Tilt/Grainfather temperature anomalies!

[Isolec82]

Just getting back in to brewing after some time away. A few days ago, I put a Pilsner into the GF fermenter and dropped in a tilt, just as I have done many times over the past few years. I am using a Gcast to send readings to the GF app to monitor and control the brew. This time I am noticing between 4 - 5 degrees Celsius difference between the tilt temp and the conical fermenter temp values. Now I am guessing this difference has always been there in the past and I had never noticed. Historically I had always used the temp sensor in the GF fermenter to control the brew. I know that the tilt is floating on the top and the GF sensor is low down in the fermenter. My gut reaction was to go with the tilt, but surely GF have given thought to the positioning of their sensor. I am supposed to be using using M84 at 12c, and this margin of error potentially takes me out of the sweet spot for this yeast. Anybody have any thoughts.

 

[peebee]

Wretched isn't it? We've been so cock-sure the thermometer is telling the truth and then we go and put more than one thermometer in it.

One thermometer and everything is peachy.

Two thermometers and I guess the beer must go down the drain.



Try putting the fermenter in a conservatory/greenhouse. Things get really kooky then (fermenters wearing sun-hats?).

 

[The-Engineer-That-Brews]

Well the temp sensor in the Tilt is probably a band-gap digital reference, so I’d tend to trust it.
Presumably the GF sensor is a thermocouple, however it still should be fairly accurate. How’s it located, in a thermo-well? If so, could it be buried in the sediment?
What strikes me is that if the brew is actively fermenting then you may have convection currents in it that could cause it to be hotter on the surface and colder where it’s in contact with the walls of the vessel - but which reading is higher?

 

[Isolec82]

Just thrown a second tilt in and it shows a 3rd reading, 1c lower than the other tilt (an even bigger difference). I can live with the 1c difference in the tilts - probably calibration changes. I am starting to believe the tilts but surely the GF30 should be more accurate than that. I have brewed 60+ successful batches in it using its temp settings that appear to be 4-5 degrees lower than i thought. It is only 3 days in so there will not be a great deal of sediment yet to surround the thermocouple. It has started and dropped from 1.045 - 1.043. I had considered convection currents, but dismissed it because the lower temps were at the top!!

 

[Isolec82]

Interestingly the GF fermenter controller allows significant temperature calibration changes almost as though they anticipate the problem.

 

[peebee]

The Tilts themselves are not calibrated except at the factory, The client reading the Tilt has the configuration. It's multi-point configuration and as accurate as you wish to make it (within the tight limits of the device ... well within 0.5C). This is an oldish trace (July 2022) from a TiltPRO (slightly better resolution) via a Raspberry Pi (TiltPi software). The saw-tooth pattern (in red) illustrates why it needed a sun-hat! The Grainfather fermenter's thermometer does the temperature control (very effectively, but it's not in a thermowell):

My opinion is it would be plainly daft attempting to control the temperature from a Tilt (floating on the surface) ... "accuracy" of reading has little to do with it.

 

[Isolec82]

Peebee, I take what you say, but the values given by the GF30 are now very erratic and some 8c different to two tilts with the same value in the brew. I have historically always used the GF to control temp. It is because of the anomalies that i went to tilt control on temp. On reflection, when i did a 3c cold chill on my last brew, the GF froze it. See pic. My thoughts are now coming down to software error in the GF controller. I will spk to Grainfather.

 

[HockleyPete]

The Tilt is sitting on the surface of the beer there's going to be a difference between that and the built in probe.

I've got a similar effect on the Kegland Bucket Buddy as the probe is down at the bottom.

 

[peebee]

Peebee, I take what you say, but the values given by the GF30 are now very erratic and some 8c different to two tilts with the same value in the brew. I have historically always used the GF to control temp. It is because of the anomalies that i went to tilt control on temp. On reflection, when i did a 3c cold chill on my last brew, the GF froze it. See pic. My thoughts are now coming down to software error in the GF controller. I will spk to Grainfather.

Think what you're doing!

At 4C thermal convection currents go into reverse, and so there isn't much (any even) thermal convection for the Tilt to react quick enough to the impending freeze. "Cold Chilling" (hey, thanks for avoiding my trigger phrase! All that gawd awful "cras... ing") to below 4C with a floating thermometer probably doesn't provide enough time for the Tilt to detect the encroaching inevitable.

But don't let that stop you complaining ... Tilt are very good at replacing items which only have a hint of being unworthy.

 

[Lunesdale Brewer]

Extract from Conical Fermenter Pro Instruction manual:

Temperature distribution changes as it cools. Therefore, when cooling at low temperatures, a reading that’s taken at the top of the fermenter with an additional probe, such as a Tilt hydrometer, may be different to a reading that’s taken in the cone.

My RAPT Pill reports 6 deg C when Brewfather controller reports 4 deg C.

I assume that stratification is happening as the more dense solution at 4 deg C falls to the bottom of the vessel where the GF sensor is located and switches off the cooling pump.

Both devices are working correctly.

 

[JockyBrewer]

I tend to chill the last few degrees to pitching temperature in my GF30 using the glycol chiller. I've found that when it gets down to pitching temperature according to the GF30 the iSpindel at the top of the wort will read quite few degrees higher (e.g. 14C on the GF30, 20C on the iSpindel), but the temperatures will eventually converge over 12 hours and then they usually agree within 1C.

I'd expect that with a bigger ambient temperature delta that they might not agree quite as well, and if you're not using the internal cooling jacket to chill then that would also change things quite significantly.

Ultimately though you need to try things out, see how they work out in the finished beer and then adjust in future. More data can be confusing at first, but it can also help you figure out what has gone wrong/right in a beer.

 

[hoppyscotty]

Well the temp sensor in the GF conical fermentor is at the bottom in the cone which is heated. It sees the change in temp before the beer above it does and gets covered in trub that is bound to have an insulating affect. The tilt is floating on top and partly submerged in the beer and a little bit out. So you're getting the two very extremeness of temperature there. The real beer temp is going to be somewhere in between most likely.

In our GF conical we just let it ride and rely on the GF temp control, ignore any numbers coming from the iSpinedel and just use that for gravity trend monitoring and all is well. In reality how accurate do you need to be...as long as your fluctuations are in control i.e. +/- 1 degree say, then you should be all good. Is it so disastrous if you want to ferment at 20 degrees and you actually ferment at 19 degrees or 21 degrees so long as the temp is controlled??