV etc and thus the oscillation. The 20% SP change caused a 10% change in out, and then PV. The PV change reduced error from 20 to 10, so the valve backed off by 5%, and PV also went to 25. Now error is 15, so valve output goes to 7.5 5 from where it started, and as you see, the PV and valve settled at around 27 %. For self regulating loops, which is what you created by looping back the output, integral action is required to get the PV to SP. I think the PV would reach SP only when SP is set back to 20. Everywhere else there will be offset. If you want to show how the GAIN affects out, don't feedback the process so the Error remains constant. When you do the SP step change, the OUT will respond and remain there since the error will be fixed.
Andre Dicaire
In reply to Andre Dicaire: