Tuning For Sous-Vide Control

[Brett]

#3529

I’m really enjoyed this post on the diy-pid-control google group. I thought that would be a nice way to start off this sub-forum.

[chizuoka]

#3694

Hi,

I have been meddling with the OSPID and various sous vide setups ranging from twin heater elements, to slow cooker, to the best i found so far, a deep fryer surprisingly. Characteristics are that it holds its temp really really well because of good insulation, it heats up extremely fast, and it has very little moisture loss when closed. However, it continues to heat up even when power is off due to the fact that its heating element is so massive, and autotune just refuses to work, it brings the water all the way up to a boil. Can anyone give me some advice what would be nice params or help me to get to some nice params?

If we see this image

The autotune brings heats up the water very linearly until boil… this was the second round already so seeing it was about to boil its head off i killed autotune and manually turned off output to 0. Why doesnt autotune want to work?

If we take a look here

The residual heat continues heating up the water. What settings would let the PID know that the water will continue being heated for X amount time before it stabilizes?

We can see the insulation is quite good.. the temp remains quite stable without additional heating, so from a logic point of view the PID should heat up the elements just a bit.. and let residual heat from the element heat up the water to set value

I tried a guesswork value of P=1, I=0.5, D=0.5 and got an average overshoot of 1.3C and under value of 0.2C. This is when the water was already around the preset temp.

I read through this but dont get a few points…
https://groups.google.com/forum/?fromgroups#!topic/diy-pid-control/0bswXHDhJAs

What did he mean by “Choose a Kp such that your output will greatly exceed your maximum
output limit until you get close to the set point”. Whats the “maximum output limit” that he is talking about?

Btw im using an SSR. So i have a output update time of 1s. Do help.. anyone! thanks so much!

[Brett]

#3695

So the PID algorithm is linear. that means that it works best with the process behaves the same going up as it does going down. if it’s not the same it’s not usually a big issue, but if rising and falling are REALLY different (as they are in your case) then you have a problem.

put another way, the pid is adding heat to the process assuming that as it gets closer to setpoint it will be able to push the process downward as fast as it could push it upward. from a control perspective, you’d be better off getting rid of that insulation!

with the process you have, you’ll probably have the most luck with a controller that just drops the output to 0 the moment it gets anywhere above setpoint, and that means having a big Kp. If I had to re-phrase the “max limit” thing from that post, I’d say that Kp needs to be bigger than the output span (max-min) which is 100 by default.

[chizuoka]

#3696

Hello Brett…

Thanks so much for replying. Lets take a look at the characteristic of my system…

I tried ranges of P from 5 to 100… and 100 does help alot… should i go 1000? would there be a difference from 100 as i dont see much difference from 100 or 50…

Then this last graph shows the different responses. one when start point is far from set point, one nearer to set point, and the stable one…

We can see that even at stable, just a little bit of output produces nearly 1C of change, while from far the overshoot can be as high as 10C, what would be nice numbers so that the system could behave something like
– when approaching set limit, dial down the power gradually
– when applying power, longer wait time to monitor the increase in temp before applying power again. (With P so high, the power applied seems to shoot all the way 100 to 0 even when very near to the setpoint. )

Thanks so much for your advice…

Chizuoka

[chizuoka]

#3697

hmm.. it seems need to open images in new tab or download as its too wide for the screen to see it all… ^_^

[chizuoka]

#3698

whooops… anyone have any tips on suitable PID numbers for this sort of behaviour?

We can see that even at stable, just a little bit of output produces nearly 1C of change, while from far the overshoot can be as high as 10C, what would be nice numbers so that the system could behave something like
– when approaching set limit, dial down the power gradually
– when applying power, longer wait time to monitor the increase in temp before applying power again. (With P so high, the power applied seems to shoot all the way 100 to 0 even when very near to the setpoint. )

[Cracked]

#3699

Is your sous vide stirred at all? I’m guessing not and I suspect that the heating you see after the output falls to zero is actually from the fluid temperature evening out (that is to say, the temperature you are measuring is not the mean temperature of the fluid), as opposed to the residual heat of the element.

Can you move the temperature sensor closer to the element? (Not too close though).

[chizuoka]

#3700

its the residual heat from the heating elements. I have a pump inside to circulate water. Anyway, I have abandoned this idea of using the fryer because of the issue of residual heat.

I have found that using an immersion heater is still the best option. Minimal residual heat and fast to start. However i do have a question, using my immersion heater, i threw it a number of PID 50 : 0.5 : 10. Im curious.. what does the I and D value do, there seems to be no effect to the system whether its at 0 or 10.

With this setup im able to get a typical +- 0.5C but sometimes can swing to +-1C for short moments.

[Author]

Posts