Slow Cooker Sous Vide Modification, Learnings

I haven't used our modified slow cooker because I wanted to make sure it had a fail-safe over-temp cutoff system first.  I found some cheap thermal cutoff switches for microwave ovens and bought one that was rated for 120C (from ebay).  I inserted it in series with the heater connection and bolted it to the bottom of the heating unit, which is an aluminum basin that holds the crock pot.  I fired everything up and it went well -- until the thermal switch opened up.  And it stayed open.  Apparently this style of over-temp protection is like a fuse -- once it opens up, it's done.

Rather than buy something with a slightly higher cutoff temperature (and possibly have the same problem), I decided to actually do it more scientifically (about time, eh?).  I have a thermocouple-based thermometer I bought from Harbor Freight awhile back, tested with ice water and boiling water so I know those two points are fairly accurate.  I left the thermal cutoff switch in place, just loosening one of the attachment screws enough to slip the thermocouple underneath it, retightened, and rewired the heater so it was operating in its "dangerous" mode.

I filled the crock with hot (50C) water and set the temperature controller set point to 78C, which should produce 160F in the crock (based on previous characterization work I did on the system).  Before the water got even close to 78C the thermocouple was indicating over 150C!  No wonder the thermal cutoff opened up.

Some investigation revealed some clues as to why the crock and over-temp sensor location were so different, temp-wise.  I loosely crumpled up some aluminum foil into a ball and put it in the bottom of the heater, then dropped the crock in there.  Pulling it back out, I saw the foil ball had been compressed into a "puck" that was almost 1 inch thick!  That means there is a _really_ poor thermal connection between the heater and crock.  This result is not inconsistent with what I initially observed, and which prompted me to move the temp controller's temp sensor into the crock itself, rather than controlling the temperature of the surrounding aluminum bowl.

Just for fun, I replaced that aluminum-foil "puck" in the heater bowl directly above the thermocouple location, put the crock back in, re-filled it with hot water, and re-did the heating experiment.  This time the heater didn't quite make it to 120C by the time the crock got to 78C.  Success!  And, I also noted that this temperature difference was worst-case -- as the setup continued to operate, the temperature difference between the crock pot and heater became much smaller.

I might declare victory here, but the downside is that it takes quite awhile for the crock pot to heat up to the set temperature.  Improving that is problematic, unless I'm willing to do a much-more intensive modification of the slow cooker.  I'm thinking about putting something like plaster of paris or the like, in the heater and dropping the (greased) crock in there.  After the plaster cures, the crock can be pulled out for cleaning/washing, but it will have a much better thermal connection between the heater and crock.  Plaster of paris isn't the most robust thing in the world though, and it will make the whole thing heavier.  I'll have to try the new setup (with a new resettable thermal cutout switch I found) to see if it's usable or not.  A mix of Portland cement with fine-grain sand would be a lot stronger.  Not going there yet!!!

One way to reduce the time to get to operating temperature would be to fill the crock with water that is close to the operating temperature.  This is an easy way to address a number of these issues.

DIY Spectrometer and CBD/THC analysis: Results

In an earlier, lengthy post I described my spectrometer build and showed the spectrum from a CFL lamp.  Since then, I have continued with experiments to see if the dye (Fast Blue B) could be used to produce more quantitative data on CBD and THC content.

My executive summary:  the color shift between CBD and THC is too subtle to use as an analytical tool.

Below, I show two spectrograms -- taken with extracts from two different marijuana clones.  One contains less than 1% THC and one contains no CBD, as determined by my wet chemical analysis (a 5% sodium hydroxide solution in ethyl alcohol turns blue if CBD is present).  The differences are very subtle, making it very difficult to distinguish between the two.

So in conclusion, I made a fairly nice spectrometer I can use for other things, but my primary goal -- being able to determine the relative amounts of CBD and THC in plant material -- was not achieved using this approach.