RSLogix PID Tuning Issue

harleymc100

Member
Join Date
Jan 2014
Location
london
Posts
21
So a bit of background. We have a tank that pH corrects dairy waste water. We variably dose acid to keep it at a setpoint of pH 4.8 for optimum treatment.

Can any experts shed any light on our PID loop here?, As its not great.

We have cut the max pump speed to 65% to stop overshooting ( down) below setpoint, but my thoughts are, a properly tuned PID ? without the need to limit pump speed?

It doesn’t have to be perfect, it can bounce 0.5 above or below setpoint if needed.

Do the values entered here seem like they should work?

Image 26.png
 

Attachments

  • Image 26.jpg
    157 bytes · Views: 42
When I've tuned pH-controllers in the past, they process together with measurement response has been slow to very slow.

Have you done a step response in manual mode?
 
As Dirt says - response tends to be slow so make sure your PID timer is about 10s or more - play with that time for a bit before attempting tuning.
 
Thanks for the replies fellas. I'm only starting out with understanding PIDs so im not sure about step response in manual.
Do you have any advice on how i can do this or a recommended reading on this?
And yes it is extremely slow, ie, minutes between oscillations sometimes.
 
Have a search for the Rockwell PID tuning guide. That might help to get started and to get you in the right ball park e.g. PID instruction timings and approx. PID values plus it will explain some of the concepts. You often end up using trial and error to finalise though. If done methodically that's not a bad way of doing it. You just have to make sure your nice steady state response is still ok when you make a step change.

The PIDE instruction has autotune although I've not used it enough to have a lot of experience with it - the few times I have autotuned it's given me mixed results... And that is with pretty simple oven temperature control loops.
 
It looks like your gains are too aggressive. The pump is basically ON/OFF at 65% if I'm looking at this correctly.
 
Last edited:
I've never been directly involved with pH control, so I can't exactly speak from experience, but my boss has had success in the past using gain scheduling in pH control. Essentially, your P term becomes different as the measured value changes. Not sure the exact in and outs of it all, but it might be something you can look into. Someone else here can feel free to debunk this, but I believe that pH is logarithmic so it can be kind of difficult to control with a single gain setting.
 
pH is not linear if follows a titration curve, if you look at a small section it is almost linear, but once it gets out either way it drastically changes. I have had good success with gain scheduling in Rockwell applications, in Siemens PCS7, there is a Smith Predication algorithm that works well also.

You first need to determine the lag time in your application. Put the controls in manual, and increase or decrease the output by 10 percent, how long before you see a change in ph?
 
You can run a titration and record at each step the amount of acid or base added and how much rise/fall you saw in pH.

You could plug that information into a function generator in Logix5000.

The problem with tuning pH is that the base chemical takes much longer to affect the pH. Proper mixing helps and standard PID controllers will struggle to keep the pH at setpoint.

Building and controlling from a characterization array is just another tool in the box.
 
Hi Guys, lots of awesome information in here. Ill give it a go and let you know the outcome. Ive got a lot of reading to do. Thanks!!!
 

Similar Topics

I am using PID instruction in RS Logix 5000 with independent equation. I am confused about its tuning. Please somebody explain under listed...
Replies
1
Views
4,921
Hi Guys, Can anyone help me i have a machine which has already been programmed but the tempertures are not reacting quite enough it drops +- 10c...
Replies
1
Views
5,285
hi guys, i have a machine which the heaters pid was set up when the machine came in, but now the heaters are taking along time to heat up and when...
Replies
9
Views
17,251
Hi, Long time not in the forum, and not in the programming. I´m getting back. I was issued a conversion from RSLogix 500 to RSLogix 5000 (studio...
Replies
0
Views
1,284
So I have a PID loop on an 1756-L61 running V17 software just for background. Also the PID PV is a pressure transmitter and the CV is speed sent...
Replies
1
Views
892
Back
Top Bottom