Hey all,
I'm sure it's possible to retrofit a gas oven with PID control, but would one gain too much if the gas can only be turned on or off, i.e. you can't control the amount of gas being burned?
If there would be faster time to stable and/or more continuous stable operation, how would one deal with the oven door being opened in the control cycle?
Thanks!
Forgive me, I'm rather rusty on PID control. For your application, I'd recommend PD control, the integration can make your system very unstable if the response time between valve control and temperature change isn't fast.
I can't really speak to the gain since I haven't worked with ovens, that's something I would recommend experimenting with, seeing the bounds of the temp oscillation when you power the valve on/off periodically.
Ideally, the system would take the door opening/closing into account. The door opens, the temp drops, the system responds by altering the gas flow, temperature is adjusted. I'd run some tests first with PD control before venturing into taking the door opening/closing into account. If you set up the oven to run at a constant temperature for an extended period (steady state), you could then open the door for X time and then observe the response on the temperature profile inside the oven. Based on that, you may be able to take the state of the door being open as an input to the plant and adjust the target temperature higher or lower depending on the state of the door (ie, set target temp +5deg when door is open). I'm assuming a very crude model, I wouldn't expect the oven to respond quickly or precisely if you're limited to a binary control (that you can't do PWM on?).
Hi, I heard several years ago about higher-efficiency gas furnaces that did a very fast on off cycling. If you can find information on those it might help.
Large ceramic kilns may have some proportional control.
My gas oven will not cycle quickly due to the igniter delay. You'd need spark ignition I think, but you could retrofit it.
Careful out there!
As a retired control engineer I have to admit I've never had need to use a system using only PD control, however I'm willing to learn. It's the "I" control term that eliminates the offset error with respect to time so if you want your oven temperature to match setpoint then "I" is essential.
If your oven thermal mass is relatively low then you may be able to dispense with the "D" term
On a similar topic I fitted up my domestic kitchen electric oven with a cheap Chinese PID controller with PWM output driving an SSR wired in place of the conventional thermostat. It uses a K thermocouple as a measurement device. I must admit it performs way above expectation, controlling oven temperature within 1 degree centigrade. The controller had an adaptive tuning function which effectively took care of thermal mass.
As to door opening and loss of temperature, I doubt you will get a controller that can cater for this sudden change in system parameters. An oven with an open door is a different system than an oven with a closed door so you are unlikely to find a controller optimised for two different systems. However a properly tuned controller will rapidly get back to setpoint once the door is closed.
How big is this oven? What kind of ignition system? Does it have a combustion air fan or is air feed from atmosphere? What would be the safety system in case the oven doesn't light but has gas feed to it?
cyclegadget:
How big is this oven? What kind of ignition system? Does it have a combustion air fan or is air feed from atmosphere? What would be the safety system in case the oven doesn't light but has gas feed to it?
Moreover - are you going to use this oven in a business or home, and do you have the proper license, certifications, etc to make modifications to a gas appliance (I am pretty sure most jurisdictions in the United States have laws about this sort of thing - in most places, you can't even plumb into a gas line as a DIY/homeowner job)...?
I agree with cr0sh here, "playing" with gas is ----> 
To design and make a controller for a gas system is not that hard in my opinon ( The electronics/ electricals parts / programing parts ) but to "touch" the gas line <-- open / re-install / cut the gas line is "BAD" <---- Remember the destruction of a neberhood in California ( I think ) of a gas line rupture and cause a big explosion and destroy the whole neberhood . But that project could be OK install into a electric stove ( with a few changes )
First thing is to install a commercial (Home depot etc) LP or Natural Gas alarm right where you're working. That's what I did when building a set of gas burners into a slab of granite. Only have tripped it when I fed it a little gas to test...
The spray can of 'real' leak tester is also better than the traditional soap bubbles. Don't mess around.