I have a setup with a raspberry pi and two DS18B20 digital thermometers that measure the temperature of the ambient air and the water in a swamp cooler that I place my fermentation chamber in. I was alerted to the fact that measuring the swamp cooler water was a poor way to estimate the temperature of the actual fermenting beer.

I have ordered a third DS18B20 and I want to measure the actual beer temperature. I want to see if there are any correlation between the ambient, swamp water, and beer temperatures. In order to accomplish this I wish to have the DS18B20 measuring the beer directly -- I do not want to put an insulation material on the outside of the fermentation chamber (however, since this method is commonly done, I think I will order another probe and see how well it correlates to the probe in the beer).

What I would like is ideas on how to best accomplish this while keeping everything air tight.

I will be using buckets as opposed to carboys for this experiment.

I plan to continually update the numbers to this if anyone else has any interest.


5 Answers 5


While not the most elegant, you can go through the lid in a larger fermentor without a thermowell. I put this together without wanting to incur the cost of a thermowell - so I used some electrical tape, an extra hole opposite the airlock and a rubber grommet. I patiently and accurately wrapped the tape around the sensor wire, building it out so it fit nice and snug in the grommet.

This is what I've done with my BrewPi setup; on my sixth brew now and things are still airtight. Next brew I'll be removing, cleaning down, and replacing the tape - the repeated soaking in sterilising solution means the tape has started to get a bit gummy as you can see in the picture - the top layer is starting to pull away.

So, perhaps not 100% ideal as the thermometer is directly in your beer, but has served me well this far.

enter image description here


Given that you're fermenting in buckets, then you need a straight-walled thermowell, like this: enter image description here

You put the thermowell through the lid of the bucket and down into the beer. The temperature probe slides inside the thermowell.

The tricky part is to then make an airtight seal with the bucket lid. You can either buy a stopper with a hole, or a grommet. The thermowell goes through the middle hole of the stopper or grommet. Both will require that you drill a larger hole into the lid of the bucket. The OD of the thermowell is 3/8", so this is the size of the hole in the stopper/grommet. Once inserted into the lid, the stopper/grommet will make an airtight seal both with the bucket and with the thermowell.

  • If I may hijack for a moment; may these by used in a carboy, side-by-side with either a blow-off tube or airlock? I would presume it's just a matter of finding a stopper that would accept both devices.
    – object88
    Commented Jul 8, 2013 at 23:01
  • Indeed - you can buy rubber carboy caps with two inlets - northernbrewer.com/shop/3-5-6-gall-carboy-cap-orange.html. The inlets weren't the same size so I could fit the thermowell in the 3/8" inlet, but not a airlock in the second smaller one, so I simply used foil instead of the airlock. It's also possible to get hold of a double hole stopper. morebeer.com/products/stopper-65-drilled-1-38-hole-14.html
    – mdma
    Commented Jul 9, 2013 at 1:07
  • @MDMA Would you please check my answer and tell me your opinion? Commented Jul 17, 2013 at 0:11

I like the idea of taping the thermometer to outside of the bucket and adding a layer of insulation over top so it is more closely registering the beer temp and not the air temp.

  • What you gain with similicity, you lose with accuracy. In my experience with the brewpi temp controller project, the temperature still follows the air temperature rather than the beer temperature when the probe is stuck on the outside. Think of the beer as one side of the cube, then the beer temperature influences the measured temperature from only one side, while the ambient air temp has influence from the other 5 sides. At best, you're getting a weighted average of the beer temp and air temp.
    – mdma
    Commented Jul 8, 2013 at 16:59
  • 1
    Yes but with the insulation you get a lot more accurate without the worry of violating the fermentation chamber. Of course, what level of control are you exerting using only a fridge vs. say a glycol wrap. In other words, how closely do you really want/need to control it?
    – uSlackr
    Commented Jul 8, 2013 at 17:27
  • uSlackr, I will do both and compare the two to see if insulation is actually accurate or not, but I definitely want to record the temperature of the actual beer. What material do you suggest to use as insulation? Commented Jul 8, 2013 at 21:14
  • I would think any type of foam would work well.
    – uSlackr
    Commented Jul 9, 2013 at 12:44

I don't know how I would go about drilling a hole and fitting it with a bung for a thermowell, so I decided to simply nail a hole in the lid and permenantly attach the DS18B20 probe using a sealant. My mom used to make and sell aquariums, and she suggested using silicone glue. Here is a post on my web page detailing the step-by-step and including all the pictures. Otherwise, here is the run down:

I made a hole in the lid with a nail and a hammer. I opened the hole up with a pen. I dropped the probe inside the hole so that it was about half way down the bucket. Would the top or the bottom be a better location? Then I sealed the probe to the lid with silicone glue. I put the glue only on the top of the lid because I feared the potential for the taste to seep into the beer.

enter image description here

What do you think?

enter image description here


I have reason to believe that your concerns about the swamp cooler temperature differing from the fermentation chamber temperature are not based in reality. I calculated that putting one ice pack in the swamp cooler should reduce temperature by about 2.3 degrees Celcius if the heat transfer between the ice pack, the swamp cooler and the fermentation chamber is ideal. I have approximately the same amount of water in the fermentation chamber and in the swamp cooler. Now, if there is practically no heat transfer between the swamp cooler and the fermentation chamber, the temperature of the swamp cooler would rise by about 4.6 degrees Celcius (but one could ask the valid question of why the swamp cooler is used in the first place, then, if that was the case).

I tested what happens if I put one ice pack in the swamp cooler. Temperature reduced by about 2.1 degrees, but there was a long delay of about an hour. This is close to the predicted 2.3 degrees Celcius temperature decrease.

I haven't actually measured the temperature difference between the fermenting beer and the swamp cooler because I don't want that oxygen gets to the fermentation chamber, but based on this simple test of putting one ice pack on the swamp cooler, I believe heat transfer between the fermentation chamber and swamp cooler is good.

I could of course calculate how great is the heat transfer rate given the area and thickness of plastic that I have, and how great is the hear produced by completely fermenting the beer. Then when assuming a particular fermentation rate, one could calculate the temperature difference. I don't believe the difference is large. But I'm satisfied with my simple experiment of putting one ice pack on the swamp cooler and seeing that part of the coolness must go to the fermentation chamber.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.