PID Control with discrete on/off for CV

Gentlemen,

I'm faced with controlling a valve that has 480 feed to it, but using 120Vac On..Off discrete signals to position it. It does have analog feedback for position. It also had a Flow Transmitter for which to control the valve/process to a setpoint.

Here's what I'm thinking of doing, and would love someone elses input if they feel like it....

'Normal' PIDE control block with the FT as the PV, but a unitless output for the CV. This value, would then somehow give me a value of which I could scale a timer, and in the logic pulse the output for open based on this value.

Am I trying to re-invent the wheel, or anyone seen anything would work better. I've never done it, but I would think thse valves have been used many times before by other people.

Thank you....

This is what you´re looking for, this example is for SLC500, but it´s very simple to modify it for CLX, please tell me if you have any problem.

Thank you widelto, I like what you have there, but I'm worried I'd fry my contact outputs if they cycle that fast? After some thought I think I can get by with some very simple on/off stuff.

If I'm reading correctly, you have an analogue setpoint and process value (SP & PV) that you want to control using digital outputs (CV).

I did this to control a stock valve and refiners on my process. I'm using RSLogix5000 and used the function blocks PIDE and STRP (split range).

I hooked the PIDE CVEU to the STRP IN and had the heat out and cold out as my digital outputs for open and close.

Attached is a pdf of the function block with the settings for the STRP which took a while for me to refine. Its probably not the best way to do this, but it has worked great for me so far.

Creativepaper, your approach is better than mine, I like it a lot. Tks for sharing it.

What I'm picturing from your description is a valve operated by a 3 phase 480V motor with two 120V contactors. It that correct?

Scale the PID EU to 0 to 1. This will give you a multiplication factor that you can use to determine the desired position of the valve.

For example, if the valve position feedback is a 4-20mA sensor and the PIDE CV is 45% then CVEU is .45. Then compute 4mA + (16mA*.45) = 11.2mA. Position the valve so that the feedback is ~= 11.2mA. But unless the valve moves very very slowly I don't think you are going to get really good valve position control. You are going to have to determine some kind of hysteresis band within which you don't attempt to move the motor. That could make tuning very difficult. On top of that, most valves are notoriously non-linear.


I use motor operated butterfly valves on large vacuum furnaces to control pumping speeds. When pumping large vacuum chambers from atmospheric pressure its desirable to gradually open the valve to minimize the wind in the chamber and minimize the load on the roots booster. In this case I don't have to have ultra-precise positioning of the valve, as long as I'm within 5% its fine. I use a linear relationship to compute position with respect to vacuum, expressed as a value from 0 to 1. Then this is multiplied by the valve time constant (how long it takes to move the valve from full open to full closed) to determine total motor run time to move the valve to the desired position. Its not super precise, but good enough for my app.

I hope that helps.

Sorry to bring up such an old thread, but I am working on a project almost identical to this when I came accross this thread. Tvich, what solution did you come up with for this?

bobbys1967 said:

Sorry to bring up such an old thread, but I am working on a project almost identical to this when I came accross this thread. Tvich, what solution did you come up with for this?

Click to expand...

bobbys1967,

It looks like Tvich has not posted for about a year, so not sure he's(she) still around. It might be best to start a new thread and post details about your process and hadware you have. I'm sure the PID guru's will be happy to help.

I didn't have to deal with it. The valve DID end up coming with it's own 4-20mA controller.

I liked creativepaper's idea.

Watchout though, as I've heard of controls 'burning out' controllers if you're turning on a motor at say 120Vac, that controls a 480Vac valve,..by cycling the on/off too much.

Good luck.