Fermentation temperature control using internal probe

I've been playing with a temperature control system I've created. It's basically a microcontroller that reads temperature probes and uses the information to turn on and off a pump / fan system and a compressor to keep things at target temperature.

I brewed up a blonde ale for my first test batch, and the system is pretty solid now (capable of keeping the temperature within a degree of target, according to readings.)

I've now got a 12" stainless steel tube that you can order (pre-welded) here:

Welcome to Brewer's Hardware

I'm wanting to put a probe in one and submerse it in my beer while it ferments to get a little more fine tuned control over fermentation temps.

When a particular brew or yeast has a recommended fermentation temperature, is it referring to the ambient temperature? If I were to try to maintain an internal fermentation temperature, what would be an appropriate strategy?

Thanks to anyone with input.

Thanks for the business!

The appropriate strategy is to try and maintain the internal fermentation temperature of the fermenting liquid since there can be a substantial difference between ambient and fermentation.

Derrin said:

Thanks for the business!

Click to expand...

You're very welcome, thanks for a great service! You did an immaculate job on the probe covers. You come highly recommended!

Derrin said:

The appropriate strategy is to try and maintain the internal fermentation temperature of the fermenting liquid since there can be a substantial difference between ambient and fermentation.

Click to expand...

That is my understanding, and will be my strategy on my next test batch. I'm mostly wondering about the target temperature. If common practice is to keep ambient temperatures for a particular ale at, say, 66, then that ale's internal fermentation temperature can reach as high as 76 (from what I've read). If I'm maintaining fermentation temperature, do I want to keep it at 66? Higher? Do I want a complex schedule of temperatures (which is part of the software design behind my system anyway

While I've got your attention, do you have recommendations for sealing these?

Thanks again!

You should try to maintain the liquid at 66. You'll probably want a significantly lower ambient temp to achieve that number. What are you using to generate your cold air?

Also, what kind of temp sensor are you using?

The temperature probes are dallas 1-wire probes that cost about $5 a pop. I've got 14 of them, I believe, so I can read quite a few temperatures if I want

I've got a compressor i ripped out of an old fridge, which is keeping a small cooler full of water cold. A temp sensor in the water allows me to know when to activate and deactivate the compressor to keep it within a particular range.

Likewise, a temp sensor in the chamber lets me know when to pump that water through a radiator, at which time a fan is also kicked in, keeping the chamber within some range.

The major pieces of the puzzle are:

1) the micro controller (an Arduino) which reads these probes and controls switches to turn the pumps/fan/compressor on and off. The Arduino has an ethernet addon, and listens over the network, doing only what it's told. It reads a temperature from a probe when asked, or turns a circuit on or off when told; nothing more.
2) A Java app I wrote which talks to the arduino over the network. It periodically reads each probe, analyzes it, and turns the various circuits on and off when appropriate. It also records every piece of information into a database; I've got very extensive data collected off my first test fermentation.

Some of the stats I'm particularly interested in collecting and analyzing are the relationship between the coolant temp, ambient temp, fermentation temp, and fermentation phase. I'm hoping to learn enough about these relationships to keep the internal fermentation temperature within a degree Ferenheit.

I've taken quite a few pictures, I've been meaning to write up some of the things I've discovered in the process. Here's a pic of the design for the cabinet:

stephelton said:

The temperature probes are dallas 1-wire probes that cost about $5 a pop. I've got 14 of them, I believe, so I can read quite a few temperatures if I want

I've got a compressor i ripped out of an old fridge, which is keeping a small cooler full of water cold. A temp sensor in the water allows me to know when to activate and deactivate the compressor to keep it within a particular range.

Likewise, a temp sensor in the chamber lets me know when to pump that water through a radiator, at which time a fan is also kicked in, keeping the chamber within some range.

The major pieces of the puzzle are:

1) the micro controller (an Arduino) which reads these probes and controls switches to turn the pumps/fan/compressor on and off. The Arduino has an ethernet addon, and listens over the network, doing only what it's told. It reads a temperature from a probe when asked, or turns a circuit on or off when told; nothing more.
2) A Java app I wrote which talks to the arduino over the network. It periodically reads each probe, analyzes it, and turns the various circuits on and off when appropriate. It also records every piece of information into a database; I've got very extensive data collected off my first test fermentation.

Click to expand...

Looks cool, no pun intended.

stephelton said:

Some of the stats I'm particularly interested in collecting and analyzing are the relationship between the coolant temp, ambient temp, fermentation temp, and fermentation phase. I'm hoping to learn enough about these relationships to keep the internal fermentation temperature within a degree Ferenheit.

Click to expand...

Here ya go...start reading...

Very cool. I have been thinking about building something very similar. I want to build a chamber with at least 2 temperature zones (fermentation and lager/cold crash/storage). I've drawn up plans for the box but that's as far as I've got.

Any thoughts on using a window A/C unit to cool?

Mongo64 said:

Any thoughts on using a window A/C unit to cool?

Click to expand...

I've seen quite a few projects that use window AC units. Some of the first plans I came up with for this project involved circulating cool air, but I think using liquid allows better means of control and more precision, plus I can optimize the runtime of the compressor since the temperature of the coolant is fairly arbitrary.

A window AC unit would be much cheaper. $100 or so for a cheap (used?) one, plus $30-50 in construction materials, and something for temperature control.

You can get used window ACs dirt cheap on craigslist, especially if you live in an area where it gets cold in the winter. At the end of the summer last year there were tons of used ACs up there for $20-$25. You don't need anything big unless you're building a walk-in. A small chamber for a few fermenters/kegs will only require a 5,000Btu unit.

I've written an in depth explanation of this system on my blog:

Stephen’s Blog » Fermentation Controller Test Run 1

In case anyone cares

Sweet Asics!

chefchris said:

Sweet Asics!

Click to expand...

Thanks, had to look that up

Application-specific integrated circuit

stephelton said:

Thanks, had to look that up

Application-specific integrated circuit

Click to expand...

I was referring to the shoes.

But I'll go with the later definition since it makes me look smarter.

That's funny. Are you referring to the Google Sketchup model's shoes?

stephelton said:

That's funny. Are you referring to the Google Sketchup model's shoes?

Click to expand...

Yup.

Now that I look at it, they look more like Pumas.

During the first few days of fermentation I use a probe in the wort. After the exothermic activity quites down I switch to an air probe to monitor the PV and use the wort probe only as a readout. My system is an older non-defrosting apartment fridge with a sanke keg for a fermenter, controlled with a Honeywell UDC 3000 controller and "J" thermocouples.
When the wort is generating heat an air probe has too much thermal lag to maintain a small error and vice-versa when the exothermic stage is over. Also the Honeywell controller has the capability to do temp ramps so I can move from fermenting to lagering temps at the rate & time I desire.
I tried several control schemes with a single t/c in the wort but none provided as tight of control as switchng the PV input. The main issue was short cycling the compressor to maintain a small error with a single temp probe in the wort.
The error is +/- .4dgf at it's worst when making the transition for the controlling input and I do this manually by switching some t/c jacks. After the transition the error is zero unless I get a big change in ambient temperature where I keep my fermenting fridge.

jdieter said:

During the first few days of fermentation I use a probe in the wort. After the exothermic activity quites down I switch to an air probe to monitor the PV and use the wort probe only as a readout. My system is an older non-defrosting apartment fridge with a sanke keg for a fermenter, controlled with a Honeywell UDC 3000 controller and "J" thermocouples.
When the wort is generating heat an air probe has too much thermal lag to maintain a small error and vice-versa when the exothermic stage is over. Also the Honeywell controller has the capability to do temp ramps so I can move from fermenting to lagering temps at the rate & time I desire.
I tried several control schemes with a single t/c in the wort but none provided as tight of control as switchng the PV input. The main issue was short cycling the compressor to maintain a small error with a single temp probe in the wort.
The error is +/- .4dgf at it's worst when making the transition for the controlling input and I do this manually by switching some t/c jacks. After the transition the error is zero unless I get a big change in ambient temperature where I keep my fermenting fridge.

Click to expand...

Very good idea. At the height of fermentation the wort is extremely well mixed, but as things die down you could easily get radial temperature gradient in your fermenter. If there is still some fermentation going on, you may get the yeast that is closer to the fermenter wall to start dropping out before the temp probe registers a change, possibly causing problems finishing the ferment.


Something you may want to consider in the future is a profile temperature probe. It's a probe that has multiple sensors are varying locations along the length. Once you start to see radial stratification you can switch over to the air probe.

EDIT: Or you could just use two probes, one in the middle and one near the edge. When the differential between the two reaches a certain point you could switch to the air probe.

Very interesting thoughts. I've observed that, during vigorous fermentation, that there is lots of motion in the wort, so I have no trouble believing that it would be thoroughly mixed at this point.

I have not considered the stratification effect you talk about. Multiple probes in the wort would be very challenging (and still would not entirely solve the problem) so I probably will turn to relying off of the ambient air temp after primary is over.

During my next test run, I will adjust the depth of the temperature probe and see what kind of differences I come across.

Has anyone done any studies on this? I'd like to see what conclusions others have made. In particular, I'd like to see the temperature increase due to exothermic activity over time. I suspect that it becomes negligible very shortly after primary fermentation slows.