Fine tuning PID

[Arcs_n_Sparks]

Mr. Walleye,

Not bad. The P term "amplifies" the error term (setpoint - current temp). The higher the P value, more correction (in this case power up or down) is applied for a given error.

For a number of reasons, Proportional control will not drive you to zero error. So the Integral term is used to sum up these small residuals errors and help drive it to zero. However, since it sums previous errors (which can be quite large if you are far away from setpoint), it must be used with care and is usually bounded.

The derivative term is used to "anticipate" what to do upon a change in setpoint. Higher values cause more or less power to be applied quickly.

He is a link with some more information: http://www.htservices.com/Applications/Process/PID2.htm

Arcs_n_Sparks

[smokin stu]

It;s supposed to go above freezing here next week, and I will try out my PID then.  This waiting around for nice weather to try out my new toy is driving me nuts.

Reminds me of when I got a new bike for Christmas one year and had to wait months for the snow to melt so I could try it. :'(

[West Coast Kansan]

The new bike for me was 4th grade. Did not get to ride it until my birthday in march ... was a killer.

When you tune the pids do you have to have a puck burning also. Just wondering if the profile is effected by that type of thing. Not sure if a smoldering puck changes things any?

[Mr Walleye]

The first 2 times I ran auto tune I did it with the SG on but no pucks and the tower was empty. Both times the numbers ended up almost the same. Then on the following weekend I was doing a couple of butts and decided to put the PID into auto tune mode. The result was my current set of numbers which are different from my other 2 attempts. The PID seems to be pretty much spot on now. Once in a while it will flash + or - 1 degree. I remember reading BubbaGrumps post about using a couple of bricks in the tower to simulate a load. I think when running the auto tune this is fairly important.

Mike

[smokin stu]

Well, I ran the autotune on the new PID today.  The first time was with the smoker shack door open and I got a PID or 472, 215, 279 respectively.

I then re-ran the autotune with the smoker shack door closed as I would normally smoke and I got 0, 185, 0........  Not being sure what the heck that meant I redid the autotune a third time with the shack closed up and got 362, 9250, 147.

I am not sure what the heck is happening here.  The temp looks good and I guess that is what counts.  I had water in the BS but no food.  I think I will go with it and when I do a butt, I will wait past the first two or three hours and then run the autotune then, with the butt in and everything warmed up.

The BS door thermometer reads 195 degrees, the ET-73 says 220 degrees, and the PID says 215 degrees, which is where I set it.  I think I will bring a thermister home from work that I know is accurate and then calibrate the actual temp on my PID with that.

Tell me, if I bought a new digital BS, this is all built in, no?

[Mr Walleye]

Here is a link to my experiences with the temp swing on the DBS. Tipically I would see a swing of 189 to 207 with a target temp of 200. I guess out of the box you do have the ability to control the temp enough feel reasonably ok to leave it un attended. Although lots of people here feel the same way about the OBS.

http://forum.bradleysmoker.com/index.php?topic=4740.0

Mike

[LoneStar]

I just put together a new pid and autotuned it.  Once its autotuned, do you have to autotune it every time you change the set value?

Rick

[Mr Walleye]

Hi LoneStar

No you don't have to. The only time you would rerun the auto tune is if you change something that influences the ability of the unit to heat the tower. Like adding a circulation fan or something like that. As I have indicated earlier I have ran auto tune several times and recorded the numbers, the most recent was while doing a butt. This setting seems to be the best for me so I would recomend running the auto tune with a load in. JMHO

Good Luck
Mike