Host Engineering Forum

General Category => Do-more CPUs and Do-more Designer Software => Topic started by: RBPLC on July 26, 2020, 09:32:13 PM

Title: PID Output
Post by: RBPLC on July 26, 2020, 09:32:13 PM

In the shown photo I've got a TC reading ambient temperature. When I change the .SP to slightly below the current ambient temp, the forward acting loop is still giving slight output with 0 Rate and 0 Reset. Why is this? I would expect that by setting the setpoint to anything lower than the current .PV would yield zero output from the loop. What am I missing?

Title: Re: PID Output
Post by: franji1 on July 27, 2020, 10:05:28 AM

The Bias term is non-zero, hence it is affecting the .Output calculation.  The Bias can become non-zero based on your "bumpless" transition behavior option or previous tuning constants.

Title: Re: PID Output
Post by: RBPLC on July 28, 2020, 07:19:27 PM

What's the likelihood of adding a tickbox that would keep (force) the bias term at zero on a manual/auto transition? For many of the ways I would use the PID, it would be nice to keep it at zero during a transition. Not sure if this would be useful or not to the general user base. For now I'll just one shot the bias term to zero during the transition.

Title: Re: PID Output
Post by: Controls Guy on July 28, 2020, 09:09:37 PM

I was thinking sort of the same thing.  If Ki is set to zero (at least from Do-More), maybe clear bias by default?

Title: Re: PID Output
Post by: Controls Guy on July 28, 2020, 09:10:58 PM

Not continuously keep it at zero, just one-shot if the user clears Ki.

Title: Re: PID Output
Post by: ADC Product Engineer on July 29, 2020, 08:26:10 AM

Quote from: RBPLC on July 28, 2020, 07:19:27 PM

What's the likelihood of adding a tickbox that would keep (force) the bias term at zero on a manual/auto transition? For many of the ways I would use the PID, it would be nice to keep it at zero during a transition. Not sure if this would be useful or not to the general user base. For now I'll just one shot the bias term to zero during the transition.

This is a fairly common issue.  If Mark/Bob can do add a tickbox or a structure field to force bias to zero, I would lend it my vote.