Sciencemadness Discussion Board - Homebuilt re-circulating chiller pump

Homebuilt re-circulating chiller pump

NERV - 3-5-2008 at 10:23

I have been in the process of creating my own re-circulating chiller pump for general lab use. The chillers at my workplace operate at about 5-10`C which is the range I am currently aiming for. I have no need at this point to create a chiller like the ones at school that can pull -60`C considering that the most volatile things I tend to work with are DCM, and ether (If I need anything cooler than that I will use my dry ice condenser). I am severely limited by finances at this time so all I have to work with is what I already currently have on hand, which is a dehumidifier, some PVC pipe, and a water pump.

The first step will be to strip my old dehumidifier down and remove the compressor, and the heat exchangers. I intended to use this as the continuous cooling supply for the system my only concern is weather or not it will be able to handle the stress of cooling down a large volume of water since it is designed for use in air. I have been unable to dig up any really useful technical data on the dehumidifier aside from the standard manufactures warranty which only tells me not to disassemble the unit, so I have been left to my own devices to test it out. I do know that the unit I own is the largest residential unit that was sold at Wallyworld when I got it, but I have sneaking suspicion that its just a larger bucket and the cooling capacity for this unit as compared to a smaller one is probably the exact same. I have clocked it on my thermocouple (K type) at a steady 6`C although I am sure it could go lower if I removed the thermoresistor although currently I do not plan to do so unless absolutely necessary. My current idea to reduce stress on the compressor is to fill the chiller full of Ice (allowing the Ice and water to equilibrate before powering up) so that all the heat exchanger would have to do was maintain the water temperature in the system. The entire cooling chamber (a 5 gallon bucket) will be heavily insulated, as well as the coolant delivery lines to keep excess heat to a minimum. On paper and in my head this looks like a plausible solution to the problem at hand, although I have been known to be wrong at times.

I am considering wiring my thermocouples temperature probe into the water output line, and then have it send a signal to an automatic shut down switch so that if the unit fails at anytime the reaction will be safe (I am not sure I am entirely confident of this setups reliability yet). I am not sure how to go about doing that yet so any ideas are welcome. Any comments, criticism, or encouragement is highly appreciated. Hopefully in the near future I will have some photos of the completed unit for everyone.

Edit: After allowing the unit to run for an hour, and re-probing it with the thermocouple I have found that the lowest achievable temperature is 3.5`C (better than what I wanted

) in the center of the heat exchanger. I am not sure weather or not to trust my thermocouple readings though since it had a notorious spell of inaccuracy less than 2 weeks ago

……..

[Edited on 22-9-2002 by NERV]

evil_lurker - 3-5-2008 at 12:00

Get a small window A/C unit, yank off the front cover, build a new cover out of acrylic, silicone to the frame, and some way or another rig it up to where you can mount a PC radiator unit inside... thus you have a recirculating cooling unit... either that, or rig up to where you can vent the output into a cooler filled with water.

NERV - 3-5-2008 at 12:30

Quote:

Originally posted by evil_lurker
Get a small window A/C unit, yank off the front cover, build a new cover out of acrylic, silicone to the frame, and some way or another rig it up to where you can mount a PC radiator unit inside... thus you have a recirculating cooling unit... either that, or rig up to where you can vent the output into a cooler filled with water.

If I could afford to get a small window A/C I would be out getting one now, for that matter I would just get a mini freezer or something like that, but all I have is a dehumidifier that I am hoping will be able to keep up with my pump which I have just found out through some disassembly that its cooling itself with the water it will be pumping out from the chilling tank! Now I have to figure out how to re-route the pump, and cool it with its own separate system so it does not raise the temp of the circulating water (this could be a negligible amount in the end, but I’d rather avoid the possibility of failure now rather than have to disassemble it all later). The only thing I am worried about is weather or not the heat exchanger has the necessary capacity to keep the whole unit at a stable temp. If it all works out I will have a re-circulating chiller that I built almost for free from crap lying around my lab!

This is a relatively simple project, but I have never seen anyone document the construction of a re-circulating chiller from start to finish I figure this thread will help some young budding chemist one day get some idea as to how to go about building his own chiller, or at least so I hope. Photos of my progress will be here within the next day or so maybe sooner.

evil_lurker - 3-5-2008 at 14:22

Keep in mind that you need quite a bit of BTUs to really keep up with a condenser.... IIRC about 6K should do it.

leu - 3-5-2008 at 19:43

Quote:

This is a relatively simple project, but I have never seen anyone document the construction of a re-circulating chiller from start to finish I figure this thread will help some young budding chemist one day get some idea as to how to go about building his own chiller, or at least so I hope. Photos of my progress will be here within the next day or so maybe sooner.

Have you studied this:

http://www.sas.org/E-Bulletin/2003-08-22/labNotesAS/body.htm...

NERV - 4-5-2008 at 04:48

Quote:

Originally posted by leu

Quote:

Have you studied this:

http://www.sas.org/E-Bulletin/2003-08-22/labNotesAS/body.htm...

That was a very interesting read Leu with a great explanation of how a refrigeration system works (I have always loved columns from The Amateur Scientist) although I think that the information provided there is best suited to someone who is attempting to build a gas liquefier, or for someone attempting to construct there own cryo storage. I am trying to keep this project simple utilizing things that most anyone could get there hands on easily such as a dehumidifier which has a pre-constructed refrigeration system (or any other item such as an A/C unit, or a refrigerator) of course simple modification can be made to these such as removing the thermocouple that controls the freezing coils temperature, and all you need to do it is some wire cutters. If anyone is up to the task of modifying there refrigeration system beyond that then go for it, I would myself, but I lack the time and money right now to undertake such a task, and I cannot foresee any reason as to why I would need a chiller capable of below zero temperatures at this time. Once I finish the development of my system I am hoping to write a small PDF if time allows which will be posted in pre-publication for the benefit of various members, and non-members who are looking for a general direction to go in while constructing there own chillers. I am aware that there are documents out there that discuss this subject, but there numbers are few and many of them such as the above article are not always within the capabilities of an amateur scientist. A brief search here yields a very limited amount of discussion on this subject.

Evil_lurker I liked the idea of using a PC chiller as the pump in a re-circulating cooler, one could use the thermocouple built inside of it to monitor the temperature so long as one had the software to connect it to a computer. I was on the verge of insanity trying to deal with the pump I had on hand which was effectively killing the efficiency of my chiller because of how hot the motor got (see my previous post). It was consuming 4 30lbs bags of ice every two hours during some pre-trials

! There was simply no way I was going to be able to compete with my pumps cooling needs, and a condensers cooling demands at the same time without the condenser loosing the battle. I am inclined to warn anyone else out there considering constructing there own re-circulating chiller to avoid at any cost pumps that are submersible because the damn things will put out more heat than you will be able to effectively cool. After looking at this thread I realized I had an old Aquagate PC cooling pump buried in my spare parts box, I just didn’t have any of the software or the PCI card necessary to use it on my computer so I ripped the pump out of its housing, and its working great in my chiller it just leaves a little bit to be desired in its ability to overcome gravity, but in order to accommodate the new pump I have decided to mount the chiller on a shelf overlooking me and Fleakers workspace so I can use the force of gravity to compensate for the pumps inability to pump water any higher than about 2 feet up (I actually think mounting the system up high as compared to a floor model is much more efficient since you are mostly using gravity to move your water instead of pump power). As for the BTU rating on the freezing coil in my dehumidifier I could not tell you, but I am willing to bet its not quite 6K. I was planning on calculating the amount of heat it moved over time using a bucket of water at a pre-set temperature then timing how long it took it to cool it down to ~20`C. I may get lucky if I can manage to convince my younger sibling to give up his 5.5K BTU window A/C unit since he has not used it for well over a year now. I must say what was supposed to be a simple afternoon project is getting mighty complex…….

[Edited on 22-9-2002 by NERV]

chemrox - 4-5-2008 at 05:35

The range of 5-10*C can probably be achieved with a radiator. I saw an oil radiator at a shop that would fit nicely in a window with a fan behind it. Add a slow garden-fountain pump and you're in business. The SA article descibes a setup that liquify ammonia! Nice but overkill maybe for your purposes? Anyway I enjoyed and saved the SA material- I may use it too. Thanks leu.

NERV - 4-5-2008 at 08:48

A radiator could possibly work so long as the area that you live in is either really cold or mild weathered throughout most of the year. I think if I tried a radiator around here this time of year it may start to boil my product in the condenser

.

I realize now that this topic is probably better suited for the Reagents and Apparatus Acquisition board so if a Moderator feels inclined to move it they are more than welcome to unless you happen to want it to stay here………..

LSD25 - 4-5-2008 at 13:43

I was actually putting some thought into this exact topic the other day, however, I was approaching it the other way - rather than using solid CO2 (expensive) or a pump (needs cooling, complex plumbing & electrics, etc.) would it be possible to utilize gaseous CO2 as the cooling system? Gaseous CO2 is fucking cheap, provided you vent the outgas properly it is non-toxic, and when it is allowed to expand it absorbs a LOT of heat. It is also very easy to purchase in bulk, it is available anywhere there is soft-drinks.

I was thinking perhaps using the gas through a thin copper line (with pressure fittings from the tank), passing this into a large diameter cooling coil, then exhaust. There would be no heat generated - the only time heat is generated is when you try and condense the gas again. There is no need for complex plumbing, electrical wiring, moving parts, etc. It would be interesting to see what the minimum temperature which could be reached using this system (for the doubters, turn on the nozzle of a CO2 cylinder and let it run out a window - feel the outside of the cylinder & also the fitting). The drop in pressure of the gas absorbs heat (or something like that).

This probably wouldn't be as cheap to operate as a recirculting system, but for simplicity of operation, ease of construction and quite probably performance, it would probably outperform most that could be built at home. Apart from that, as stated above, full tanks of CO2 ain't precisely expensive. The ulimate performance could probably be tweaked by using the CO2 to make 'CO2 Snow' in-situ, then allowing that to evaporate.

Attachment: EvapCooling.pdf (447kB)
This file has been downloaded 948 times

NERV - 4-5-2008 at 22:11

Using compressed CO2 to cool something down is a definite possibility, but that is more like constructing your own refrigerator. IMHO it would be a serious pain in the ass to try to hook a high pressure CO2 line to a condenser unless you custom built a CO2 powered condenser or the likes. For the most part your average chemist just needs something that will supply cold water continuously to whatever experiment he may be performing. I cant imagine anything easier than picking up a window A/C unit, dehumidifier, refrigerator, or whatever, as well as a 5 gallon pail, and a water pump throwing it all together and BAM there you have it a continuous supply of cold water all the time no need for re-filling a cylinder, or running out of CO2 during your experiment you just have to plug it in to the wall and there you have it stable efficient cooling for all your lab needs (so long as there are no blackouts :p) .

Currently I am facing some setbacks with the chiller I am currently constructing. I was able to acquire for free a 5050 BTU/Hr window A/C unit from my younger brother that was in perfect working condition. Everything was going as planned with the construction until I went to install the freezing coil on the bottom of my water cooling tank. As I adjusted the copper line that led the vaporized Freon back into the compressor I accidentally cracked one of the Aluminum lines inside the cooling coil causing a wonderful gush of Chlorodiflouromethane (R-22) to burst from the system! Arg now I have to go and fork money out to some A/C repair guy to re-fill my A/C unit because I just can’t bring myself to go back to the dehumidifier after experiencing the cooling power of a good window A/C unit. In a comparison of the two it took the dehumidifier about 4 minutes to reach its maximum coil temperature of ~6`C (with another hour it was able to get to 3.5`C), while the A/C unit was able to get down to a minimum temperature of -25`C in less than 30 seconds

! That right there alone sold me on A/C, but even better is that A/C units actually give you all of there specifications unlike the dehumidifier allowing me to calculate that the A/C I was using was theoretically (in a perfect world) capable of cooling 5.6 gallons of water at a rate of 1`C a minute that may not sound like a whole lot, but when you look at the large heat capacity of water (4.18 J/g) that’s a significant amount of heat being removed! Since the A/C is capable of sub zero temperatures I could use propylene glycol as the cooling fluid for reactions that need to be super cooled, even if I don’t have plans to do anything that cold its nice to know that I could. So to hell with using dehumidifiers Evil_lurker had it right from the get go use window A/C’s! I finally have some images of my project uploaded to my computer, but I still need to sort, and re-size them before I can get them up. Right now I am feeling abnormally tired though so I will have to put them up later on today when I wake up.

Ahh to hell with it I will at least attach an overhead shot of the apparatus halfway through construction.

[Edited on 22-9-2002 by NERV]

100_4924 small.jpg - 105kB

S.C. Wack - 4-5-2008 at 22:15

Something missing here is what sort of scale we are talking about needing. A window AC, that's a lot of cooling. And a lot of heat in the lab on the high-pressure side. 60 pounds of ice an hour?

evil_lurker - 4-5-2008 at 22:38

By gollly, I do believe that is a $96 walmart special right there gutted out.

MagicJigPipe - 4-5-2008 at 22:49

Uhhhhh... Just blow the heat out of a window like the A/C unit was designed to do.

BTW, LSD25, evaporation removes heat. Just like when you feel cold after you get out of the shower, the CO2 is doing the same thing at a much faster rate.

The CO2 snow isn't very efficient because the heat must be removed from the CO2 by evaporation. So, basically, you are using some of your CO2 to remove heat from the rest of your CO2 to -109*F. However, like you said, CO2 is cheap and this inefficiency isn't such a big deal.

I was thinking. If you could get a coolant liquid down to, say, -20*C. Even though that is not cold enough to condense ammonia perhaps you could use a mixture of the coolant and dry ice to decrease the time it takes for the CO2 to sublime by an order of magnitude.

It would be REALLY nice to construct a chiller that could get the coolant around -30*C so one could condense ammonia in a normal condenser. This is exciting!

NERV - 4-5-2008 at 23:47

Quote:

Originally posted by S.C. Wack
Something missing here is what sort of scale we are talking about needing. A window AC, that's a lot of cooling. And a lot of heat in the lab on the high-pressure side. 60 pounds of ice an hour?

I am not to terribly concern with the excess heat that will be generated on the high pressure side, there is allot of space in the lab for it to dissipate to, and I am not the one responsible for the air conditioning bill

. To control the amount of heat removed from the coolant I am going to be using a universal PID controller with a K type thermocouple that Fleaker, and I bought some time ago for a different project that keeps being pushed back further, and further. All I have to do is connect it to a relay in series with the window A/C’s power supply in order to keep the temperature of the coolant at a pre-set value since I am not too terribly interest in a coolant reservoir that is just a solid block of Ice. That should also help reduce the amount of heat expelled by the high pressure side. I am very seriously considering using either ethylene, or propylene glycol as coolants eventually since the window A/C is clearly capable of at least -25`C max. Even though I don’t really need this much cooling power I might as well utilize it for all it is worth since so far this chiller hasn’t cost me a single penny (aside from having to go refill the A/C with R-22 later on today which was my fault). And as a convenient side benefit I no longer have to worry so much about the pump heating up the coolant because the A/C will be able to handle it :cool:

Quote:

It would be REALLY nice to construct a chiller that could get the coolant around -30*C so one could condense ammonia in a normal condenser. This is exciting!

Looks like the thermocouple reading on the A/C is getting pretty close to your goal there, I am thinking that the coolant will only get to within a few degrees of that mark in reality though.

[Edited on 22-9-2002 by NERV]

[Edited on 22-9-2002 by NERV]

100_4912 small.jpg - 104kB

LSD25 - 5-5-2008 at 02:04

Not so much looking at using it for a condenser, more like this sort of setup:

http://www.asynt.com/product_ranges_view.php?ID=29

A thin capillary line carrying CO2 from an inverted CO2 cylinder into the unit - where it expands & evaporates and then is exhausted through a much larger line (say a magnitude of 10-20X).

Using something like this sub-zero reactions should be rather easy, although I must admit NERV has something going on here, if only this can get down to the requisite temp's, I for one would be more than happy to gut and modify a cheap room airconditioner. The uses of such a system are manifold - vacuum traps, liquification of volatile liquids, low-temperature reactions, etc (I was thinking something like a dewar-type vacuum trap rather than a standard condenser).

PS I remember from past-experience that similar ideas are used by industry in the provision of extremely dry air for clean-room environments and extra-touchy pneumatic instrumentation. They add in a number of filters, pneumatic type, which cause any condensate which is present to collect in a trap from whence it may be removed rather simply.

[Edited on 5-5-2008 by LSD25]

tentacles - 7-5-2008 at 09:29

Quote:

Originally posted by MagicJigPipe
It would be REALLY nice to construct a chiller that could get the coolant around -30*C so one could condense ammonia in a normal condenser. This is exciting!

Just come up here in Canada in the winter and run your coolant through a radiator outside... -36C actual temp isn't too uncommon here in the plains.

-jeffB - 7-5-2008 at 10:28

Quote:

Originally posted by NERV
To control the amount of heat removed from the coolant I am going to be using a universal PID controller with a K type thermocouple that Fleaker, and I bought some time ago for a different project that keeps being pushed back further, and further. All I have to do is connect it to a relay in series with the window A/C’s power supply in order to keep the temperature of the coolant at a pre-set value since I am not too terribly interest in a coolant reservoir that is just a solid block of Ice.

Does your controller impose a lockout cycle so that it doesn't cycle the power off and back on too quickly? As far as I know, quick-cycling is still a reliable way to kill a compressor.

unionised - 7-5-2008 at 10:43

I just love this quote from that amateur science article ". Methyl chloride, also formerly used in domestic refrigerators, is noncorrosive and nontoxic, although it can interfere with respiration if it is inhaled in large amounts."
It seems to have changed somehow these days.
http://msds.chem.ox.ac.uk/CH/chloromethane.html

MagicJigPipe - 7-5-2008 at 17:26

Quote:

Originally posted by tentacles
Just come up here in Canada in the winter and run your coolant through a radiator outside... -36C actual temp isn't too uncommon here in the plains.

Wow, it would be so cool to see liquid ammonia, chlorine, methyl chloride, dimethyl ether or 1,1,1,2-tetrafluoroethane sitting outside, not boiling.

undead_alchemist - 7-5-2008 at 17:39

You know you can refill it with propane.
It works very close to the properties of R22

Myself, I have lots of compressors laying around same goes for coils. Have one compressor that is rated for 20K BTU, have another that is 5K BTU.

Propane also know as R290 works very well.
Mixed with ISO-Butane will work as a replacement for R12 or R134a.

Quote:

Originally posted by NERV
Using compressed CO2 to cool something down is a definite possibility, but that is more like constructing your own refrigerator. IMHO it would be a serious pain in the ass to try to hook a high pressure CO2 line to a condenser unless you custom built a CO2 powered condenser or the likes. For the most part your average chemist just needs something that will supply cold water continuously to whatever experiment he may be performing. I cant imagine anything easier than picking up a window A/C unit, dehumidifier, refrigerator, or whatever, as well as a 5 gallon pail, and a water pump throwing it all together and BAM there you have it a continuous supply of cold water all the time no need for re-filling a cylinder, or running out of CO2 during your experiment you just have to plug it in to the wall and there you have it stable efficient cooling for all your lab needs (so long as there are no blackouts :p) .

Currently I am facing some setbacks with the chiller I am currently constructing. I was able to acquire for free a 5050 BTU/Hr window A/C unit from my younger brother that was in perfect working condition. Everything was going as planned with the construction until I went to install the freezing coil on the bottom of my water cooling tank. As I adjusted the copper line that led the vaporized Freon back into the compressor I accidentally cracked one of the Aluminum lines inside the cooling coil causing a wonderful gush of Chlorodiflouromethane (R-22) to burst from the system! Arg now I have to go and fork money out to some A/C repair guy to re-fill my A/C unit because I just can’t bring myself to go back to the dehumidifier after experiencing the cooling power of a good window A/C unit. In a comparison of the two it took the dehumidifier about 4 minutes to reach its maximum coil temperature of ~6`C (with another hour it was able to get to 3.5`C), while the A/C unit was able to get down to a minimum temperature of -25`C in less than 30 seconds

tentacles - 7-5-2008 at 19:35

Yep, the propane works well. The stuff in most cans isn't necessarily pure, though, you really want refridgeration grade. I used "Enviro-Safe" to charge my in-laws' house aircon, and it worked well. Reportedly, better than it ever did on R22.

It's also pretty cheap.. Just make sure you patch up that hole you made. If it's in one of the lines you could cut it and repair with a compression coupler or something. If it's in a coil, good luck!

Some aircons are R134a (newer ones especially). You can buy R134a in convenient refill cans at walmart, cheap. You will most likely need to cut off the fill stem (should be on the compressor) and solder on a fitting of some sort to use to refill the unit. Ordinary lead solder is fine for this.

MagicJigPipe: A look is about all you'd want, then you'd go back inside. I've considered setting up a radiator in my window to cool my condensor when I distill nitric, but it's easier to just use the effluent from my vacuum aspirator. When your tap water is 1C that's not such a bay way to go.

[Edited on 7-5-2008 by tentacles]

MagicJigPipe - 7-5-2008 at 19:52

Quote:

Propane also know as R290 works very well.
Mixed with ISO-Butane will work as a replacement for R12 or R134a.

I've read that some air conditioning systems will have bad problems when filled with hydrocarbons (not to mention flammability that it wasn't designed to handle). I believe it was about vehicle A/C units but I would still be careful.

ShadowWarrior4444 - 9-5-2008 at 14:33

A heatpump/cryocooler that may be very interesting to home chemists attempting to liquefy gasses (including air) is the Stirling engine:

http://en.wikipedia.org/wiki/Stirling_engine

"At typical refrigeration temperatures, Stirling coolers are generally not economically competitive with the less expensive mainstream Rankine cooling systems, even though they are typically 20% more energy efficient. However, below about -40 to -30 deg.C, Rankine is not effective because there are no suitable refrigerants with boiling points this low. Stirling cryocoolers are able to "lift" heat down to -200 deg.C (73 K), which is sufficient to liquefy air (oxygen, nitrogen and argon). They can go as low as 60K - 40K, depending on the particular design. Cryocoolers for this purpose are more-or-less competitive with other cryocooler technologies. The coefficient of performance at cryogenic temperatures is typically 4-5%.[2] Empirically, the devices show a linear trend, where typically the COP = 0.0015 × Tc - 0.065 , where Tc is the cryogenic temperature. At these temperatures, solid materials have lower values for specific heat, so the regenerator must be made out of unexpected materials, such as cotton."

It should be fairly easy to build for someone with a bit of talent in mechanical engineering. Its higher efficiency would also likely be a boon to those not wishing to cascade vapor-compression coolers to get the required liquification temperature.

[Edited on 5-9-2008 by ShadowWarrior4444]

LSD25 - 19-6-2008 at 22:07

Here is an interesting idea:

http://en.wikipedia.org/wiki/Gas_absorption_refrigerator

Now they use a strong aqueous solution of ammonia in a system which is charged with hydrogen, the idea is that by heating the solution some (most) of the ammonia boils off - it is then condensed at a high pressure to a liquid - this travels to the evaporater where it mixes with gaseous hydrogen, which lowers the vapor pressure and allows it to evaporate. The ammonia/hydrogen mixture is then allowed to come into contact with the low ammonia solution, which absorbs the ammonia (but not the hydrogen) and is then ready to start again

http://www.arizonaenergy.org/AltEnergyClub/SMALL%20AMMONIA%2...

http://www.me.gatech.edu/energy/andy_phd/

Now, I was thinking, how would one go using a diffferent inert gas which will not absorb in water - like argon/helium/nitrogen (gaseous hydrogen is also on the unobtainium list - it can be generated yes, but not at a decent pressure)?

This would work particularly well if it were used with a hot water line - use the water line to provide the heat which causes the ammonia to distill off (check out the design of the water circulation pump or bubble pump - they are essentially using the water left in a bump trap to transfer some of the water to the absorption section of the design).

The other trick is that it appears that the inert gas is kept confined to one section of the pump - dunno, maybe a non-return valve from the bottom of the compressor (ie. under liquid level) and a similar setup from the bottom of the absorbtion area (also under the liquid area) to the boiler?

Would be far simpler to construct, especially if other gasses could be used.

YT2095 - 20-6-2008 at 00:58

how about going to a scrap yard and ripping the AC unit out of a car?
I`m sure it would have lots of useful bits and be Really Cheap as well

MagicJigPipe - 20-6-2008 at 21:37

I have noticed that the only problem with vehicle A/C units is the compressor. It is run by the same belt in the motor that turns the alternator etc. Basically, to use it you would have to have an electric motor with a belt. Also, they use a magnetic clutch to (de)activate. That is so it's not always running whenever the engine is in use.

It would take up a lot of room to set it up but would be interesting...

Mr. Wizard - 20-6-2008 at 22:28

Here is a guy that made a quick and dirty refrigeration unit using propane.
http://home.earthlink.net/~lenyr/refrig.htm
He then went on to modify a refrigerator using propane. Lots of pictures and practical information.
http://home.earthlink.net/~lenyr/refrig2.htm

F2Chemist - 11-7-2008 at 12:49

Although it doesn't have even CLOSE to the refrigeration power as a common recirculator, an inexpensive option one could use for something as simple as a cold condenser finger is this:

http://www.premiumaquatics.com/Merchant2/merchant.mvc?Screen...

Also, keep an eye out in the aquarium section of your local classifieds...you can sometimes pick up an inexpensive probe or recirculating chiller used for aquariums. These are anywhere from 1/10 hp to 1 hp models.

[Edited on 11-7-2008 by F2Chemist]

peach - 1-5-2010 at 02:01

I have an even more basic suggestion.

Go to the local dump and salvage a freezer. Remove all the drawers and compartments.

Install into said freezer a large bin that uses up as much room as possible. Fill the bin with water and some anti-freeze.

Drop a recirculating pump into the bin. Drill two holes in the freezer and connect your in/out lines to the pump.

If my memory is working right, a normal household freezer can manage up to -18C.

The compressor won't be able to keep up with solvents that have a high latent heat capacity, but the huge thermal mass of the water in the bin will mean the temperature won't change all that much during an at home scale distillation.

You could even improve the thermal mass of this idea further still. Drop a bunch of sealed ice packs into the bin first, or even 2l coke bottle full of tap water would do. Add enough water for your pump to function. The change of state occurring as either of these two melts will soak up huge amounts of energy before the bin starts to warm up - and forms a mirror image of the state change occurring in the condenser.

It should also form a nice contained system that doesn't need reboxing to stop it getting broken when it takes a knock. It'll take zero intelligence or effort to put it together and be virtually free.

And now you have somewhere to stick all your stir bars.

Branded recirculating chillers are ridiculously expensive, yep.

[Edited on 1-5-2010 by peach]