Freezer cooling, heating, and duty cycles

Today I finally installed my BrewPi sensors inside the freezer, although just for monitoring and not for controlling. I’ve been fiddling with compressor delay time and differential (STC-1000 controller) because I was worried based on my first BrewPi graphs that the freezer seemed to be running a lot.

This is why toward the far right of the graph you see longer cooling and warming cycles. But the thing I’m most puzzled about is how fast the freezer temp starts to climb once the compressor shuts off.

It’s a pretty new freezer, not a garage sale special, leaving me with these as the best explanations:

  1. There’s nothing in the freezer other than a 5-gallon batch of beer. No frozen meat, etc. that normally inhabits a freezer of this size to provide thermal ‘inertia’. In fact nothing in it is actually frozen.
  2. The temp probe for the STC-1000 and the two temp probes for the BrewPi are making their little indentations on the gasket / seal around the door. So there is likely a bit of cool air escaping from there; I don’t have a perfect seal.

But I’ll ask the community: does a freezer warming up this fast seem normal? (nearly 20 degrees in about 90 minutes). We’ve all experienced electrical power outages and probably not seen our freezers climb like this.

if the temperature outside the fridge is 20 degrees warmer than the inside, then yes it will equalize pretty fast. thats why there is also a cooling cycle. is your freezer outside or in a hot garage or something?

The current set point of the fridge is 50F, and the freezer is located in a finished basement where the ambient temp averages about 67F. So yes, it’s nearly a 20-degree difference outside the freezer vs. inside, and about to increase to 35+ degrees since my aim is to cold-crash.

I’d do a test with a profile and some water to see if it’s able to hold temps on your beer. You may be surprised with the outcome after you get about 5 gallons of water stabilized in there.

I am more puzzled that you don’t have a fridge setting line, beer setting line and actuator activity in your chart.

Can you reset your settings?

echo -ne 'E\n' > /dev/ttyACM0
1 Like

I reset the settings based on Elco’s recommendation, and have the following to show for it. I am currently stabilizing at 40F with one last push down to 35F coming later today.

You can see a gap in the logging where the device was reset, and I set beer/fridge settings at 68/67 respectively so as to separate their plots (they used to both be at 68).

I find it interesting that the warm-up of the chamber exhibits a logarithmic shape; initially the climb is pretty rapid but then as the chamber temp approaches the temp of the thermal mass inside the fermenter it rolls over to the more linear / flatter shape I was expecting, until the next cooling cycle of course.

Let me know if the hardware reset accomplished what was expected. I wasn’t anticipating actuator activity on the plot since the software is in the “Off” mode and I’m only monitoring temps (the STC-1000 manages the freezer cycling). My Spark still shows only the hardware test screen, with two temp sensors displayed + the BrewPi logo, but I’ve come to expect that.

Below are two graphs. In both cases the BrewPi is being used only for logging of temperatures, while an STC-1000 handles the freezer cycling. The 2nd graph is the result of increasing the delay time and differential (F2 and F3). (I am not fermenting, just keeping kegs ~38F).

In the 1st graph the freezer cooling periods are about 4.5 minutes long, after which the freezer idles for about 13 minutes. Therefore the freezer duty cycle = 25%.

In the 2nd graph the freezer cooling periods are ~40 minutes long, with the freezer then idling for 4+ hours. In that case the freezer duty cycle is about 13% but the outer walls of the freezer are definitely warmer to the touch during the cooling cycles.

Q: which kind of duty cycle is better for the long-term life of a typical freezer compressor?

p.s. I trolled homebrewtalk / etc.trying to answer this question. Didn’t find a satisfactory answer.