How do you know what temp the keg is?

[htims05]

So my first foray into kegging...I set the co2 to 30psi for the first 24hrs then set the pressure based off “the chart”.

My problem is I really don’t know the beer temp. I thought I did with my temp controller submersed in a bottle of water but when I pull a sample from the tap the beer is much warmer. 10ish degrees.

Should my temp controller be in water or not (I know it’s a hot debate). Or should I go based off what I measured at the tap/pour?

 

[Robert65]

To avoid getting into the hot debate you correctly identify... could you get one of those adhesive, liquid crystal thermometer strips and stick it on the keg? Might be the simplest way to get a good idea of the beer temperature.

 

[htims05]

To avoid getting into the hot debate you correctly identify... could you get one of those adhesive, liquid crystal thermometer strips and stick it on the keg? Might be the simplest way to get a good idea of the beer temperature.

Yeah I thought that would be a good idea. I googled for those a while ago but never found any in the correct temp range. They were all for fish tanks.

 

[Robert65]

Yeah I thought that would be a good idea. I googled for those a while ago but never found any in the correct temp range. They were all for fish tanks.

Huh. My LHBS used to have them that went down to 34° or 36°F, now they only have them going down to 44°F. Here's one that goes to 39°, might almost be good enough. https://www.mibrewsupply.com/fermometer

 

[day_trippr]

The first pour of an idle faucet is going to produce beer that's warmer than that in the bottom of the keg. A 10 degree rise wouldn't be unusual at all. If you want to get closer, pour another glass or two and see the temperature get ever closer to the keg temperature.

fwiw, I strap a digital probe lead to the side of the fullest keg in the keezer with a pad of inch-thick closed cell foam over it as insulation against the cabinet air temperature. Along with four other probes feeding a multi-channel monitor program I can see exactly what's going on inside my keezer. Here's a plot for the last 24 hours, with the actual temperature controller set for a beer temperature of 35°F +/- 1°...

Cheers!

 

[htims05]

The first pour of an idle faucet is going to produce beer that's warmer than that in the bottom of the keg. A 10 degree rise wouldn't be unusual at all. If you want to get closer, pour another glass or two and see the temperature get ever closer to the keg temperature.

fwiw, I strap a digital probe lead to the side of the fullest keg in the keezer with a pad of inch-thick closed cell foam over it as insulation against the cabinet air temperature. Along with four other probes feeding a multi-channel monitor program I can see exactly what's going on inside my keezer. Here's a plot for the last 24 hours, with the actual temperature controller set for a beer temperature of 35°F +/- 1°...

View attachment 656219

Cheers!

Ok I’m trying that now...looks a little crazy in there.

 

[day_trippr]

As crazy as this?

Cheers!

 

[htims05]

As crazy as this?

View attachment 656227

Cheers!

Haha no...just a lot of tape I used to counter act the condensation.

What temp gauges/system are you using?

 

[bracconiere]

IR Thermometer is what i use....just shoot it at the keg in the fridge....$15 or so on ebay.....

 

[htims05]

IR Thermometer is what i use....just shoot it at the keg in the fridge....$15 or so on ebay.....

I have one of those - never sure how accurate they are...I get weird readings sometimes - presumably due to reflections...usually shiny surfaces cause the invalid readings...for example top part of my keg in the kegerator said 72...bottom said 43....keg only had 4 gallons in it. maybe my infrared is just junk.

 

[bracconiere]

I have one of those - never sure how accurate they are...I get weird readings sometimes - presumably due to reflections...usually shiny surfaces cause the invalid readings...for example top part of my keg in the kegerator said 72...bottom said 43....keg only had 4 gallons in it. maybe my infrared is just junk.

stuck my digital probe in the fridge, got 37f, IR read 40f at the bottom of the keg, 50f or something at the top....but water has more thermal mass, and just having the door open for that few seconds the probe went up....

 

[Robert65]

Unless your sensor has an adjustable emissivity setting, and you can calibrate it against a contact probe, you cannot use an IR thermometer to measure a reflective surface; the device relies on simplified assumptions about the emissivity of non reflective surfaces. You can somewhat account for this by putting a piece of non reflective tape on the surface to be measured, but at a considerable loss of accuracy. In short, wrong tool for the job.

 

[bracconiere]

Unless your sensor has an adjustable emissivity setting, and you can calibrate it against a contact probe, you cannot use an IR thermometer to measure a reflective surface; the device relies on simplified assumptions about the emissivity of non reflective surfaces. You can somewhat account for this by putting a piece of non reflective tape on the surface to be measured, but at a considerable loss of accuracy. In short, wrong tool for the job.

i guess i'm lucky i just want them cold enough to drink.....lol

edit: but for the record, i just stuck my digital thermometer in my freshly poured beer, and pointed my IR therm at the bottom of the keg, both read exactly the same....41f within tenths of a degree....

 

[day_trippr]

What temp gauges/system are you using?

I have a RaspberryPi under the keezer than runs my metered tap list (RaspberryPints) and controls the keezer and my three ferm chambers (BrewPi). It also runs my temperature monitoring/charting program. The probes are ds18b20 "one-wire" sensors on 3 meter wires...

Cheers!

 

[doug293cz]

One issue that can occur in kezzers is temperature stratification. The cure for this is putting a fan in the kezzer to move the air around. Ideally, there are some gaps between the kegs to allow for air to flow around them.

Thermodynamics tells us that if the air temp in the kezzer is reasonably constant, kegs in the kezzer will come to equilibrium at the average temp of the air. No real need to place the temp probe anywhere other than hanging free in the air. Might take a couple of days for a warm keg to come to equilibrium. You do want to make sure you have the compressor start delay on your temp controller set to more than 10 minutes.

Brew on

 

[day_trippr]

Logical, but the difference is if you control a keezer via its air temperature the compressor will cycle WAY more frequently than the five times in 24 hours shown.

If one examines the air temperature curves vs the keg temperature curves in the plot above and imagine where the temperature control points would exist if using air temperature it's pretty clear what will happen. Which is why I do what I do

Cheers!

 

[Unicorn_Platypus]

a better question is how do you know how much beer is left in the keg? Solve that one for me please


In terms of gauging the approximate temperature in the keg this is what I do.

I either pick the fullest keg in the keezer or stick an empty keg with water in the keezer if there's room. Then I duct tape the probe to the middle/bottom of said keg.

The probe touches the outside of the keg directly. I then insulate the probe in a few layers of ripped paper towel, then duct tape on top of the paper towel. This gives me a pretty good gauge of the kegs temperatures in the keezer (assuming all kegs have been in there long enough to equalize). I

I set my temp controller @ 41 degrees with a +2 variable, so figure its at ~42 on average. I've had good success at using carbonation charts and yielding accurate levels based on my temperature read.

note: I also use this method for my carboys during fermentation. This way I get an active read on the actual ferment temp

 

[Unicorn_Platypus]

As crazy as this?

View attachment 656227

Cheers!

what kind of tubing and connectors are you using for the gas side on this beast? looks uber LODO

 

[Robert65]

a better question is how do you know how much beer is left in the keg? Solve that one for me please

Answer: How many people are you having over tonight? There's at least 1 pint less than that number. [emoji57]

 

[day_trippr]

a better question is how do you know how much beer is left in the keg? Solve that one for me please.[...]

Did you notice the flow meters atop six of the kegs in my keezer pic above? They provide the input to make my RaspberryPints tap list happen:

fwiw, I recommend going heavy with the probe insulation. The more the better to isolate the probe from the cabinet temperature. One inch thick foam virtually eliminates coupling to ambient.

what kind of tubing and connectors are you using for the gas side on this beast? looks uber LODO

It is now
I just recently replaced all of the cold side tubing in my humble brewery with EVAbarrier, both beer (I used 4mm ID for that) and gas (5mm ID there). That's the keezer and three fridges outfitted for both fermentation and cold-conditioning/carbonation up to four kegs each. Sounds like a lot but tubing wise it was only one roll of the 4mm and two rolls of the 5mm.

It was the fittings that were the killer. With six beer runs, 20 gas drops (including beer gas and a high pressure line for my fizzy beers), and the runs from three regulators to the keezer and three other regs to the three fridges I used almost five dozen PTC connectors from John Guest, Duotight and DMfit, to connect QDs, flow meters, beer shanks, regulators, manifolds, etc. Even the new rinser got fitted with a PTC (JG 1/2" BSP to 5/16" OD tubing, fwiw).

But I'll say this: I leaked tested the hell out of every bit and those fittings are the bomb when used with the EVA tubing. I even found (and fixed) a glacially slow regulator PRV thread leak that had probably been there since bought over a decade ago because everything else is holding tight so well.

So I'm pretty pleased so far, and to that extent I'd definitely recommend this tubing system...

Cheers!

 

[Unicorn_Platypus]

Did you notice the flow meters atop six of the kegs in my keezer pic above? They provide the input to make my RaspberryPints tap list happen:

View attachment 656275

fwiw, I recommend going heavy with the probe insulation. The more the better to isolate the probe from the cabinet temperature. One inch thick foam virtually eliminates coupling to ambient.

It is now
I just recently replaced all of the cold side tubing in my humble brewery with EVAbarrier, both beer (I used 4mm ID for that) and gas (5mm ID there). That's the keezer and three fridges outfitted for both fermentation and cold-conditioning/carbonation up to four kegs each. Sounds like a lot but tubing wise it was only one roll of the 4mm and two rolls of the 5mm.

It was the fittings that were the killer. With six beer runs, 20 gas drops (including beer gas and a high pressure line for my fizzy beers), and the runs from three regulators to the keezer and three other regs to the three fridges I used almost five dozen PTC connectors from John Guest, Duotight and DMfit, to connect QDs, flow meters, beer shanks, regulators, manifolds, etc. Even the new rinser got fitted with a PTC (JG 1/2" BSP to 5/16" OD tubing, fwiw).

But I'll say this: I leaked tested the hell out of every bit and those fittings are the bomb when used with the EVA tubing. I even found (and fixed) a glacially slow regulator PRV thread leak that had probably been there since bought over a decade ago because everything else is holding tight so well.

So I'm pretty pleased so far, and to that extent I'd definitely recommend this tubing system...

Cheers!

I'll stick with my paper towel for now, but maybe treat myself with an upgrade to thicker foam in 2020

I'm planning to upgrade to the EVA with duo tights on the gas side at some point, but have been worried about leaks. The current distributor in my keezer uses MFL connectors. I've read mixed reviews regarding leaks on the duo tight MFL fittings when connecting to stainless steel MFL connections. I see your distributor is not using MFL, what kind of connectors are you using and how tough is it to rig that up? assuming barbs fittings can't be used with john guest style connections.

 

[bracconiere]

i'm do drunk to finsh reading the last post, but it gave me a hard on....

edit: or post before last....lol.

does that flow meter count ounces? i'd love to stop having to use gradients on my glasses and a ticker for keeping count....

 

[Robert65]

I've had no trouble with the Duotight MFL x PTC adapters. Recommend.

 

[Unicorn_Platypus]

Did you notice the flow meters atop six of the kegs in my keezer pic above? They provide the input to make my RaspberryPints tap list happen:

View attachment 656275

fwiw, I recommend going heavy with the probe insulation. The more the better to isolate the probe from the cabinet temperature. One inch thick foam virtually eliminates coupling to ambient.



It is now
I just recently replaced all of the cold side tubing in my humble brewery with EVAbarrier, both beer (I used 4mm ID for that) and gas (5mm ID there). That's the keezer and three fridges outfitted for both fermentation and cold-conditioning/carbonation up to four kegs each. Sounds like a lot but tubing wise it was only one roll of the 4mm and two rolls of the 5mm.

It was the fittings that were the killer. With six beer runs, 20 gas drops (including beer gas and a high pressure line for my fizzy beers), and the runs from three regulators to the keezer and three other regs to the three fridges I used almost five dozen PTC connectors from John Guest, Duotight and DMfit, to connect QDs, flow meters, beer shanks, regulators, manifolds, etc. Even the new rinser got fitted with a PTC (JG 1/2" BSP to 5/16" OD tubing, fwiw).

But I'll say this: I leaked tested the hell out of every bit and those fittings are the bomb when used with the EVA tubing. I even found (and fixed) a glacially slow regulator PRV thread leak that had probably been there since bought over a decade ago because everything else is holding tight so well.

So I'm pretty pleased so far, and to that extent I'd definitely recommend this tubing system...

Cheers!

this is one sick frickin setup by the way!

 

[day_trippr]

[...]I see your distributor is not using MFL, what kind of connectors are you using and how tough is it to rig that up? assuming barbs fittings can't be used with john guest style connections.

Yes, the keezer has a 6 port manifold with barbs that frankly I wasn't inclined to replace with an MFL version ($$).
Using a heat gun it took about ten minutes to fit all seven 5mm ID tubing runs to it, NBD.

My three fridges, otoh, were acquired after I had wizened up to the whole flare fitting thing, and each sports a four way MFL manifold. Those got the full Duotight treatment, like these two:

I dunk tested everything I could - entire manifolds and all their QD runs - and also did protracted gas tests (charge up, close cylinder, watch valves) and literally went two weeks without any of the systems losing pressure all the way to the QDs. I really am impressed with the PTC performance of all three brands used.

The only thing worth noting is you don't use flare gaskets with these connectors (any of the brands) and instead rely on the deformation of the interior mating surface itself to form a gas tight seal. So you need to put some pressure on the connection - but stop shy of blowing the back out of it. I did manage to find that breaking point on one connector and put a nice circumferential crack in it. Don't do that

[...]does that flow meter count ounces? i'd love to stop having to use gradients on my glasses and a ticker for keeping count....

Yes, I have my setup configured for ounces (zoom in a bit), but there's a version available that goes metric for those so inclined.

this is one sick frickin setup by the way!

Thank you kindly

I call it my Low Art/High Tech Keezer, and it keeps me off the streets at night

Cheers!

 

[doug293cz]

The first pour of an idle faucet is going to produce beer that's warmer than that in the bottom of the keg. A 10 degree rise wouldn't be unusual at all. If you want to get closer, pour another glass or two and see the temperature get ever closer to the keg temperature.

fwiw, I strap a digital probe lead to the side of the fullest keg in the keezer with a pad of inch-thick closed cell foam over it as insulation against the cabinet air temperature. Along with four other probes feeding a multi-channel monitor program I can see exactly what's going on inside my keezer. Here's a plot for the last 24 hours, with the actual temperature controller set for a beer temperature of 35°F +/- 1°...

View attachment 656219

Cheers!

Seems that you should be able to obtain the same compressor cycle time by hanging the probe in the air, and setting the total hysteresis range to the air temp delta seen when controlling with keg temp.

Also, how did you perform the thermodynamically impossible feat of getting the beer colder than the surrounding air?

Brew on

 

[htims05]

Seems that you should be able to obtain the same compressor cycle time by hanging the probe in the air, and setting the total hysteresis range to the air temp delta seen when controlling with keg temp.

Also, how did you perform the thermodynamically impossible feat of getting the beer colder than the surrounding air?

Brew on

This kinda what I am doing...if I just find the right temp on the controller that gives me a beer temp of 38-40 then I can get rid of all the tape and pia of redoing it every time I switch one out (I only have 2 kegs).

But for the sake of figuring this all out...I’ve taped my sensor to a keg and threw in one of my wireless outdoor weather station thermometers in there to see what the air temp is.

 

[TheMadKing]

How close is close enough on carbonation volume for you guys?

As long as I'm between 38-44F and 11-15psi I'm happy. I tend to prefer lower carbonation than 2.5vols anyway so I do it by feel.

Without having a side by side reference, I can't feel the difference between 2.2-2.7 volumes anyway. Once you start getting over 3 or under 2 you can really feel a difference that's immediately apparent.

As long as it's in the ballpark and not distractingly high or distractingly low I don't really care what my carbonation level is.

 

[day_trippr]

Seems that you should be able to obtain the same compressor cycle time by hanging the probe in the air, and setting the total hysteresis range to the air temp delta seen when controlling with keg temp.

Also, how did you perform the thermodynamically impossible feat of getting the beer colder than the surrounding air?

Ahahaha! Whoops! I recently took everything out of the keezer and removed the lid assembly for rust remediation, a rinser installation, and changing out all of the beer and gas lines for EVAbarrier tubing and PTC fittings. Apparently I did not put the chamber probes back where they belong so aside from the keg and tower channels they're way out of whack.

Here's what that plot should have looked like:

While I don't know exactly how everything would behave if I used a chamber air probe to control the compressor, intuitively I expect much more cycling because the probe response is so quick to air temperature changes that the compressor would run for a very short time then await the next set point crossing - which instead of being 4-6 hours later would be more like minutes.

One of these days I might give it a try and see what happens...

Cheers!

 

[doug293cz]

Ahahaha! Whoops! I recently took everything out of the keezer and removed the lid assembly for rust remediation, a rinser installation, and changing out all of the beer and gas lines for EVAbarrier tubing and PTC fittings. Apparently I did not put the chamber probes back where they belong so aside from the keg and tower channels they're way out of whack.

Here's what that plot should have looked like:

View attachment 656316

While I don't know exactly how everything would behave if I used a chamber air probe to control the compressor, intuitively I expect much more cycling because the probe response is so quick to air temperature changes that the compressor would run for a very short time then await the next set point crossing - which instead of being 4-6 hours later would be more like minutes.

One of these days I might give it a try and see what happens...

Cheers!

If you do try letting the probe hang free, just don't let it be to close to the sides of the keezer.

Brew on

 

[bracconiere]

Ahahaha! Whoops! I recently took everything out of the keezer and removed the lid assembly for rust remediation, a rinser installation, and changing out all of the beer and gas lines for EVAbarrier tubing and PTC fittings. Apparently I did not put the chamber probes back where they belong so aside from the keg and tower channels they're way out of whack.

Here's what that plot should have looked like:

View attachment 656316

While I don't know exactly how everything would behave if I used a chamber air probe to control the compressor, intuitively I expect much more cycling because the probe response is so quick to air temperature changes that the compressor would run for a very short time then await the next set point crossing - which instead of being 4-6 hours later would be more like minutes.

One of these days I might give it a try and see what happens...

Cheers!

looks like your date isn't set, or that's 4.5 years old.....

 

[day_trippr]

Meh. The datetimes are correct, I pulled the plot from my archive, didn't bother checking the date because it doesn't matter a bit...

 

[Unicorn_Platypus]

Yes, the keezer has a 6 port manifold with barbs that frankly I wasn't inclined to replace with an MFL version ($$).
Using a heat gun it took about ten minutes to fit all seven 5mm ID tubing runs to it, NBD.

My three fridges, otoh, were acquired after I had wizened up to the whole flare fitting thing, and each sports a four way MFL manifold. Those got the full Duotight treatment, like these two:

View attachment 656298

I dunk tested everything I could - entire manifolds and all their QD runs - and also did protracted gas tests (charge up, close cylinder, watch valves) and literally went two weeks without any of the systems losing pressure all the way to the QDs. I really am impressed with the PTC performance of all three brands used.

The only thing worth noting is you don't use flare gaskets with these connectors (any of the brands) and instead rely on the deformation of the interior mating surface itself to form a gas tight seal. So you need to put some pressure on the connection - but stop shy of blowing the back out of it. I did manage to find that breaking point on one connector and put a nice circumferential crack in it. Don't do that



Yes, I have my setup configured for ounces (zoom in a bit), but there's a version available that goes metric for those so inclined.




Thank you kindly

I call it my Low Art/High Tech Keezer, and it keeps me off the streets at night

Cheers!

Thanks for the detailed feedback on those MFL connections. I'll make sure not to over tighten and crack them.

So no need to use plumbers tape on the connection then I'm assuming, right?

 

[day_trippr]

Flare fittings seal at their opposing faces; the threads are strictly there to provide compression force.
So, no, just like swivel nuts, no tape should be used...

Cheers!

 

[Noob_Brewer]

Ahahaha! Whoops! I recently took everything out of the keezer and removed the lid assembly for rust remediation, a rinser installation, and changing out all of the beer and gas lines for EVAbarrier tubing and PTC fittings. Apparently I did not put the chamber probes back where they belong so aside from the keg and tower channels they're way out of whack.

Here's what that plot should have looked like:

View attachment 656316

While I don't know exactly how everything would behave if I used a chamber air probe to control the compressor, intuitively I expect much more cycling because the probe response is so quick to air temperature changes that the compressor would run for a very short time then await the next set point crossing - which instead of being 4-6 hours later would be more like minutes.

One of these days I might give it a try and see what happens...

Cheers!

OK, so to revive a slightly old thread, I have a question. I can't tell by looking at this graph. How long is the compressor on during each cycle (approximately) over the 24hr period? 1hr and 15minutes? I imagine it is still better for the lifespan of the compressor to limit the number of times per day it turns on.

Also, I have had the probe simply zip tied to a 12oz can of beer so the probe is tight to the beer can but also taking in ambient temps. My keezer cycled for only about 15 minutes (set temp was 35 degrees with a 2 degree differential) but this was 15 minutes for every hour! So today I FINALLY added insulation to my keezer collar. BOOM! the temp differential from the top of the keg to the compressor hump went from 6 degrees to 1 degree. However, the compressor still is cycling for 15 minutes out of every hour. So to (hopefully) reduce the number of times the compressor turns on I just put a 1"thick dense styrofoam around the probe strapped to the can of beer.

 

[day_trippr]

That plot spans 24 hours, so 5 cycles over that period. The image obviously isn't interactive but the gui it came from responds to mouse-overs on the plots, and those compressor-on periods are 55-60 minutes on average (plot resolution is 5 minutes). So the average duty cycle is between 20-25%. There have been times (read: winter) where it drops to 4 cycles per day.

Using a probe strapped to a keg of beer and well-insulated from the cabinet air temperature provides a huge hysteresis or dampening factor to the control function as can be clearly seen in that plot. Using a smaller volume of fluid for the same purpose will work - but to a much lesser effect. That can will chill and warm much quicker than a keg.

But now you can prove that to yourself

Cheers!

 

[Noob_Brewer]

That plot spans 24 hours, so 5 cycles over that period. The image obviously isn't interactive but the gui it came from responds to mouse-overs on the plots, and those compressor-on periods are 55-60 minutes on average (plot resolution is 5 minutes). So the average duty cycle is between 20-25%. There have been times (read: winter) where it drops to 4 cycles per day.

Using a probe strapped to a keg of beer and well-insulated from the cabinet air temperature provides a huge hysteresis or dampening factor to the control function as can be clearly seen in that plot. Using a smaller volume of fluid for the same purpose will work - but to a much lesser effect. That can will chill and warm much quicker than a keg.

But now you can prove that to yourself

Cheers!

Lol yes I agree with my can not being up to par with a 5gallon keg! But my kegs aren’t lasting that long either. Wondering if strapping it to the CO2 tank inside my keezer would be better than can but still less thermal mass than a 5g keg (which doesn’t stay full long anyways). Thanks for the response! I’ll report back on the performance of the cycles in a 24hr period with it insulated and strapped to a 12oz can

 

[day_trippr]

Of course if you have the room you can use something larger than a 12 ounce can. A two liter soda bottle filled with water comes to mind as a cheap jumping-in point that you could leave there forever.

That is the one hitch with the keg-probe method. While I believe it provides the best control, efficiency, and lightest compressor abuse, it does require eventually swapping the probe/straps/insulation over to a more full keg when the currently monitored one kicks. A large enough vessel assigned to probe duty would solve that...

Cheers!