Logix PIDE Scan

[rigsha97]

We have a routine that we trigger with a JSR every 10 seconds that has PIDE instructions. This resides in a Periodic task that triggers every 40ms. This PIDE controls well. I want to move into its own Periodic task that triggers every 10 seconds. When I do this, the PIDE goes out of control quickly. When you compare the two instructions the Delta T for the first PIDE is 40ms. The Delta T for the second PIDE in its own task is 10 seconds. If I am triggering the JSR every 10 seconds in the first periodic task of 40 ms why is the delta t at 40 seconds and not 10 seconds like the first task?

 

[Mispeld]

It sounds like that routine with the PIDE instructions is being run at the 40 ms periodic interval. Are you sure there are not multiple JSR instructions for this routine?

Edit: one thing you can try is putting an ADD instruction in the PIDE routine that unconditionally increments a DINT tag value. You can watch if the value only changes every 10 seconds, or if it is jumping as if incremented every 40 ms.

 

[rigsha97]

There is only one JSR for this and when you watch the output of the PIDE, it only changes the output every 10 seconds as expected.

 

[Mispeld]

There's a detail on page 1084 of this manual regarding the timing mode of PIDE:

Logix 5000 Advanced
Process Control and Drives
Instructions

It implies that the instruction uses the periodic task interval as the instruction's time interval, regardless of whether it is scanned at a slower rate by a timer or other means.

So when you moved it to the 10 second task, the I and D gains would be way off, like 250x for the independent gains, if this is actually how it works.

It sounds like the instruction only calculates a true delta-T when the instruction is contained in a continuous or event task. My experience is only with the continuous task approach.

 

[rigsha97]

I understand and what you state does reflect what I am seeing. Thank you for your time and help!!

 

[drbitboy]

We don't have enough information to answer the question. Please print the project, with all configuration (tasks, task timings, etc.), to a PDF, and attach (

) that PDF it to a post.

@Mispeld put his finger on the direction to the solution: for a start, what TimingMode is configured for this PID?

PLC programming and configuration is about time, and that is doubly important for PI, PD and PID controllers.

 

[drbitboy]

Which form of the PID equation is this PID using: independent gains or dependent gains?