Announcement

Collapse
No announcement yet.

PID Bias problem

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts


  • PID Bias problem

    I have a PID loop configured for Bumpless initialization. In my experience, Bumpless transfer should set the bias equal to the output when the PID initializes. However, when the PID switches to Auto, the bias drops from around 50% to around 0, which is causing a lot of problems on startup.

    Is this intended behavior, and what value is the bias using to initialize?

    Click image for larger version

Name:	what is the bias doing.png
Views:	48
Size:	35.5 KB
ID:	123895
    Attached Files


  • #2
    I think it sets it to the .Output MINUS the current ERROR * Kp. Hence, the net result of the first .Output calculation at that instant is that the .Output value is bumpless??? Otherwise in that situation, the .Bias would be the "old" output (~45%), then when we add that the FIRST PROPORTIONAL error term to set the initial auto .Output term, and .Output will change dramatically. It appears you have a large-enough error, hence the Bias going to 0 (actually -20%)???

    However, I do see that you still got a good bump on the .Output, regardless. Not sure what happened.

    BobO will have a better idea, but he's not in at the moment.
    There are 10 kinds of people in this world, those who know binary, and those who do not.

    Comment



    • #3
      Originally posted by franji1 View Post
      However, I do see that you still got a good bump on the .Output, regardless. Not sure what happened.

      BobO will have a better idea, but he's not in at the moment.
      I think I found it. I have a rung that writes the .Output from another source when the PID is in Manual so that the transition is as bumpless as possible. I changed the limits of the PID output, but forgot to change the scaling. Thanks!

      Comment

      Working...
      X