temp. loop tuning

We have a machine that has a gas fired oven on it that heats mold frames running in and out of it. It has a GE PID loop instruction on it that controls a burner modulating motor. During steady state with no load and both doors closed the oven keeps it's setpoint within +/- 2 degrees. However, once we start loading the frames and opening and closing doors the process starts to oscillate to a point where it is about +/- 30 degrees at times. Were using P = 17, I = 8.5 and D = 0. My question is now I need to tune this better for loaded conditions and would like some reccomendations on howe best to achieve this. I am assuming that this fine tuning should be done while the operator is running the machine and loading and unloading molds. Thanks for the help.

With only single doors (ie goes in door and goes out door) on the oven I see no way to accurately control temperature when you open one of the doors. Kinda like your garage when you open the door in December.

I think I would turn the flame off (or low fire) while the doors are open - why use the gas to heat up the shop area?? After the doors are shut and a 15 second delay to let things normalize a bit then turn on the PID sensing to control temperature.

Dan Bentler

The oven can hold up to three molds at a time and oven times vary from 45 seconds to 3 minutes and during that time molds are entering and exiting through the doors. We have 5 other machines with this kind of control loop and they perform within +/- 5 degrees while in process but on this system the oscillations are quit wide. Just want to get a better idea of how to eliminate such an oscillation. By increasing/decreasing P and I, but not sure which would be better to tweak.

Where is your point of measurment (process variable PV)?

If PV is near the door(s), and you are opening and closing them, a PID is not really controlling the process.

Can you move the measuremnt point to the center of the process away from the doors?


Plot or record your control variable output (CV) percentage during idle state, and during heating state (with doors and process). After you analyze the CV, find the AVERAGE CV during heating. You may obtain a more stable process by using a steady state CV during heating, and then switch back to PID during idle.

You say you have five other ovens and they perform well, but this one oven does not.
I'm thinking this is might not be a tuning problem. Anyone else concur? If there is an underlying correctable problem then re-tuning the oven control loop probably will help, but it could still render less than satisfactory performance.

What is different about the one under performing oven?

Without knowing more about your process no one here can reliable say adjust P or adjust I. If you have a plot of the oven performance and can post that plot that will help. An ideal plot is one where the oven is allowed to reach steady state with the loop in manual, then the loop output is increased in manual, and the oven is allowed to again reach steady state.

here's an example of the type of graph mentioned by my distinguished colleague Alaric ...

http://www.plctalk.net/qanda/showthread.php?p=27858&postcount=10

"Were using P = 17, I = 8.5"
Is P Gain or proportional band, if it's gain it seems very high to me
Is I minutes per repeat or repeats per minute?
Does this machine have the identical controller, perhaps you are getting the terms mixed?

Alaric said:

You say you have five other ovens and they perform well, but this one oven does not.
I'm thinking this is might not be a tuning problem. Anyone else concur? If there is an underlying correctable problem then re-tuning the oven control loop probably will help, but it could still render less than satisfactory performance.

What is different about the one under performing oven?

Without knowing more about your process no one here can reliable say adjust P or adjust I. If you have a plot of the oven performance and can post that plot that will help. An ideal plot is one where the oven is allowed to reach steady state with the loop in manual, then the loop output is increased in manual, and the oven is allowed to again reach steady state.

Click to expand...

I DONT KNOW ABOUT CONCUR BUT I HAVE A COUPLE OF WHAT IFS
What is this units age and operating life compared to others?

Uhhh this may be a dumb question BUT
You are not using PID control to merely control a solenoid valve or even a two stage burner setup ie high low fire?

How is the flame controlled - is it done with modulating control valve (makes sense if you have PID).
Could valve be worn or sticking?
What about regulator has the pressure to the valve changed or is it changing?
Dirt or other contaminants in control valves?
Draft control or draft fan control?
Door seals shot?

I would consider going thru the entire machine to ensure there is not another problem.
Dan Bentler