Mini fridge modification to improve accuracy?

[KMoffett]

I am not a refrigeration person, so thus the inquiry!!!

I have a researcher that wants to store a small volume (I think 1/2L) of a chemical at 4°C +/-1°C. He has heard of people supposedly using a mini (dorm) refrigerator with an external controller (Honey Run Apiaries) with a probe in the compartment. My first question is about the accuracy of the unmodified mini refrigerator...? The other is, if the normal thermostat is replaced with the controller, will there be problems...like frequent cycling the compressor...? I'm assuming I would have to add a circulating fan to minimize a thermal gradient.

Out of my depth here.

Ken

Also, he wants to do it on the cheap.

 

[MrAl]

Hi there Ken,

It would be very hard to specify a run of the mill mini frige i think because they probably vary from manufacturer to manufacturer, and the hysteresis would vary too i would think, and not be very dependable for very tight control.

Will there be frequency cycling...the tighter the control band is, the more frequent the cycling. If this worries you then you can think about introducing some controlled loss, where you might even consider a small heater in conjunction with the cooling. I guess it depends partly on how much energy you can get away with using for this thing. Also, a big block of something helps to maintain temperature, such as a big block of steel acting as a heat capacitance.

Circulating fan...well, if you want to minimize gradients what you could do is build up a very simple sub chamber, where you have a good heat conductor chamber inside a bad heat conductor chamber, with the sample held inside the good conductor chamber. The idea there is that any drafts have to first get through the insulator, then once they do get through whatever makes it has to distribute through the good conductor. The whole thing could be a thick wall metal can inside a styrofoam enclosure like a cheap cooler. A very thick wall would help reduce cycling also.
The thermo probe would of course be inside next to the sample, or you could use several probes along the height of the sample depending on how fussy you are.
Testing of course would be a must, to make sure that the temperature stays within the predetermined levels.
I would bet you could get away with thermistors with the right circuit as long as there was a calibration phase followed by periodic re cal.

I would ask how long the sample has to be kept like this...days, weeks, months, years, etc., and possibly make some mods depending on if it is going to be a very very long time (cryo for example).
I would also ask what are the consequences if something goes wrong...this would help to define what level of reliability we would have to build into the system.
I would also ask how long the sample can be subjected to indifferent heat levels, for example when the sample is transferred into the chamber.

BTW if you are good with electronics, then you are good with heat too...simply transpose variables where heat conduction is current, temperature is voltage, etc. We could actually draw up an electrical circuit describing this systems heat profile.

 

[Reloadron]

A few years back I gave one of our new young engineers a nice Omega PID controller out of a scrapped environmental chamber. He was using it for a kegerator design to keep beer at a tight temperature tolerance. I know it worked out well. When I see him on Monday I'll ask him just how well it has worked out, I know it was up and running but don't know how tight of a temperature it maintains the beer at after stabilization. Plus or minus one degree C is pretty tight for a converted fridge though.

Ron

 

[Mr RB]

Providing your thermostat produces a reliable average temperature you can just use some thermal averaging (lag) by putting the target in a thernally insulated enclosure (even wrapping it in bubble wrap etc).

That will damp the thermal variation caused by the thermostat cycling.

I would also put a remote thermometer in the target and run a wire out to external temp display, to help fine tune it.

 

[KMoffett]

Off work today because of snow. The concept is to keep a sample liquid at 4°C until it's put into an apparatus that is held at 4°C by a precision refrigerated water bath circulator and heat exchanger coils. The apparatus is not always on so I can't use that circulator. I found a perfect circulator on the internet for ~$180 with free shipping, but he wants to go with the "cheaper" mini fridge idea. He was supposedly going to have someone contact me that he thinks has done this.

Ron, I would be interested in your friends experience.

RB,
If I have to go with the mini fridge, I'll follow your recommendation for trials....and ask a lot more questions.

Ken

 

[Reloadron]

I asked Blake and currently the kegerator maintains the beer at +/- 2 degrees F with a SP of 38 deg F. Keep in mind that is just a keg of beer with the TC stuck to the side (center level) of the beer keg. Actually sort of humorous that he is using a pretty expensive PID loop controller to maintain beer. With tuning it could likely do better.

Ron

 

[KMoffett]

Thanks Ron! +/- 2°F = +/- 1.1111°C. A keg is a huge thermal mass. I suppose we could surround the sample with box wine bags filled with water. I do that in my home freezer to reduce empty space.

Ken

 

[Reloadron]

Thanks Ron! +/- 2°F = +/- 1.1111°C. A keg is a huge thermal mass. I suppose we could surround the sample with box wine bags filled with water. I do that in my home freezer to reduce empty space.

Ken

That sounds good. Yeah, a keg, even low to empty is a large thermal mass which weighs in on the stability.

Ron

 

[Mr RB]

Those small electonic fridges with peltiers etc use electronic thermo sensors and keep a surprisingly constant temperature, provided the external ambeint temperature isn't too redical.