Allen Bradley 5000 P.I.D. Setup

Cookz615

Member
Join Date
Feb 2017
Location
tennessee
Posts
18
Hello everybody,

It's been a good while, hope everyone is doing well!

I've recently been task to setup a P.I.D. to control the water level in our sludge collection hoppers. I am using an Einfield S2-025-U-04 proportional control valve with an SMC cylinder and a SICK linear transducer for positional feedback of the cylinder to the Einfield valve.

I have a P.I.D. setup and controlling somewhat but cant seem to figure out how to completely take out the pulsation of the water level due to the output of the P.I.D. constantly updating and changing the output.

This is the first time I have ever touched a P.I.D. "As you can probably tell" so I don't know if this can be tuned out in the P.I.D. instruction completely or if I need to write logic to open up the acceptable window for level and only allow the P.I.D. to run when its outside that window.

Ideally I would like to keep the water level at around 15 in. I believe I could stand to do a plus or minus 1 in using compare statements to enable or disable the P.I.D. Not sure the best course to take.

Any help would be greatly appreciated.

Thanks,
 
Enabling/Disabling the PID is not the solution. You probably just need to actually tune it better.



Check here for the Zeigler-Nichols approach, which is fairly simple to do (you can create a trend in Logix to determine the time constant of your process).
 
Is it possible for you to post an image of a trend charts showing the controller output percent and level measurement? Ideally, this would be during a time when the process is somewhat stable (i.e., water input to the hopper is steady), and you have the PID in manual mode at a fixed percent output. Then change ("bump") the controller output in manual by an amount that changes the level, but still within your acceptable safety tolerance. Stop the trend chart after it appears the level has reached a new equilibrium.

This request assumes a couple things: that your water/sludge input is fairly stable over the response time of the valve, your output flow is reasonably linear in the control range (valve percent) where you normally operate, and that level measurement is responsive and repeatable. With a correctly configured PID loop, this type of trend can be used for ballpark tuning, or to identify problems where tuning might not be root cause of observed issues.
 
Ditto the suggestion about getting a trend.

Check here for the Zeigler-Nichols approach, which is fairly simple to do (you can create a trend in Logix to determine the time constant of your process).
NO, NO, NO, NO!!!! NEVER!!! Z-N keeps control techniques in the dark ages and isn't applicable to many system like tank level control

You should always state whether the valve controls the inflow or outflow. It makes a different.
I am assuming the valve controls the outflow.
I am assuming maximum valve outflow matches the max inflow

Why use a PID? Keep it simple.
Just use proportional band control where the valve is closed at 14 inches and open 100% at 16 inches. At 14.5 inches the valve is open 25% and at 15 inches the valve is open

Most proportional valves do not control flow proportional to the control signal. Many proportional valve are non-linear or have a huge dead band so it may take 20% control signal to get any flow. This too should be stated or a link to the valve specs should be provided.
 
NO, NO, NO, NO!!!! NEVER!!! Z-N keeps control techniques in the dark ages and isn't applicable to many system like tank level control

Not wanting to hijack the thread to much but I'm a big opponent of ZN as well. The fact that you are willingly pushing a process to the point of instability is just bonkers in the real world
 
I only mentioned ZN, as it is a very basic, and simple method of at least getting something under control. It seems a lot of people just drop in a PID(E) instruction and expect it to work by randomly popping in numbers.


Almost every write up of ZN at least goes through the basics of figuring out the time constants of your system.
 

Similar Topics

Hi all, the machine I am programming the Rockwell PLC for has about 100 parameters that need to be adjusted for every customer. I would like to...
Replies
20
Views
1,818
Hey all, I'm doing a more standardize version of one of my programs and organizing things into UDTs and whatever else I can do to shrink the...
Replies
0
Views
409
Hello everyone, I have recently started a new project using ST in Studio 5000. Previously, I have programmed in ST in Siemens. As i was writting...
Replies
4
Views
2,837
Hello everyone, I have recently started a new project using ST in Studio 5000. Previously, I have programmed in ST in Siemens. As i was writting...
Replies
1
Views
1,384
Ok so I assume the answer is no, because it's AB. Other Manufactures have small utilities for this. Is there any way I can "Download" a program...
Replies
7
Views
3,758
Back
Top Bottom