Prevent overshooting

Michel Kraaij

New member
Any tips on how to prevent these overshoots?

1599300849303.png

The first few spikes aren't even that high (and maybe I just needed a but more patience). The last one is after I opened the lid.
My current PID output parameters (on Celcius) are:

1599300778683.png

Oh, a quick addition: Even with these spikes everything did stabilize without intervention.
 

GSpinelli

New member
Try lowering the fan max (start with 60%). I would guess the fire gets stoked up too much. Also sealing any leaks helps. What type of smoker, and how is the blower/servo attached?
 

Bryan Mayland

TVWBB Hall of Fame
Yeah if you're using something with a Damper, I also highly recommend "Fan on above" set to something like 50% and "Servo fully open at" whatever number you pick. This way the Damper goes first, then only uses the fan if it needs to and cuts out earlier as it approaches temperature.

The last overshoot is contributed to control coming back online while the temperature is still very low after the lid was open-- at the start of the temperature recovery which makes the temperature seem lower. You can raise the length of the lid open duration, or I actually reset my lid timer (left button twice) right as I close the lid to give the temperature enough time to equalize and get a true reading before the control algorithm kicks back in. I'd rather have the temperature be low for a couple extra minutes rather than have it overshoot, which takes longer to recover from.
 

Michel Kraaij

New member
Yes, using an Adapt-a-Damper. Okay, trying now:
1599306968699.png

And the last overshoot... I have the nasty habit of disabling the 'lid open' whenever I have closed the lid. If I understand you correctly that's something I should not be doing and let the HM correct itself. Right?
 
Last edited:

Bryan Mayland

TVWBB Hall of Fame
Looks good with those numbers, Michel. Yeah the "lid open" period should encompass both the time the lid is open, and the time for the temperature to recover and stabilize.
chicken-51-400.png

I closed the lid at the lowest part of the dip, 200F. If the lid mode turned off there, HeaterMeter would go FLANK SPEED AHEAD to try to get back to 400F and get a lot of overshoot from the extra blower action. Instead, see how the temperature comes back up and recovers to ~340F and then HeaterMeter just takes it up from there when the lid mode ends. Ideally, the software could detect when the temperature comes back up after the lid closes and end lid mode early, but it is only 30 seconds or a minute max and I think it would be harder to detect than I'd think:
  1. Detect entering lid mode
  2. Wait for a "Low Peak" event while still in lid mode
  3. Watch for a period of increasing temperature with a greater gradient than just normal fluctuations with the lid open
  4. After temperature increases, wait for it to level off
  5. End lid mode early
#3 is where I think we'd see a lot of problems, considering that most lid events aren't perfect like my example. #4's problem is that it takes time to detect an increase (due to how the trailing average is calculated) so it might not be able to detect the leveling off before lid mode would normally end.

I'm just sort of thinking out loud here, as the lid mode is (waggles hand) useful but not very smart and I'm always keen on trying new things to make the system better.
 

Gary V

TVWBB Fan
Michel,
Nice looking graph. What kind of grill are you using. A Couple of things here to add. The control is slow, so biggest mistake I see out there is trying to tune the control to be fast. Your current setup looks good. I use a Primo Jr grill which is on the smaller end of the ceramic grills out there. Over a few years I have learned the following:
1) Leave the default lid setting where they are. They work good and overshoot does not happen. You always want to come back to temp slow and steady
2) Make sure you do not have pit probe too close to flame. My probe hangs down from top center of dome and is close to where the mechanical probe is in the dome.
3) NO LEAKS. Fix gaskets and make sure there are no leaks
4) Use the highest quality temp probes you can. Maverick probes do not fit well in probe jacks, so I do not use them because of that.
5) Make sure the damper can move easily. Do not tighten screw so much that the damper cannot move easily.
6) Over time I have got my config setting so the fan is used when I need a little temp or heating up, but the damper pretty much does the work after that.
7) Last is to use the exhaust damper of the grill to fine tune. As temp raises, I open my damper a little more to keep the control steady. My damper is usually only open 2.5-6mm, depending on set temp.

adding my current config and last cook. Opened lid a few times to turn corn.
 

Attachments

Michel Kraaij

New member
Michel,
Nice looking graph. What kind of grill are you using.
Thanks! I'm using a Bastard Large, a dutch brand kamado.

IMG_20200822_101242751.jpg

A Couple of things here to add. The control is slow, so biggest mistake I see out there is trying to tune the control to be fast. Your current setup looks good. I use a Primo Jr grill which is on the smaller end of the ceramic grills out there. Over a few years I have learned the following:

1) Leave the default lid setting where they are. They work good and overshoot does not happen. You always want to come back to temp slow and steady
True. However, the default settings were for fahrenheit instead of celcius. So some adjusting was mandatory. ;-)
2) Make sure you do not have pit probe too close to flame. My probe hangs down from top center of dome and is close to where the mechanical probe is in the dome.
The pit probe is at grill level and I use a heat deflector.
3) NO LEAKS. Fix gaskets and make sure there are no leaks
My Bastard is equipped with fiberglass gasket. Haven't discovered any leaks so far (except for a very small one where the pit probe goes through. I know, I can run in through the top vent. However, I noticed hardly any difference in leakage.
4) Use the highest quality temp probes you can. Maverick probes do not fit well in probe jacks, so I do not use them because of that.
Agreed. I had a set of GrillEye Pro probes lying around, which seem to do pretty fine. However, I didn't use them this time because I'm doing ribs right now. So a 6-hour slow burn is fine for now ;-)

The pit probe is a low-budget thermocouple.
5) Make sure the damper can move easily. Do not tighten screw so much that the damper cannot move easily.
It runs as smooth as a baby bottom! ;-)
6) Over time I have got my config setting so the fan is used when I need a little temp or heating up, but the damper pretty much does the work after that.
I think the last adjustment Bryan suggested fixed that for me. The fan is pretty much idle right now. The damper seems to do all the work, which I think is good!
7) Last is to use the exhaust damper of the grill to fine tune. As temp raises, I open my damper a little more to keep the control steady. My damper is usually only open 2.5-6mm, depending on set temp.
The top vent is almost closed. Whenever there is hardly any airflow the temperature starts to drop, which I think is a good thing (If i read all the threads correctly). That's when the HeaterMeter kicks in.
adding my current config and last cook. Opened lid a few times to turn corn.
Thanks! Almost done with this 6-hour run. Time to eat!! ;-)

IMG_20200905_163256955_HDR.jpg
 
Last edited:

Tom O.

TVWBB Member
The top vent is almost closed. Whenever there is hardly any airflow the temperature starts to drop, which I think is a good thing (If i read all the threads correctly). That's when the HeaterMeter kicks in.

Just a word here about the top vent. On MOST smokers no matter the type, having the top vent open as much as you can while still maintaining the temperature you want is MUCH better from a smoking flavor point of view than closing the top vent down hard. For the best flavor you need a good flow through of air. Running with top vents shut down has the tendency to make the smoke taste like creosote or more acid-like.

Tom
 

Michel Kraaij

New member
I agree, Tom. Next time I'll experiment with the top vent more open (in different settings) and see how the HM reacts.
Nevertheless, the ribs tasted absolutely fabulous last time ;-)
 

Top