Explaining controlling algorithm

Hi all guys,

I have just yesterday discovered this super cool project. As I am electrical engineer, I would like to ask something more about the controlling algorithm behind the project. For this reasin I would like to ask a few questions.

The major goal is to keep beer temperature to set value and as flat as possible (I would say). How is prediction of optimal frisdge temperature done in order to reach wanted beer temperature?

At the begining of the fermentation the yeast will release energy in the form of heat which means that the we need lower environment temperature. At the end of fermentation in the conditioning phase there beer temperature should more closely follow environment temperature.

Should environment temperature be much more prone to changes than actual beer temperature, so keeping beer temperature stable shouldnt be a big problem once the close vixinity is reached?

That would be all for now, but I think I would have like tents of questions.

Thanks for the answers in advance and cheers,
Nemanja

This is traditionally done with a thermal probe and a PID (proportional integral derivative) controller. Whether the brewpi is using a PID algorithm I’m not sure, but that’s the common method to avoid temperature over/undershoots based on thermal mass lag. You’re right that environmental temperature is quicker to change, it’s got to do with the ‘specific heat’ of a liquid versus air. If it’s truly of interest to you, there’s plenty of online reading just a google search away, since most of the concepts are platform-independent.

Yes. As you say the controller keeps the temp lower than beertemp to compensate for the yeast activity.Check my current brew at http://www.sollidensbrewery.se:8888 . You can clearly see that it keeps the fridgetemp at aproximately 2 degrees Celsius below beertemp. But as the yeast activity lowers you are able to see tht this 2 degree differnence lessens as well. When it stops completely it will be nearly identical with beertemp. Give or take difference for where the probes are located.
The PID controller with overshoot prediction is really good. In my case it actually hold the beer at 18 ±0.05 degrees!