SLC PID Tuning gone awry -- please help

[Lakee911]

Hello,

I'm having a lot of trouble tuning a PID loop on a SLC 5/05. Can someone tell me if I'm doing something wrong? Please see attachment to this thread.

Basically, I have a gate opens into a wetwell. I am opening the gate to allow flow into the wetwell. The wetwell has a PID loop (that works just fine) that varies the speed of pumps which maintain a wetwell level. What comes in, gets pumped out. After the pumps, there is a magnetic flow meter. The loop I am having trouble with, modulates the gate to maintain a flow rate on the magnetic flow meter.

My problem is, it jumps around from max to min doesn't seem to track properly at all. It's reverse acting. The P parameter seems to track properly, but when I add some I, it tracks the opposite way that I'd expect. I'm not sure if my parameters/ranges are correct.

I'm scaling my setpoint (F16:22) and the process variable (F14:226) from 0 - 16383 before sticking them in the PID. My control variable (output) gets scaled down to 0 - 100. Temporarily, I've limited the output from 0 - 40% to prevent completely draining another process as this wildly swings around.

Unfortunately, this snap shot came when the process was finally shutdown (and I had pulled out most of my hair).

Can anyone help?

Thanks!

 

[Lakee911]

I went and took a look at my PID loop that maintains wetwell level, which is working fantastically! Unfortunately, I noticed a discrepancy here....

I'm scaling the Setpoint PV from 0 - 16383, but I'm scaling the SP from 0 - 2000. My P (Kc) is 14.5 and my I (Ti) is 8.9. I have no D (Td).

I'm afraid that if I adjust the range of my SP, it's going to need to be re-tuned. Does this matter if its working correctly?

I'll admit, I have more experience with ControlLogix than these old beasts (SLCs).

Thanks.

 

[Mickey]

It would be better if you could post your complete program the .RSS file for your SLC5/05. Zip it first then post.

Don't fix something that's not broken.

 

[Lakee911]

Sure thing. See attached.

Thanks.

 

[Mickey]

Just a couple of things while your waiting for the PID gurus to check in.

Your PID is set for "STI" time mode but the "STI" is set to control ladder 10, the PID instruction is in ladder 9. This will give unpredictable results.
Also all of your PID instruction are writing to the same control variable address N7:64 , I'm not sure what problems that may or may not be causing but IMO should be corrected.

 

[Mickey]

One other thing before I have to go do my taxes.

How stable is the input signal ( flow from the flow meter)?

One tool to help the gurus is trends. ( PV,SP,CV)

 

[Lakee911]

Just a couple of things while your waiting for the PID gurus to check in.

Your PID is set for "STI" time mode but the "STI" is set to control ladder 10, the PID instruction is in ladder 9. This will give unpredictable results.

I've got many PID loops. I think I should change them all to timed mode, no?

Also all of your PID instruction are writing to the same control variable address N7:64 , I'm not sure what problems that may or may not be causing but IMO should be corrected.

You're right! That might have been the biggest problem. Not all of them will be used at the same time, so perhaps they were all fighting eachother. I'll work on getting that corrected.

One other thing before I have to go do my taxes.

How stable is the input signal ( flow from the flow meter)?

One tool to help the gurus is trends. ( PV,SP,CV)

I don't have trends with me at the moment, but if I recall correctly, when I'm not varying the gate position, the flow is fairly stable. It definately does not vary more than 5%. I'll try to capture some on Monday. The SP won't be changing much at all.


Thanks!

 

[Lakee911]

So, I made the changes ... things are well. I've yet to get it tuned, but its responding like it should. I can take it from here. Must have been the timing and the overwriting of data. Thank you!!!

 

[Mickey]

Your welcome. And thanks for the update.