How To Build A Cheap-ass Thermowell For Fermenter In 5 Minutes

[shadders]

Was racking my brains looking for a way to get my fridge mate probe into the beer without drowning it. So I was basically looking for a tube that was sealed at one end and had a collar at the other end so I could hang it in the airlock hole. I came up with this.

We've all got one of these lying around somewhere, probably from our very first instant-home-brew-kit. If not I think they are about $6 at a supermarket

If you pull it apart you'll find all the bits you need for a quick and dirty thermowell

basically take the little blue pointy bit, discard the spring, cut off the long part if you want so you can get the probe further down. it has a rubber o-ring already so just push it into one end of the tube point first and you have an instant plug. The little knob on the end is handy for removing it when you want to sanitise. Mine wasn't a perfect seal so I took off the o-ring and wrapped a couple of turns of plumbers tape around the end then put the o-ring back over the top. The extra thickness was enough to form a water tight seal but be careful not to let the tape hang over the edge or you'll make some nice crevices for nasties to hide in. I suppose you could also just put a squirt of silicone in the end and use the plug the keep the beer away from the silicone.

plug in:

At this point it will fit into an airlock grommet though it's a tight fit. If you want to hang it the larger blue bit makes a perfect collar. The hole at one end is just a fraction too small for the fridgemate probe so you can either drill it out or just cut the end off. It seems to be slightly tapered inside so you need the end closest to the two prongs to get a tight fit.

A dab of blutack should seal the end to stop air getting in and out and keep the reading relatively accurate.

Another alternative for a collar which I just tried it get one the nozzles you use on silicon guns. If you cut it at the right place you get a snug it for the tube and the taper flares out so you can wedge it tightly in the airlock hole (sans grommet). They have ridges at the fat end which act as a collar so you don't have to worry about accidently pushing it through the hole. I got mine short enough that I can sit another fermenter on top without touching it.


To sanitize just pull out the plug, pull off the collar and throw them all in.

 

[shadders]

Seem the silicone nozzle collar forms a much more airtight seal than I thought even without a grommet. I'm using one of the bucket style fermenters with a clip on lid that doesn't need an o-ring, didn't think these would be completely airtight either but I just checked it and the top was bubbled up nearly to the point of bursting. temp probe didn't have a drop of beer on it though even with all that pressure so the plug seal must be holding. I've popped it out of the hole and just sat it on top so I'll just have to pray my fridge is sterile. Just gave it a bleach bath after a couple of years on non-use so cross-fingers it should be good.

When this brew is finished I might take collar off and make a tiny slit down one side to give the air a way out when it's sealed in. Maybe just one swipe with a hacksaw.

 

[Frag_Dog]

Seem the silicone nozzle collar forms a much more airtight seal than I thought even without a grommet. I'm using one of the bucket style fermenters with a clip on lid that doesn't need an o-ring, didn't think these would be completely airtight either but I just checked it and the top was bubbled up nearly to the point of bursting. temp probe didn't have a drop of beer on it though even with all that pressure so the plug seal must be holding. I've popped it out of the hole and just sat it on top so I'll just have to pray my fridge is sterile. Just gave it a bleach bath after a couple of years on non-use so cross-fingers it should be good.

When this brew is finished I might take collar off and make a tiny slit down one side to give the air a way out when it's sealed in. Maybe just one swipe with a hacksaw.

Just drill another hole in the lid and install another airlock...

 

[jonocarroll]

I wouldn't be surprised (especially given the thickness) to find that the thermal conductivity of the bottle-filler plastic is worse than that of the fermenter wall... i.e. in this case you may actually be better off with the probe on the outside.

Just sayin'.

 

[pbrosnan]

I wouldn't be surprised (especially given the thickness) to find that the thermal conductivity of the bottle-filler plastic is worse than that of the fermenter wall... i.e. in this case you may actually be better off with the probe on the outside.

Just sayin'.

Yeah that's what I would have thought. I don't know that you'd be getting an accurate reading. I've used dip tubs from kegs with a crimp in one end. I still fill the thermowell with sterile water as the air in the thermowell will provide some insulation so the probe is in liquid but not beer which is probably not good for it.

 

[shadders]

I wouldn't be surprised (especially given the thickness) to find that the thermal conductivity of the bottle-filler plastic is worse than that of the fermenter wall... i.e. in this case you may actually be better off with the probe on the outside.

Just sayin'.

I thought about that but 95% of the tube is inside the fermenter and the top opening is sealed with blutack so it's not exchanging air outside. Considering the large thermal mass of the beer and the time it takes for small temp changes the lag is pretty negligible. End result is that it's working well. The temperature strip on the outside is hovering 0-1 degree below probe temp depending on whether the fridge is in an on cycle or not (when air temp is around 3 degrees). That sort of lag might be problematic for a mash but not for fermentation.

I wanted a solution that wasn't taping to the outside, I've had heaps of problems with that, firstly having to deal with tape residue, condensation on the outside of fermenter making the tape fall off etc. Also with the tape-on solution you've got to worry about sealing it against the fermenter and insulating it so the probe isn't getting ambient air near it. Don't have to worry about any of that. In terms of thermal conductivity I think this is probably better, when you've got a cylinder pressed against a cylinder you've got very little contact area, this way the entire thing is surrounded and the air volume that needs to match is very small. Outside fermenter if you have got it sealed properly you're potentially dealing with a whole fridge full of air volume passing over the probe.

 

[shadders]

After a few weeks of using this thing in two different ferments I'm happy enough with the results to avoid spending money on a stainless version.

I've have the fridgemate set to 19 deg with a differential of 1 deg. 95% of the time the temp reading on the fridgemate sits between 18-20, I've seen a couple of short spikes to 17 and 21 but it only last a short time so it must be just inside the range i.e. 17.9 and 21.1 The fridgemate doesn't kick in until 21 deg (sensor temp must be > 19 set temp + 1 deg differential) so it's tighter than the limits of the fridgemate on the warm side with about of a degree of overshoot on the cool side. I suspect with a temp controller with 0.1 deg resolution this would be much tighter range and I could also tighten it up by filling the tube with liquid. The only reason I haven't is because the seal at the top of the probe looks slightly perished and I don't want to risk it getting water inside.

Overall I think compared to the water in a jar or taped to side of fermenter methods you'll get a slightly wider band of temp variation with longer on/off cycles of the compressor but the band will be *much* more accurately centered around wort temp. This method will also get you down to pitching temp faster as it basically holds the fridge on for as long as it takes to get it down there. I've found with the tape method it still cycles on/off a few times in that initial phase. Downside is that you do get about a 2 degree overshoot on that initial cooling phase so on the second brew I set the fridgemate 2 degrees higher for about 24 hrs then dropped it back to 19 once it stabilised.

What I might try on my next brew is turning the fridge's thermostat up to it's warmest temp after the initial cooling phase. This should reduce overshoot on the cooling phase as there's less 'coolth' stored in the fridge air and insulation after the fridgemate switches it off.

 

[jonocarroll]

What I might try on my next brew is turning the fridge's thermostat up to it's warmest temp after the initial cooling phase. This should reduce overshoot on the cooling phase as there's less 'coolth' stored in the fridge air and insulation after the fridgemate switches it off.

coolth?

Setting the fridge thermostat higher should do nothing at all. The external temp controller doesn't 'override' the internal one in any way, it just dictates when the internal one can activate the fridge. Setting the internal to minimum ensures that the internal thermostat won't interfere too much (and depending on how it's build, how hard it will try to get down to the internal-set temperature). Setting it to maximum will only mean that it might interfere and say, turn the fridge off when it gets below 6C (if you were, say crash chilling). Doing this at ferment temps will do nothing.

 

[shadders]

coolth?

the opposite of warmth

Setting the fridge thermostat higher should do nothing at all. The external temp controller doesn't 'override' the internal one in any way, it just dictates when the internal one can activate the fridge. Setting the internal to minimum ensures that the internal thermostat won't interfere too much (and depending on how it's build, how hard it will try to get down to the internal-set temperature). Setting it to maximum will only mean that it might interfere and say, turn the fridge off when it gets below 6C (if you were, say crash chilling). Doing this at ferment temps will do nothing.

When the wort gets up to 21 deg and the fridgemate kicks in then until the wort drops below 19 the internal temp of the fridge is governed by it's own thermostat. With it turned way down the fridge temp can be below 0. At the other end it can be up to 10 deg. So at the warmer setting, during the fridgemate's cooling cycle you have a longer gentler cooling phase. You also don't cool down the fridge air or insulation anywhere near as much. Once the fridgemate switches off this stored cold continues to cool the fermenter until the fridge reaches equilibrium with the fermenter (minus some minor loss to external air). If you've got your fridge air and insulation cooled to 0 deg that's a much bigger differential in thermal mass that needs to be balanced. At 10 degrees the fermenter will stop cooling much sooner after the fridgemate switches off.

It's probably all academic, I don't know how big a deal a 1-2 degree variation in ferm temp really is, I doubt it's significant unless you are trying to ferment very close to the temp where the yeast goes to sleep and you need to be sure you don't overshoot on the cool side of the cycle.

 

[jonocarroll]

the opposite of warmth

Ah.

When the wort gets up to 21 deg and the fridgemate kicks in then until the wort drops below 19 the internal temp of the fridge is governed by it's own thermostat. With it turned way down the fridge temp can be below 0. At the other end it can be up to 10 deg. So at the warmer setting, during the fridgemate's cooling cycle you have a longer gentler cooling phase.

Only if the internal probe gets down to the T_internal (what you have the internal set to) will you get a longer cooling phase. Does the fridge actually turn off while above set temperature?

You also don't cool down the fridge air or insulation anywhere near as much. Once the fridgemate switches off this stored cold continues to cool the fermenter until the fridge reaches equilibrium with the fermenter (minus some minor loss to external air). If you've got your fridge air and insulation cooled to 0 deg that's a much bigger differential in thermal mass that needs to be balanced. At 10 degrees the fermenter will stop cooling much sooner after the fridgemate switches off.

The heat capacity of the air in the fridge itself will be comparatively very low (doesn't take much to heat the air). I question the ability to chill the insulation down to a lower temperature than the wort to such a degree that it will then provide significant heat-sink to the wort once the fridge is turned off.

The majority of the overshoot comes from the wort not being at a uniform temperature (hence the use of a well-placed probe). Some of the wort will be over-cooled, while some is not yet at the desired temperature. The lag-time is due to the over-cooled wort heating, and to a lesser degree the air/inside components heating.

As a physicist, I can't help but correct people on the notion of "transfer of cool(ness)/'coolth'" or "stored cold". No such thing exists. There is transfer of heat energy, but the difference is important. Something cold doesn't "store cool", but has the potential to be heated.

It's probably all academic, I don't know how big a deal a 1-2 degree variation in ferm temp really is, I doubt it's significant unless you are trying to ferment very close to the temp where the yeast goes to sleep and you need to be sure you don't overshoot on the cool side of the cycle.

Remember, it's not a "1-2 degree variation in ferment temp" - it's a 1-2 degree variation in the probe temperature. The ferment temp will not be constant anyway since you have a macroscopic heat-generating vessel with a non-homogeneous cooling system with a poor active surface area, measured at a single point.

In the end, you will make beer. It's probably even reproducible, so yes, largely academic.

 

[shadders]

Only if the internal probe gets down to the T_internal (what you have the internal set to) will you get a longer cooling phase. Does the fridge actually turn off while above set temperature?

Yes. Probe temp = 21 deg. Set temp is 19 deg, fridge temp is 0 degrees = 21 deg differential. fridge temp 10 deg = 11 degree differential -> slower cooling

I question the ability to chill the insulation down to a lower temperature than the wort to such a degree that it will then provide significant heat-sink to the wort once the fridge is turned off.

Which raises an interesting question that a physicist might be able to answer. I presume thermal conductivity and thermal capacity are only loosely related. Do you know if this is the case or if there's any relationship at all?

As a physicist, I can't help but correct people on the notion of "transfer of cool(ness)/'coolth'" or "stored cold". No such thing exists. There is transfer of heat energy, but the difference is important. Something cold doesn't "store cool", but has the potential to be heated.

Probably the same way that flow of current from positive to negative doesn't exist but it's much more convenient way of talking about it. You can blame Dr Karl for that one I think I heard the term 'coolth' from him.

 

[jonocarroll]

Yes. Probe temp = 21 deg. Set temp is 19 deg, fridge temp is 0 degrees = 21 deg differential. fridge temp 10 deg = 11 degree differential -> slower cooling

A question first: Where is the 0C measurement coming from? That differential is only relevant if it is maintained, which it's not if the fridge is off.

Which raises an interesting question that a physicist might be able to answer. I presume thermal conductivity and thermal capacity are only loosely related. Do you know if this is the case or if there's any relationship at all?

They are not related. Several examples exist which contra-indicate a relationship (Argon gas, Carbon as graphite and diamond to name a couple).

Probably the same way that flow of current from positive to negative doesn't exist but it's much more convenient way of talking about it.

But 'flow of current' is the wrong thing to talk about if you want to understand the physics. 'Flow of electrons' is perfectly acceptable, and just happens to go the opposite way to convention for current.

 

[shadders]

A question first: Where is the 0C measurement coming from? That differential is only relevant if it is maintained, which it's not if the fridge is off.

Separate thermometer inside fridge (outside fermenter). The fridge is on at this point because probe temp is above set temp. Until probe temp drops below set temp the fridge temp will be determined by fridge thermostat.

 

[jonocarroll]

Separate thermometer inside fridge (outside fermenter).

Yes, but is it anywhere near either the cooling plate or the fridge probe?

Of course if the fridge gets/stays cooler for longer the cooling will be faster, but with a relatively warm fermenter (as in above the 10C set on the fridge) how long will the fridge stay turned off for once it hits that 10C? I can't believe it's cooling the wort [large, warm, reacts slow] significantly while the fridge probe [small, reacts fast] stays below 10C (i.e. fridge off).

 

[shadders]

Yes, but is it anywhere near either the cooling plate or the fridge probe?

Of course if the fridge gets/stays cooler for longer the cooling will be faster, but with a relatively warm fermenter (as in above the 10C set on the fridge) how long will the fridge stay turned off for once it hits that 10C? I can't believe it's cooling the wort [large, warm, reacts slow] significantly while the fridge probe [small, reacts fast] stays below 10C (i.e. fridge off).

No the temp probe is in the thermowell inside the fermenter (see thread title). The other thermometer is sitting on a shelf in the fridge a fair way from the cooling plate so it's just measuring ambient air temp in fridge.

fridge is on until wort temp drops below 18.

 

[jonocarroll]

fridge is on until wort temp drops below 18.

This was my point. If the fridge is on then the fridge has not reached the internal-set temperature, in which case there is no difference in changing the internal setting... i.e. if it's not getting down to 10, setting it to 0 doesn't do anything.

The instantaneous ambient air temperature inside the fridge is not a useful number, since it's not a system at equilibrium.

 

[shadders]

This was my point. If the fridge is on then the fridge has not reached the internal-set temperature, in which case there is no difference in changing the internal setting... i.e. if it's not getting down to 10, setting it to 0 doesn't do anything.

The instantaneous ambient air temperature inside the fridge is not a useful number, since it's not a system at equilibrium.

You're missing the point. While the fridge is on the fridge's thermostat takes over. If fridge thermostat is set to 10 deg then when ambient air temp gets below 10 deg the fridge will remain on but compressor will switch off. End result is ambient air temperature bottoms out at 10 deg. While the fridge is on the air temp will drop rapidly to 10 then hover pretty close to that temp until the wort temp drops below 18 and the fridgemate switches the fridge off.

Air temp will get to 10 a lot faster than wort temp will get to 18 so the fridge has a window of self determination before it fridgemate steps in and swipes away it's free will.

 

[jonocarroll]

You're missing the point. While the fridge is on the fridge's thermostat takes over. If fridge thermostat is set to 10 deg then when ambient air temp gets below 10 deg the fridge will remain on but compressor will switch off. End result is ambient air temperature bottoms out at 10 deg. While the fridge is on the air temp will drop rapidly to 10 then hover pretty close to that temp until the wort temp drops below 18 and the fridgemate switches the fridge off.

Air temp will get to 10 a lot faster than wort temp will get to 18 so the fridge has a window of self determination before it fridgemate steps in and swipes away it's free will.

I now see you're discriminating between 'fridge on' and 'compressor on' - hadn't assumed that in your text yet, which may have caused some confusion on my part.

The above looks fine, but there's a lot missing between "air temp will drop rapidly to 10" and "wort temp drops below 18".

To get back on track, my complaint was with the position that the ambient air does some reasonable cooling once the fridge compressor turns off, viz

What I might try on my next brew is turning the fridge's thermostat up to it's warmest temp after the initial cooling phase. This should reduce overshoot on the cooling phase as there's less 'coolth' stored in the fridge air and insulation after the fridgemate switches it off.

...

Once the fridgemate switches off this stored cold continues to cool the fermenter until the fridge reaches equilibrium with the fermenter (minus some minor loss to external air).

I'll use science;

The heat that can be input to the (say, 40L of) air (from the, say 20L wort, assumed similar to water) to raise the air from 0C to 18C is

ΔQ = mcΔT = (40 x 1.2) x 1.012 x 18 = 874J.

To raise it from 10C to 18C requires less, of course

ΔQ = mcΔT = (40 x 1.2) x 1.012 x 8 = 389J.

This is the energy available to be extracted from the wort to heat the air (or 'energy to cool the wort' to use incorrect phrasing).

In order to reduce the wort by 1C, say from 19C to 18C, we need

ΔQ = mcΔT = (20 x 1000) x 4.18 x 1 = 83600J.

Clearly the energy just isn't there. Any overshooting here is not significantly contributing to the cooling. Every time the compressor turns off, the energy absorbed by the air warming from 0C to 18C reduces the wort temperature by 0.01C. Changing the internal thermostat to 10C would mean that each compressor-off cycle reduces the wort temperature by 0.005C - surely not a big issue. I could run through the same exercise for the insulation, but I don't think it's going to matter.

 

[shadders]

glad we're back on the same page. Ok so you've convinced me the residual 'coolth' will have negligible effect. There is one other factor I thought of while we were having our little back and forth...

The lower temp differential between ambient air and wort means the rate of cooling is slower, or rather the rate or air warming I would assume that this gives more time for the different temperature zones in the wort to mix (and more time for the thermowell to soak it up), thus a lesser variation in wort temps. End result the probe temp lags the average wort temp by a lesser margin.

Again maybe academic but if I can gain an extra 1/2 - 1 degree of temp stability simply by twisting a dial I may as well.

 

[jonocarroll]

The lower temp differential between ambient air and wort means the rate of cooling is slower, or rather the rate or air warming I would assume that this gives more time for the different temperature zones in the wort to mix (and more time for the thermowell to soak it up), thus a lesser variation in wort temps. End result the probe temp lags the average wort temp by a lesser margin.

What you're explaining here is two objects reaching thermal equilibrium, which technically takes infinite time (needs to be done infinitely slowly), but the approximation that slower = better is good. It does indeed allow for more time for the wort to become homogeneous and thus for the probe to measure a better average temperature.

Now, the problem is, "is there a smaller temperature differential?". Assuming that a fridge runs as hard as it can while above T_internal (the lower temps being achieved by running longer) then it's a matter of "is it [the compressor] on for longer when the internal dial is set lower?" or does it not reach T_internal until the wort gets close to T_external? The way to tell would be to add up the total "on" time for the compressor during getting the wort down to T_external - sure, it may turn off a bit more with T_internal set high, but over the course of the temp drop is it really all that different from when T_internal is set low?

My suspicion would be that given the effective heater sitting in the insulated box, the ambient air would not be at T_internal for longer than the compressor delay.