Oddly, it looks like the data are, shall I say, “different” upon updating to 0.4.2 (I’m using a BrewPi Spark too)… I updated my BrewPi/BrewPi Spark during fermentation and I see the following before and after the update w/ the same beer in the same fridge & same room:
The temperature data reported are less granular than before. In other words, it looks like the data are being reported with less sensitivity than they were while I ran 0.2.11
The fridge temperature now averages ABOVE the fridge temperature that’s being reported…
–> I understand that the algorithm was updated in 0.4.X, but the data still violate the 1st law of thermodynamics…
I’ve attached screenshots of some sample charts before & after the update from 0.2.11 to 0.4.2. Does anyone have any input on this one
Since 0.4 all plotted temperatures are unfiltered: it’s a raw temperature directly from the sensor. That is why it looks more grainy. The input to the PIDs is still filtered.
My guess is that your beer is close to the back of your fridge, so it is cooled directly instead of via the fridge air. Which is why your fridge air can be higher than the beer temp. Obviously, it is cooling the right amount to keep the beer temperature within 1 bit of the setpoint.
Ahh, I see. Is there a way (and/or would it even make sense) to plot the filtered instead of the unfiltered data?
Hmm, that’s a good point regarding beer being towards the back of the fridge… It’s possible that I moved my fridge temp probe vs. where I had it before the update (I just let it dangle in my kegerator lol). I’ll play around w/ it a little
Filtering also creates a delay, so by showing the temperatures unfiltered, you have a representation that is closer to reality.
Showing the filtered temperature is not possible now, but later we will make the graphs much more configurable and you could for example plot the PID input temperature (which is the filtered temp).
