I've been using a converted mini-fridge as my fermentation chamber for years. It's about 18"x18"x30" high, I built a simple 5 sided-box out of 3/4" thick insulating board (that's a box with one side open) that I butt up against the fridge (fridge door and hinges are removed). The box is the same width and height as the fridge opening, and about 30" long, so basically I just increased the inside area of the fridge. Nothing but 1/2" weather-striping to seal the box where it meets the fridge and just pressed against it. (Surprisingly, the fridge runs about the same amount of time as before the mod, even in my FL garage)
My simple ebay controller (which has programmable hysteresis, but I left at the default 4 degrees F) uses a temp probe measuring CHAMBER AIR temp (not wort). So the fridge air stays right within the setpoint and setpoint+4 degress F (or you can call that setpoint+2 degrees +/- 2 degrees)
I measure the wort temp using a Tilt and also a separate temp probe - NEITHER of which are in the fridge control loop. My wort temp never fluctuates by more than 0.5 degree.
Yes, the wort runs warmer than the air temp by some amount, which is a function of the total thermal mass of the wort, vs the air volume of the fridge, and how active the yeast are. After learning from two batches on my system, I am now able to account for that difference in the first two days of fermentation (dial down the fridge by 3 degrees in my case), and keep the wort exactly where I want.
My point is this simple air control loop keeps my wort exactly at temp (+/- 0.5 F) perfectly.
Controlling the enviroment appears to be the best way to regulate wort temp. Makes total sense given the thermal mass and temperature lag of the wort vs the chamber air. Off the cuff I would expect something like the ratio of (mass of air)/(mass of 12 gallons wort) as the reduction in hysteresis between the controller and actual wort fluctuation.
So why in the world is there so much added complexity in "predictive systems" trying to prevent overshoots, which only exists when you control fridge temp by measuring wort temp instead of air?
Hats off to BrewPi and what they have done for homebrewing (eBrewing/mashing needs their technology more than fermentation). But the plots I have seen for chambers controlled by measuring wort temp show about +/- 2 degrees variation in the wort temp. About 2X to 4X more than I am getting with a simple ebay controller measuring air temp.
I bring this up, because I am working on a electric brewery controller and I see the same sort of overkill when talking about cycle-based triac control vs simple on-off duty cycle via relay. Again, many seem to ignore the fact that the thermal mass simplifies the problem and allows for much simpler control loops (as compared to driving a router motor or IV pump, as examples).
My simple ebay controller (which has programmable hysteresis, but I left at the default 4 degrees F) uses a temp probe measuring CHAMBER AIR temp (not wort). So the fridge air stays right within the setpoint and setpoint+4 degress F (or you can call that setpoint+2 degrees +/- 2 degrees)
I measure the wort temp using a Tilt and also a separate temp probe - NEITHER of which are in the fridge control loop. My wort temp never fluctuates by more than 0.5 degree.
Yes, the wort runs warmer than the air temp by some amount, which is a function of the total thermal mass of the wort, vs the air volume of the fridge, and how active the yeast are. After learning from two batches on my system, I am now able to account for that difference in the first two days of fermentation (dial down the fridge by 3 degrees in my case), and keep the wort exactly where I want.
My point is this simple air control loop keeps my wort exactly at temp (+/- 0.5 F) perfectly.
Controlling the enviroment appears to be the best way to regulate wort temp. Makes total sense given the thermal mass and temperature lag of the wort vs the chamber air. Off the cuff I would expect something like the ratio of (mass of air)/(mass of 12 gallons wort) as the reduction in hysteresis between the controller and actual wort fluctuation.
So why in the world is there so much added complexity in "predictive systems" trying to prevent overshoots, which only exists when you control fridge temp by measuring wort temp instead of air?
Hats off to BrewPi and what they have done for homebrewing (eBrewing/mashing needs their technology more than fermentation). But the plots I have seen for chambers controlled by measuring wort temp show about +/- 2 degrees variation in the wort temp. About 2X to 4X more than I am getting with a simple ebay controller measuring air temp.
I bring this up, because I am working on a electric brewery controller and I see the same sort of overkill when talking about cycle-based triac control vs simple on-off duty cycle via relay. Again, many seem to ignore the fact that the thermal mass simplifies the problem and allows for much simpler control loops (as compared to driving a router motor or IV pump, as examples).
