Temperature oscillations. Normal?

Hi,
I received my rotodamper and case (Thx Ralph) and tried a long run test with no meat load. Despite an overshoot at the beginning, the temperature was pretty consistent. The only thing that caught my attention was the hyper-activity of the whole setup trying to keep the target temperature.
Someone could help and tell me if this behavior is something expected or it means that the HM is having hard time keeping the targeted 220F?
My smoker is an Akorn Kamado, which has a fairly high temperature efficiency. Note in the graphic that rising the temp looks easy but lowering it takes more time…
PID setup:
B=4; P=3; I=0.005; D=5
Fan Max=40%; On at max only
I used a can in the middle and a ring of lump charcoal around. Only ignited one little spot in the charcoal ring and let the HM to take control from there. The average temperature was Ok, but those oscillations got me a bit worried. May be they are normal?
Thanks anybody who could “bring some light” 

https://onedrive.live.com/redir?resid=3EFE861449B7E531!768&authkey=!AK9-u7D118RshGc&v=3&ithint=photo%2cjpg

I would say you've got air getting in from somewhere. When the output is 0% and the rotodamper is fully closed, your temperature should drop pretty quickly, not take an hour to go down 10 degrees. As a test, you can run with the HeaterMeter connected, then when you're about 10 degrees too high, remove the rotodamper/heatermeter and fully close the vent you had it installed on. Observe how long it takes the temperature to go down.

If it still takes an hour for the temperature to go down 10 degrees, figure out where you're getting extra air into the grill. If it drops pretty quickly, figure out where the rotodamper installation isn't sealing well.

Because your whole system reacts so slowly, you will want to adjust your PID coefficients once you've got that figured out. Fast response, low P (3 like you have). Slow response, high P. I'd double it as a starting point.

Thanks a lot Bryan!... I'll check for leaks, run additional tests and get back here with the results...

I am getting the same results with my Broil King Keg (Bubba Keg). Takes forever to cool down, due to the crazy insulation. I am getting closer on the tuning though. Saturday's cook, with it running the same way (fan on max only) I found it ran better with a higher P and D. How did you have your exhaust vent set? I have to run mine at 1.5 (markings of 0-5) so 30%. I also put aluminum foil tape on the intake to seal it 100%, so the only intake air is going through the damper.

B=4

Bryan posted somewhere that B should now be 0, it was a 'fudge' term that you don't need. Bryan can confirm?

I have better results with the old presets (4,3,0.005,5) than the new. Bryan is right about the temp dip, when the RD is closed the temp should drop much more rapidly than what is shown on your graph. As Bryan said, seems like there may be air leaks somewhere if the temp is that slow to drop. On the Char Griller Kamado there may be cracks in the edge around the ash pan, or under the handles, and sometimes the clips do not pull the ash pan up tight enough to seal around the gasket, and some Akorns didn't have the vent sealed to the ash pan very well (or at all).I find the top vent setting can have effect the oscillations, recently I had a nice stable cook going (pretty much flat line on the target temp), about half way through the cook I opened the top vent about twice as wide as it had been and my temp started to oscillate. So do not ignore the top vent setting, I tend to adjust that on a per cook basis depending on the fire size, whether I am using a water bowl or not, the amount of food in the cooker etc. Also make sure the RD is opening and closing completely from 0-100%, if not the HM can get very "confused". Also make sure the RD is rotating smoothly from 0-100%, if it is assembled too tight or the wires put too much drag the servo may struggle to rotate the RD and it may stick for a bit and then jump a good distance when it finally moves, this can cause oscillations like that. When you first press the RD together (without the screw installed) rotate it back and forth and note the force it takes to move the servo, after you tighten the screw rotate it again, if it is now tight and harder to rotate then loosen up the screw a bit until it moves freely like it did before you tightened the screw. Don't worry about that screw coming loose, those parts don't move (relative to each other) so the screw wont work loose even if it is not cranked down....

Hi Mike,
Aluminum tape is a great idea to expedite the test before the leaking search task. I'll try it and play with a higher P as starting point...
Related to the exhaust vent, I would say that mine was 20% or less open during the run.

Thank you Ralph. The RD is working awesomely right. It rotates smoothly from full close to full open and the limits are just when it reaches the "full" positions open or close. I tested it before mounting and closes the air intake very well (blowing with my mouth instead of the blower, no air passing when closed).
So, my plan is to seal all the possible air leaks with aluminum tape and run a new test with the original parameters (4,3,0.005,5). If it works better, oscillations can be attributed to air leaks; if not, I'll play a bit with the PID parameters (higher P and D?) to see what happens.
Thanks a lot for your help gang!

Another question for Ralph... How open do you set the top vent in your Akorn? I got it open 2 or 3 times the size of the small holes at the beginning of the vent wheel but not sure if that is right.

There really is no "right" setting for the top vent. I find it depends on the fire and cook, target temp, whether I am doing a water smoke or dry, how much food, how well the rest of the grill is sealed up etc. For low and slow I generally run from a small crack wider than the dimple to around 2-3 times the size of the dimple, for mid to high heat cooks I usually run about a quarter to half way open on the top vent.
For low and slow a good way to go about it is to light a small section of coals and let them burn a little bit, then close the lid with the top vent cracked just a bit more than the dimple, this will help minimize the initial overshoot and get your pit temp stable ASAP. From there watch the RD position (or HM graph) and if you find the RD running more than 50% open most of the time you could crack the top vent a little wider. I tend to like to run the top vent as open as possible and have the RD running in the 30-50% position the majority of the time. If you have the top vent too tight and the RD is running near wide open when it hits 100% and the blower kicks in you will get a temp dip and then overshoot (due to the sudden increase in air flow), so try to avoid that. I have been encouraging Bryan to change the fan config for the HM instead of just "On at Max Only" we could set a range where the fan would operate, like "Fan on above XX", and have the fan speed span the entire range from the start point to 100%. I think if the HM could be set that way it would be much easier to have one config that would work well for low and slow as well as mid to high temp cooks. For instance, if the fan were to come on at a trickle pace at 75% and ramp up from there it would not cause a massive change in air flow suddenly like it does in "Max Only" mode. As it is now I generally change my config for low and slow VS mid/high heat cooks, and try to keep the RD running less than 75% open during low and slow cooks to avoid the fan from coming on and causing the dip then overshoot. For mid to high heat cooks I generally let the fan blow all the time.

Mike Wallace said:

I am getting the same results with my Broil King Keg (Bubba Keg). Takes forever to cool down, due to the crazy insulation. I am getting closer on the tuning though. Saturday's cook, with it running the same way (fan on max only) I found it ran better with a higher P and D. How did you have your exhaust vent set? I have to run mine at 1.5 (markings of 0-5) so 30%. I also put aluminum foil tape on the intake to seal it 100%, so the only intake air is going through the damper.

B=4

Bryan posted somewhere that B should now be 0, it was a 'fudge' term that you don't need. Bryan can confirm?

Click to expand...

Mike, i have a the Big Steel Keg using the follow setting for a long time
B=0
P=12
I=.004
D=7

I use dampener and no fan.

For low and slow 225 keep top set at 1.5 to 2.5 depending on wind. If going hot 275 or up I open it up to about 3. I only use lump, large chunks on bottom, small on top. I light it with a small piece of starter in the center on near top of pile. I set temp about 25 degrees low wait for it to hit that then increase temp to target. If it not too windy the temp holds about +- 3 degrees, for 12 to 20 hours.


Dale

Cool thanks Dale. Lump only here too. I'll try your settings next run.

Ok... here is what I did...
1. Sealed all the possible air leaks with aluminum tape. Ash pan and bottom damper...
2. Around 7:30 PM – 8:00 PM, started a small fire on top of the fire bowl loaded with charcoal...
3. After a few minutes, when the fire was ignited, started the HM with B=4, P=3, I=0.005, D=5...
4. It took a bit more than ½ hour to reach the target temp, overshoot and get stable @ 220F
5. The Kamado ran all night with minimum oscillations around the target temp of 220F, going up and down 2 to 4 grades but keeping the average temp...
6. At 9:45 AM, next day; I changed “P” to 6 instead of 3...
7. Almost immediately started to oscillate up to 10F from target...
8. After 2 hours running with P=6, I returned to P=3 and the oscillations decreased...
9. The test ran for 24+ hours with only one initial load of charcoal. I think that after the change on “P”, turned very difficult for HM+Kamado to get a steady temp, may be because of the fuel getting low? Or, it may be because of the “confusion” the change on “P” produced in the system?
10. Anyways, it ran for 24 hours and despite the oscillations due to the P increase, the whole system kept an acceptable (for me at least) average temp with small deviations up and down lower than 5F...
So, the initial struggles to keep a steady target temperature seem to be related to air leaks in the ash pan and the bottom damper...
As always, experts’ diagnose will be extremely helpful 

Here is the 24 hours graph...

https://onedrive.live.com/redir?res...authkey=!ACsJr2SAJ6vhmYw&v=3&ithint=photo,jpg


Only to take advantage of the burnt charcoal, next test run will have a pork shoulder on the grill... Is "P=2" a good idea?...