Rockwell PIDE PID parameters explained
The Rockwell PIDE function block is unusual in supporting two algorithm forms explicitly, selected by a parameter. In dependent form the block behaves like a conventional ideal PID with gain scaling all three terms. In independent form the proportional, integral and derivative gains are genuinely separate.
This matters more than any single tuning value, because the same three numbers produce different behaviour depending on which form is selected. Establish that first.
Which direction makes the action stronger
On your system a HIGHER proportional number means MORE aggressive proportional action.
Reset is in minutes/repeat, so a LOWER number means FASTER integral action.
This is the single most consequential thing to get right. Proportional band and gain move in opposite directions, and so do repeats-per-time and time-per-repeat. An engineer who has moved from one system to another and carried the old habit across will make loops worse while believing they are improving them.
- Proportional term: gain
- Reset units: minutes/repeat
- Rate units: minutes
- Algorithm form: ideal noninteracting
Dependent versus independent
In dependent (ideal) form, changing the controller gain scales the integral and derivative contributions along with the proportional one. That is usually what you want, because it means a single adjustment moves the whole controller's aggressiveness together.
In independent (parallel) form, changing the proportional gain does not touch integral or derivative action. Tuning rules written for the ideal form will not transfer, and adjusting one term in isolation changes the loop's character rather than just its speed.
The older PID instruction and the PIDE block do not express tuning the same way either, so a tuning set copied from a legacy PID into a PIDE needs converting rather than transcribing.
Why the algorithm form matters
The algorithm form matters because the same three numbers behave differently in each. In the series (interacting) form the derivative term multiplies the proportional-plus-integral term, so changing rate also changes the effective gain and reset. In the ideal (non-interacting) form the three terms are independent of each other but all scaled by gain. In the parallel (independent) form even that scaling is absent — gain does not affect integral or derivative action at all. Moving a tuning set between systems without converting for form is one of the commonest ways a loop that worked on one plant fails on another.
Before you change anything
Confirm the units on the actual controller rather than trusting the platform default. Systems get configured in unusual ways, engineering units get changed during a migration, and a controller imported from another site may carry conventions from there.
The fastest check is to make a small deliberate change and watch which way the loop responds. If it gets more aggressive when you expected less, the convention is not what you assumed.
Questions that come up
Which PIDE form should I use?
Dependent form is the more common choice and matches most published tuning guidance. The important thing is knowing which one is configured before you interpret the numbers, because both are in use and the faceplate does not make it obvious.
Related
Last reviewed 2026-08-01.