Small Heat Treat Oven

[Brian Rupnow]

My heat treat oven has a round hole in the door (see post #8 for picture). I assume that is where the temperature testing probe connects to tell the controller how hot it is inside. that seems to me to be very inconvenient. Can I put a new hole in thru the side of the oven and mount the probe permanently in the side? What can I use to plug the hole in the door?---Brian

[Brian Rupnow]

The heat treat oven is in it's final home position. I drilled thru the side of the oven near the top, for the thermocouple, not interfering with the heating coils, about  4" in from the front (the oven is 8 1/2" deep.). I have power to the controller and the oven is heating up and the digital display on the controller is lit up. I have to put a new 3 wire lead on the oven, as it only had a two wire lead and plug on the cord, with no ground. I went down town this morning trying to find a length of 3 wire heat resistant cord, but all the businesses are shut down for covid. I'm happy with what I have here, and hopefully will be even happier when I get the controller figured out.

[Brian Rupnow]

After what seemed to be an incredible lot of faffing about, I have the oven controller operating, holding the oven at a constant 1100 degrees Fahrenheit. I may want to tighten up the parameters, (its cutting off power at 1130 degrees and restoring power at 1020 degrees. I will search about in my instructions and see if there is a way to do that.

[crueby]

Nice setup. How hot can the oven go? This runs on 110volt? Whats the current draw?

[ddmckee54]

Brian:

That variation could be caused by a couple of things, among them are the deadband setting, and the proprotional gain (P) being too low.  According to the manual the factory default for the deadband is 0, so if you haven't changed that parameter, just leave it alone for now.  If you didn't change the proportional value then it's still at the factory default of 30, this may be WAY too low.

The way your PID controller is working is probably something like this.  The PID controller looks at the process variable (PV), this is the signal from the thermo-couple telling the controller the actual temperature.  The PID controller also looks at the set-point variable (SV), this is your set-point temperature.  It then uses both of those variables along with the P, I, and D variables in its' PID equation to determine the error signal.  This error signal is used to control the PID controller's output, in your case by turning the SSR (Solid State Relay) on and off.  For example if the PID equation says the error signal is 30%, then the SSR will be turned on for 30% of its' duty cycle.  The deadband gives the PID a +/- range of values around the set-point where it will do nothing.  Clear as mud?  It gets better.

The P variable will determine how much of a kick in the butt the error signal is going to get each time the PID equation is calculated.  A lower P value means not so much of a kick so it will take more cycles of calculating the equation before the effect is noticed.  The I and the D values are used to smooth out the error signal.  You want the PID to start backing off the error signal as the PV approches the set-point, not after it over-shoots it, I believe that's the function of the D parameter.  I believe that the I parameter is used to smooth out the oscillations around the set-point.  Ideally your PV should graph out to be a nice smooth curve that approches the SP as rapidly as possible without overshooting it.  Tuning a PID loop is something of a black art to tune it well.  I was doing it for a couple of decades, and every now and then one would STILL kick my butt.  That being said, you can get them to run, and run fairly well, without too much trouble.

I would start by doubling the P value, setting the I value to 1 (Most loops will run quite adequately as a PD loop), then make a major change in the temperature set-point and see how the controller reacts.  How well you need to tune the loop will depend on how tight you want the temperature control.  In case you haven't figured it out yet, manually tuning a loop involves a lot of standing around waiting for things to stabilize.  If you want a "by the book" method of tuning a loop along with "rule of thumb" parameter settings for a BASIC tune, I can provide them too.

Don

[Don1966]

Great explaination Don. , just to clarify, if the oven remained below temperature, “I” the integral would act to increase the head delivered. However, rather than stop heating when the target is reached, "I" attempts to drive the cumulative error to zero, resulting in an overshoot.
Derivative tuning attempts to minimize this overshoot by slowing the correction factor applied as the Pv value is approached.  It was always my understand that ‘I” was repeats per minute why as “D” was minutes per repeat. One adds time while the other grabs it and pushes it to PV.
Just my two cents.


Regards Don

[Brian Rupnow]

I didn't do any more messing around with the controller today. I still have to learn more about it and I've seen that I can get the oven to go to 1100 degrees F and hold it there, which was my biggest question. This oven had the poorest hinge plates on it that I have ever seen. They are only .025" thick material and were bent all to heck. I whittled out some new hinge plates this afternoon from 3/16" aluminum and tomorrow I will concentrate on getting the "door" to fit better and install the new hinge plates.

[Art K]

Brian,
You've got yourself quite a project here. Sounds like Don knows his stuff, really good in these circumstances.
Art

[Brian Rupnow]

This morning I installed the new hinge plates on the oven. The door fits much better with no gaps now. I managed to turn on the Auto-tune function and with the temperature set at 1110 degrees the oven shuts off at 1112 and turns back on at 1108. I'm very happy with that.---and for now, that is all the programming I need. in future I may want to do other things, which may mean that I have to learn more functions, but for now, I'm fine. This essentially ends my work on the oven. Thanks so much for the help you fellows gave me. I would have been lost without some guidance from forum members.---Brian Rupnow

[ddmckee54]

The auto-tune takes all the black magic out of tuning a PID.  If you are holding within a +/- 2 degree range for this application that's more than acceptable.  Unless you want the controller to give you alarms, or use it to control a ramped heating cycle, there is nothing more you need to do with it.

I don't see anything in the DTXG manual that would indicate the ability to control a timed heat or cool cycle.  You can limt the output so with experimentation you might be able to get a controlled heat ramp to temperature.  But the time to soak at temperature would be all on you.  It looks like this controller will either heat to a temperature, or cool to a temperature, depending on if you put it the heating or cooling mode.

Don