PID cycling beyond set point

OccamsBlade

OccamsBlade

Jim Dobbler
Joined
Jan 23, 2014
Messages
227
Been a stalker on the site for some time now, and finally gearing up to make my own blades.

So here's what's happening. I bought a toaster oven to use as a heat treat oven, and used Stacy's thread on how to build it. Once I got it all together I ran autotuning as the instructions for the PID suggested. I set the temp at 100 degrees to test it. In the first cycle it hit 100 rather quickly, and the ssr switched off. It overshot by about 45 degrees. Once it came back down to temp it switched back on, hit the temp and switched off. The second cycle it overshot by about 22 degrees. Came back down and started the 3rd cycle. Once it hit temp it started cycling on and off repeatedly and would not stop until I turned off the controller. It wen't about 60 degrees past temp before I switched it off. It will do it every time now almost immediately unless I run autotuning.

Not sure if I wired something wrong or if it's a PID setting like hysteresis. Haven't been able to find anyone having the same issue in the forums or elsewhere. I remember reading somewhere that it will cycle like this to prevent overshooting.

Anyway, was hoping someone could point me in the direction of a thread or basic instructions (for noobs) on how to set up a pid for the first time. The instructions it comes with aren't that great. In the meantime I'll be looking into exactly what hysteresis are.

Thanks!
 
Set hysteresis to about +;-5 degrees. Run autotune, and then run the program. In the beginning it will overshoot. As it "learns" it will find how long to turn "ON" to not overshoot. It usually takes 15 minutes to zero in the program.


That said, 100F is a but low for a PID to focus on, Ambient is only 20-30 degrees below the target. Try tuning at the temperature you want to use. 400F is a good tuning target.
 
You can also add a fire brick to inside of toaster oven for a bit of extra mass. Takes a bit longer to heat up, but also tends to stabilize a bit more. My toaster oven normally controls ±1ºF
 
Yes, a plate of 1/2" steel (or even better, aluminum) on the shelf will greatly smooth out tempering. If your toaster oven has two shelves, place a 1/4" sheet of aluminum on each one and put the knife blade on the lower shelf with the TC's tip right next to it ( only the tip reads the temperature).

Placement of the TC is important, as it may receive too much direct radiation from the heating elements. Optimally, it should be where the knife is. A clad TC can be inserted in a hole drilled in the steel/aluminum plate, which will make for a really close tolerance once the oven stabilizes. It may take 20-30 minutes for the plate to fully soak and the oven to stabilize at the correct temperature.
 
Great, thanks for the help and suggestions! I'll work on it this evening.

I was wondering about the size of my oven and how it might effect single blades. I got a larger one in case I wanted to do batches and larger blades later on down the line. I'll have to pick up a steel plate, and maybe some fire bricks, next time I'm at the hardware store.
 
So here are the results from last night. I set the temp to 400 degrees, set the hysteresis to 395 degrees, and ran autotuning. It got up to temp after fiddling with the oven settings, hovered around the set point for a bit, then switched off and displayed 625 over 0.0. I assume that means autotuning was complete. Turned the controller off then back on to run the program. It got up to about 396 degrees and hovered around that temp until I switched it off. So is the idea to slightly adjust the hysteresis setting, in my case to about 398-399 degrees, to get closer to my temp of 400?

About the oven settings. The only way it would get to temp is if it was set to broil, which only uses the top two heat elements, not the bottom two. Not sure if this will be an issue with even heat distribution, or is that where the metal plate/fire brick comes in? I had it on the highest bake setting (450) at first but it would only reach about 370ish degrees, because the oven was controlling the temp. And the toast setting would only work if the toast timer was on.

So, overall...success!
 
First, set the oven to whatever is the hottest setting. The PID will turn it on and off from there.

If the oven is digitally controlled, that is probably your trouble. If this is the problem, you may have to bypass the ovens controls and wire the coils directly to power cord. Then, the SSR will turn them on and off directly. If they are being controlled by the toaster oven controls the program may reset every time the PID shuts off the power. A quick test for this is to plug the toaster oven into the wall socket and set it to BROIL . Let it get hot for a minute, and then unplug it from the wall socket. Plug it back in. If it has reset to OFF, then your oven will need to be re-wired to run on the PID. If it heats back up with no buttons being pushed, it is an analog controller, and should work fine with the PID.

Also. IIRC, the hysteresis program setting for these PID controllers is a +/- setting. If you want it to run between 395 and 405, the hysteresis setting would be 5 degrees. A hysteresis of 395 would try to run an oven set for 400 in swings between 5 degrees and 795 degrees.
 
It's an analog toaster oven. In this case BROIL is the hottest setting. But again, it is only utilizing the top elements, and not sure if it's an issue for heat treating a knife evenly. If it's not that big of a deal, considering I'm using a kitchen appliance, we're in business. It's only when in BAKE that it won't hit temp because of some internal switch controlling the elements/temp.

I set the hysteresis to 5, and it hovered between 395 and 397. So I'm guessing I need to set it to maybe 1 or 2 so it's closer to 400, my set temp.

Sorry, I'm bad with electrical stuff. Surprised I even got as far as I did wiring the controller, and didn't kill myself in the process.
 
Last edited:
Give it some time to settle down. It should zero in on 400 +/- a degree or two once the learning is done. You can run an auto-tune again to see if that helps.

I don't know why it isn't passing 400. As an experiment, try setting it to 410 and see what it hovers at.

Make sure your hysteresis range is 395 to 405 or is a +/- 5 degrees. It sound like yours is set to range between 395 and 400 right now.

As a final experiment, try setting the target at 403 and see where it levels off.
 
So I double checked to make sure the hysteresis was set to 5. Set the temp to 410, and ran autotuning. Once autotuning was finished I let it run for about 20 minutes. It was giving me readings between 407 and 408. Then I set the temp to 403, autotuning, and ran it for 20 minutes or so after autotuning finished. It hovered between 399.8 and 400.9. But, while autotuning it does hit the set temp. It's only after autotuning is finished that it hovers around those lower temps. It might be some setting I'm missing or screwing up. In hysteresis it appears I can only plug in one number. Should I be able to put in, for example, a +5 and a -5 setting, or is it just 5?
 
OK, then just set the target at 402 and it should hold at 400 with a hysteresis of 5.
 
Great! This should give me a starting point to read up and learn more about heat treating, using a pid controlled toaster oven.

One other thing, my oven has slots for racks above and below the middle rack where I have the thermocouple, and will be putting the blades. So there are three slots for racks. I was wondering what would be the ideal setup as far as steel or aluminum plates would go. Will probably go with steel at first because of cost. Would leaving the grate rack in the middle (with blade), and plates above and below it be ideal? Or maybe all three metal plates? Or just like you suggested, a plate for the knife and a plate above to block direct heat?
 
If you are only heating from above, just a 1/4" or 1/8" plate of steel sitting on the upper rack would be sufficient to block direct radiant heating. Make the plate so it has about 3/4" to 1" clearance all around for convection. Additionally, the thermal mass of a second plate below the blades would be a good thing.

The plates will add a good bit to the initial heating time, but should stabilize the temperature swings and make for the blade getting very even heating.
 
Just an update.

This weekend I added a piece of 1/8 inch plate steel between the heating elements (at the top) and where the blade(s) will be sitting on the rack. As well as a thinner piece of steel below the middle (blade) rack. It took a little over an hour just to reach 350 degrees, and struggled to get beyond that. It took another 20 minutes just to hit 352 degrees. So it looks like that set up isn't going to work.

For now I will just put the heavier plate steel below the rack and have nothing above it. I will also try putting the blade directly on the plate steel, and not use the wire rack at all. As you suggested in you tutorial. At some point I'll have to look into wiring the controller directly to the coils to be able to utilize all four heating elements, top and bottom.
 
You must log in or register to reply here.
Back
Top