Help Troubleshooting Temperature Reading Problem

[kgranger]

Hoping to get some ideas of where to start troubleshooting. I'm just going to post everything I'm doing to help get a sense of where to begin looking.

I am trying to cold crash in my Spike CF10 with their temp control coil and a DIY glycol chiller I've built. Glycol temp is set to 28° F.
The chiller is holding it's temp perfectly, and I can verify there is plenty of flow through the insulated tubing leading to the conical.
I checked the coil and tubing and nothing is building any frost.
I am using one of Spike's temp controllers with the probe in the thermowell on the conical, set to 36°.
There have been small stretches of a few hours where the beer is reading +/- 36°, but then out of nowhere it starts building up to ~42°, and with constant cooling, will barely make it down to 40°.
I'll check for frost build up, glycol temp, and flow from the tubing, everything looks good.
I've checked every thermometer and temp probe I have against each other, and they are all reading nearly the same, +/- a half a degree.
I then put the screw-in temperature gauge that came with the conical into the thermowell, and it is reading ~50°, even higher than the probe.
I know they aren't the most reliable, but I do have a Tilt hydrometer in there, and it is reading 36°, which is what I was hoping for.
I even poured out a sample, and checked the temp. It was about 38°, but of course I had to assume it warmed up slightly. But still it was lower than any reading I'm getting in the thermowell.

I don't know if it's possible, but could there be something with the thermowell that is unable to accurately read the internal temp? Like could trub or yeast build up affect it that much?

Maybe something with the controller or the glycol setup?

Any ideas greatly appreciated! Let me know if you want any more clarification on things.

 

[tracer bullet]

Is the location of the cooling and the location of the temp probe far apart? Maybe the beer's getting a ton of cooling in one spot and it's taking a while for that to migrate over for a reading? The tilt is closer and so it reads more like you expect? Something like that...

I'd calibrate the thermometers too. Do the thing with them in ice water and maybe boiling water as well. But ice water at least. Or even just get all of them together and turned on in some room temperature water for starters, see if you can identify the oddball. 50 sounds worn if others read 36. I could see a degree or two or three but not 14.

Might have to wait til after fermentation for a big experiment, and just babysit it for now.

Later on could "ferment" some tap water and open it up and reach in there and check the water temp at various locations.

 

[kgranger]

Or even just get all of them together and turned on in some room temperature water

I tried this, and they are all reading just about the same as each other. Even the temp gauge that was reading 50. I calibrated them all a while back with ice water, they don't seem to have strayed much.

Later on could "ferment" some tap water and open it up and reach in there and check the water temp at various locations.

Definitely going to try this when the batch is done. I'd do it now, but I want to avoid opening and closing as much as possible. Already had to so I could make sure the coil wasn't building frost.

Is the location of the cooling and the location of the temp probe far apart?

I'd say maybe 6-8" apart, not sure exactly without opening it up. Everything with the conical is OEM Spike, so I imagine the geometry "should" be designed to read accurately.