Temperature Monitoring - Seems off?

So i finally got around to setting up my temperature monitor onto COSM, and got two DS18B20's measuring. One is taped to the side of my 8 gallon ferment bucket, then some maybe 1/4th inch foam is over it and that is taped down.

The thing that concerns me is that the beer seems to be fluctuating as much as it is. It seems way too in sync with the room temperature, where i would expect that because of the thermal mass and the fact its in probably another 5-10g of swamp cooler water that it would be moving a lot less than the ambient temperature.

This is my first time doing this, so im not sure if im just being paranoid, or if i do need to come up with a better way to measure the temperature of the fermenting wort, or insulate it more?...i was under the impression that measuring on the side was almost identical to measuring with a thermowell, from reading other peoples accounts on this forum. Granted i guess it is only moving 1 degree, i guess i just thought that with that much thermal mass it would take longer than 10 minutes of my house's heater being on to move 10+ gallons of water temperature by that much, the wort is nowhere near a heating vent.

On a side note it is interesting to see how accurate my house's thermostat is since you can see the cyclical pattern, its set to 64F to hopefully keep my Belgian Hefe in the mid to upper 60's so its pretty close.

https://cosm.com/feeds/83208

Anyone? Looking at a day chart for my Ferm temps, you can obviously see where the house heater is coming on...im pretty sure that my insulation isnt good enough on my probe thats taped to my bucket for the temperature to be moving that quickly and in sync..obviously its a very small amount of change. It just makes it more difficult to determine the actual temperature of the wort when its bouncing around like that...

Either higher insulation, or just lower the accuracy/reporting time of the sensor so i get more of a straight line?

My guess is that it is accurate and the 1F fluctuations are mostly because you have it taped to the side of the fermenter. If you were to imbed it deeper into the liquid, it would level out quite a bit more.
Also, a delta of 1? I really wouldn't worry about it. Remember beer was made before decent temp control

You have a link with your electronics setup somewhere? I'd be curious to see it.

I missed this thread earlier, but the "swamp cooler" part leaves me wondering where the probe is actually located. Is it submerged in the swamp water, above it, or ?

Also not clear exactly how much beer volume and swamp water volume is in play, but even assuming just 5 gallons of the former and 5 gallons of the latter, it is rather hard to believe even that much mass could be moved a degree in 10 minutes by a mere room temperature shift.

So...how thermally stable is this whole project - leaving beer and swamp cooler out of the equation? Is it possible your circuitry is drifting with the room temperature and affecting your data?

Cheers!

[edit] I wrote the above before looking at the plots. Note that there is virtually no time offset between the various significant temperature shifts shown by the two sensors. There's something amiss with either the electronics or the programming, because there is no chance the slopes from the two sensors could be in near perfect sync if everything was working properly...

day_trippr said:

I missed this thread earlier, but the "swamp cooler" part leaves me wondering where the probe is actually located. Is it submerged in the swamp water, above it, or ?

Also not clear exactly how much beer volume and swamp water volume is in play, but even assuming just 5 gallons of the former and 5 gallons of the latter, it is rather hard to believe even that much mass could be moved a degree in 10 minutes by a mere room temperature shift.

So...how thermally stable is this whole project - leaving beer and swamp cooler out of the equation? Is it possible your circuitry is drifting with the room temperature and affecting your data?

Cheers!

[edit] I wrote the above before looking at the plots. Note that there is virtually no time offset between the various significant temperature shifts shown by the two sensors. There's something amiss with either the electronics or the programming, because there is no chance the slopes from the two sensors could be in near perfect sync if everything was working properly...

Click to expand...

This is what i was seeing and curious about. The fact that the plots are so in sync, i am almost positive its because i am reading from the side of my fermenter and not inside the liquid. Because yesterday @ 6pm you can see in the graphs i covered the already covered probe in 2 paper towels folded over like 6 times to make a fairly thick square and it lowered the sway a lot, by about half..it still maps directly of the temp moving when the house heater goes on, but its only moving like .5 degrees...which still seems off but better.
Also fixed the ambient sensor as it had fallen off the wall and was sitting on the floor lol

Its out of the swamp cooler now since fermentation is done and i want to bring it up in temp a bit, but it was in a big one, i'd guess 5+ gallons atleast, the probe was always kept above the water line. Im using a big oblong storage container that probably submerges 3/4ths of the 5 gallons in the fermenter, my probe was in the middle of that upper 1/4th, on the liquid not the krausen.

As for the equipment i just bought an Arduino UNO and Wiznet ethernet shield off of Dealsextreme (www.dx.com) for like 30 bucks, way cheaper than anywhere else just had to wait like a month and a half for it to show up lol. The ethernet shields alone are like 35-40 bucks for the Arduino. Bought some cheap thermostat 4 wire cable @ home depot since its small and flexible like a mini ethernet cable, i didnt want rigid cables, bought some DS18b20's, wired them to the cable and the cable to the arduino pins.

The software is all pretty straight forward, just merged some other projects that read onewire values, with the code COSM publishes to use their library, which is made to work with the Wiznet ethernet shields. So theres no real chance its the software doing anything, all it does is read the value, converts it to F, push it into a cosm struct, and sends it to cosm.

I'd have to agree that the probe itself is being affected by the ambient temperature. If it was the contents of the carboy, it would be on more of a lag than is evident when looking at the two graphs. Add more insulation over the probe!

usfmikeb said:

I'd have to agree that the probe itself is being affected by the ambient temperature. If it was the contents of the carboy, it would be on more of a lag than is evident when looking at the two graphs. Add more insulation over the probe!

Click to expand...

Any ideas? I covered it with a bit of flexible shipping foam my arduino came with, and now with like 6-8 layers of paper towels...im running out of ideas lol. Short of making my own thermowell.

I tape a washcloth that's been folded a few times over my probe with blue painter's tape, and then run some saran wrap around the outside of the carboy to seal it in. However, I've never tracked via COSM, so I don't actually know how good a job it's doing. If you choose to try this method, I'd definitely appreciate you posting the results, so I know how well it works!

Ok, having the probe out of the swamp water was the missing piece that would have made the conclusion that much more evident.

Yes, I totally agree that the insulation wasn't adequate, and probably still is less than ideal.

I have controllers on a keezer and two fridges, and I use Velcro straps to press ~ 4" square pads of 1" thick closed cell foam sheet against kegs and carboys, with the controller probes centered and trapped underneath. If the probes see any significant effect from ambient shifts it hasn't been evident...

Cheers!

One thing to point out is that you're only see about a half degree impact, which isn't too bad, and isn't likely to impact your quality.

usfmikeb said:

One thing to point out is that you're only see about a half degree impact, which isn't too bad, and isn't likely to impact your quality.

Click to expand...

Yea, its not that big of an issue....it could probably be fixed and more 'accurate' by lowering the bits i use and not pushing that sensor data so often, so it would be more of a 'straight' best fit line rather than jumping all over the place every time the temperature changes by .15 degree's lol

So, end result is i could never get it to stop fluctuating with my house heat just trying to insulate it. I could lower it but never it it to stop.
So i basically did a DIY Thermowell for 5-10 bucks from Lowes using the chromed toilet supply line with some aquarium silicone to seal the end.

Now it doesnt fluctuate at all.

Well, at least that totally confirms the theory...

Cheers!

Awesome