“PID Help Micrologix 1200”

Opening post
Greetings All J
Its been a while since I have been on. Someone hacked my email account and was able to create some serious havoc on me. I had to create a new user on the forum because of it. Never the less I have still been reading all the great informative post, but now it is time to post my own.
I currently have a waste treatment project that I am working on. It was kind of shoved my way if you will, in the name of trying to save money. This is my first project working with PID loops and I have to say it’s kicking me around. I have read many posts, on this forum from a lot of the experts here but still I am having issues. I think I need some help at this point. I have worked with analog for speed references and valve positioning but I have never programmed a PID before.
I am going to try and supply all the information required so that you may have a good picture. Also more importantly maybe I can leave something behind here for the next guy to find and it may help his journey too.
PROCESS DESCRIPTION
1. Intent of control system is to limit the influent flow of industrial waste (IW) from the equalization (EQ) tank to the Dissolved Air Flotation (DAF) tank to 25 gpm but if the EQ tank becomes full override the control valve to a full open position.
2. Influent Sump Pump from the factory is operated by an on/off level float( high and low floats) with an internal dead-band to prevent short cycling.
3. On rising level on the Influent Sump Pump, the Influent Sump Pump turns on and fills the EQ tank. At the same time the Feed Pump from the EQ Tank turns on and feeds the waste to the DAF through the Flow Control Valve.
4. As IW flows to the DAF, the magnetic flow meter controls the modulating control valve to a set-point of 25 gpm. On rising level in the EQ tank , FIC-101 will increasingly close the control valve to limit the flow. On falling level, FIC-101 will increasingly open the control valve to maintain the set-point flow.
5. If the EQ tank becomes totally full, the control loop between FIT-101 and FIC-101 will be overridden and the modulating control valve will be fully opened to prevent overflow/damage to the EQ tank. EQ Tank level is read by ultrasonic transmitter.

I know the above description will raise some questions J


Devices:
PLC
1. 1762-L40BWAR MicroLogix 1200
2. 1762-OF4 4 Channel Current/Voltage Analog Output Module
3. 1762-IF4 4 Channel Current/Voltage Analog Input Module
Analog
1. LVU832 Ultrasonic Level Transmitter 4-20mA
2. Siemens Magflo Meter MAG5000
3. Triac Electric Actuator WE series 4-20mA in and out


I am trying to run my pid off the 4-20mA that I receive from the flow meter. I currently have the analog input configured for scaled for pid. I also have the output for the valve configured the same. I thought that since I was not trying to totalize. This would not be to terribly bad. But I have a ton of questions at this point. Things related to scaling, filters, etc.
If anyone has time and your bored. I would like the educationJ

Lancie1 said:

To get dozens of answers to your questions, post a picture of your PID rungs, or ZIP your RSLogix program and post it here.

Click to expand...

Lancie1 thanks for the reply

here you go. I have been on this all day so far. I just dont understand what I am missing.

are you having trouble with the system in general – or with the PID control in particular? ...

if it's the PID, please give us an explanation of what type of problem you're having ... (example: not enough output; too much output; erratically fluctuating output; etc.) ...

your PID is currently set up for Proportional action only ... have you tried dialing in a setting for Integral action? ...

your PID Loop Update time is currently set for 5.80 seconds ... if this isn't a VERY slow acting loop then that setting will need to be made faster (smaller number) ... something like 0.50 seconds is much more typical than what you currently have ...

Ron
Yes, the PID in particular. and Welcome. I have read many of your post. very educating....

I finally was able to achieve some progress after several hours of making some adjustments to the flow meter settings and playing with the P and I portion of the PID.

I currently have the PID functioning and it will hold PV(process varible) to within 1/2 gal of SP(set point)once I get to setpoint. I do notice that the PID has some issue still though.

I do have some erratic behaviour.. when I graph the data to and from the pid meaning the PV and CV(control varible). the control varible tends to be sharp and jagged. not smooth as I believe it should be. which I am not quite sure how I am staying within 1/2 gal of setpoint. I am going to try and collect some trending data tomorrow to post for you all to see.

I also can see that the PID is slow to react. At first I thought that this could be my actuator being slow. but watching the data in the output word address. I can see the data being calculated is not being updated or changing fast enough.

The PID is also slow trimming back the CV if setpoint is well below the current flow. and once it does get to setpoint it over shoots by 7 gal and then comes to meet the SP Value.

These are my observations thus far.

taking your advice I changed the loop update time setting in the pid setup screen. which also got me to thinking about update time of the flow meter that sends the 4-20mA signal for the pid process varible. I found that the update on the flow meter was set to 5 seconds. I changed it to .5 seconds.

I dont think that I am still scaled quite right yet though. I need to look at my flow meter scaling for the 4-20mA output to see what the scale is. currently have the SCP scaled for 0-518gpm. I dialed these numbers in by making adjustments to my values while watching the LCD display on my flow meter. these are the same values that I am using in the pid set up screen for setpoint min-max values.

I would really like to see how tight I can get this PID loop functioning.
I am posting my updated code so that you can see my new PID settings as of right now. and by all means I welcome your suggestions.

dont critic me to hard some of the code is still in progress