Getting stuff *really* cold is difficult, dangerous, and lots of fun. Our group is preparing to send an autonomous glider up to about 100,000 feet, so we needed a way to chill our electronics to -70c to test their use before these edge-of-space flights.

Basic concept:

Just using dry ice in a cooler will result in about 0c air. In order to chill air much colder, a fan is used to circulate the air to reach -42c. In order to hit -70c or colder, liquid nitrogen and a heat exchange (coil) is used.

This test chamber uses dry ice and liquid nitrogen to cool the air.

This instructable is part of our edge-of-space project. Watch for updates over on Hackerbot Labs

Step 1: Acquire parts for building

You'll need a bunch of parts to build a test chamber, along with some supplies. Here's what we used:

  • 25' of 3/8" copper pipe (hardware store)
  • Fittings to adapt up to 1/2" ID pipe (hardware store)
  • 2' of 1/2" pipe to use as a reservoir
  • Fiberglass wrap insulation for reservoir
  • Teflon tape for the threads (hardware store)
  • 4" diameter PVC pipe (just for forming the coil, about 2' will do)
  • A large ice cooler which you can remove the lid from.
  • 2" thick pink insulation foam. a full sheet is handy, but it needs to be the size of your cooler
  • A powerful fan which can fit in the ice cooler for air circulation
  • One or more thermal probes for monitoring the temp (K type thermocouples work best here)
  • A large cooler/ice chest with a lid which can be completely removed.
  • "ice" packs. The sealed kind you reuse. A bunch of them
  • A funnel
  • Some stiff plastic tubing

For operation of your test chamber, you will need the following consumables:

WARNING: This stuff is cold - so cold that it will burn you. Seriously. It can blind you. It can freeze off a finger or a limb, and I'm not even going to get into what would happen to your insides if you drink it. Use proper care while handling. Eye protection is very important while interacting with liquid nitrogen. Ventilation is extremely important as well. USE AT YOUR OWN RISK.

  • Several blocks of dry ice. 10 lbs will do, 20 is better. This can be sourced at most grocery stores. Some states require you to be 18 or older to purchase and consume Dry Ice.
  • a large dewar of liquid nitrogen. This can be sourced at most welding supply shops. Dewars can be rented or borrowed.

Step 2: turn your coil..

Start by dry-fitting all of the fittings, ensuring all the threads line up and have the correct gender and size so you can get up to the diameter of the small end of the funnel.

Next, it's time to make the coil. Copper pipe is stiff and can kink very easily. Kinks are not desired. Using a large cylindrical object of some kind (I used 4" diameter PVC pipe), 'turn' the copper pipe around the object to get a nearly perfect coil without any kinks. This requires much hand/arm/upper body strength, and may require a helper or two, depending on the wall thickness of your copper coil. It's important to leave extra at each end of the coil for connecting adapters and allowing for venting from the other end of the coil. You should end up with something that looks like the below photo.

Next, install all of the fittings on the end of the copper pipe to adapt it up to a larger size. Caution, the press fittings on the copper pipe are a pain to install, and it can be tricky to get a good fit. Don't over tighten the press fittings, or the threads will fail. It will still be somewhat 'loose' on the fitting, but should be water and air tight. Compression fittings are not designed to be load bearing, so don't stress the system too much or it will break or kink.

We created a reservoir by using a 2 foot length of 1/2" pipe; this extends the after the fittings. don't install this section of pipe until you've completed your lid..

Step 3: Prep the cooler

We found this rather large budwiser compliant cooler at a local grocery store. It claims to keep 10 cases of beer cold for 10 days. It was selected due to the ease at which the lid was removed, and the size of the internal volume. Due to the size of the coil and circulation fan, much space is needed. Make sure your cooler is dry and clean. Having water or moisture in the cooler is bad.

Start by removing the lid on your ice chest/cooler by whatever means required. In our case, it was 5 screws. We will be building a replacement lid for the cooler, which we can then drill holes into. You could, in theory, use the stock lid and drill holes in it. This reduces it's effectiveness for cooling beer later, unless you like your beer at -70c.

Next, cut and form your pink insulative foam to fit the recess left by the removed lid. Cutting and shaping foam is messy prospect; the chips get stuck everywhere, have a shopvac handy. Cutting foam can be done with a skillsaw, handsaw, or even a table saw. If you happen to have a hotknife made out of nichrome wire, it cuts foam rather effectively. For final shaping, simple 100 grit sandpaper works best. The goal is to create a new lid which needs to be 'press fit' into the cooler. It should compress slightly when it's properly installed on the top of the cooler. This will create a very effective seal.

Finally, install your circulation fan. We used a duct fan from a hardware store. It fit nicely inside of our cooler. Somehow, the specifications on the fan do not list the lowest temp it will operate at. There is some risk that the oil used on the fan will freeze; keeping the fan running should allow the heat from friction to keep the bearings at a happen temp. In practice, this has shown to be true.

When you are done, you should have something which looks like the below photos..

Step 4: Final Assembly

  • Drill a 1/2" hole in the foam lid.
  • install coil and reservoir pipe on the lid
  • Insulate reservoir pipe

When you are done, you should have something which looks like the below photo.

Step 5: Using the test chamber

Make sure you are well stocked on liquid nitrogen and dry ice. SAFETY is very important here. You can really hurt yourself with these items; they WILL freeze you if you come in contact with them. Use strong insulative gloves, safety glasses, closed sealed boots, pants & long sleeved shirts. Also, getting moisture inside of the coil could cause it to clog and burst with very hazardous results. Don't do that. Ensure your coil is dry and empty of all moisture and fluids (best method is to make sure water doesn't get in there in the first place..) In addition, ventilation is very important. nitrogen can displace o2 in the air, which means you've got yourself a vapor version of iocaine powder. Odorless, tasteless, deadly. Make sure you've got lots of room and a source of fresh air.

Liquid nitrogen can splash, which sends -321F liquid all over the place. This is bad, don't let it happen. It will boil at room temperature. Seek the guidance of a professional on the safe handling of dry ice and liquid nitrogen.

Start off by staging your work area. Have cleanup devices handy, and plenty of room to work. Get out your supplies and objects you wish to freeze.

Load the cooler with the dry ice and reusable ice packs. Don't use normal ice; this will create moisture in your test chamber, which is a pain. It will freeze and condense on your coil, which reduces the effectiveness.

The reusable ice packs increase the thermal capacitance of the system. Pre-freezing them in your freezer is a good idea.

Place your thermal probes inside of the cooler, at positions you wish to measure the temperature. On your first run, you will likely discover that your thermal probe won't measure as low as required. Some testing will be needed.

The dry ice should be placed between the coil and fan, staged such that the fan blows air over them. Once the lid is sealed and the fan is turned on, you should start to see a rapid drop in temperature. This will not cool past -30c to -42c due to heat loss. Once the chamber has hit a stable temperature, it's time for the liquid nitrogen.

This is the hazardous part. Use due care.

The liquid nitrogen when it enters the reservoir will phase change almost instantly. This creates nitrogen vapor, which is of a much larger volume. So it will tend to spray and burp until it cools the reservoir down to a point where it is not boiling so rapidly. If you just 'dump' much liquid nitrogen into this reservoir, it will likely splash back into your face blinding you. Don't do that.

One trick to help with the out gassing is to vent the funnel. This can be done with a simple tube run down inside of the funnel. This allows the gaseous nitrogen to escape without it blasting up liquid nitrogen into your face. In addition, ensure that the other end of the coil is not blocked or clogged, or it could burst/explode.

Slowly pour the liquid nitrogen into the funnel. At first, the temp of the system will not change, as the liquid nitrogen needs a chance to pre-chill the external reservoir and coil. Closely monitor the temp to ensure that you don't get below your target temp. Do not overfill the reservoir.

As shown on the multimeters below, you can reach very cold temperatures very easily.

That is it! you've now built yourself a el-cheapo thermal test chamber!
<p>Gents, is there anyway to induct the LN2 into the heat exchanger with out having to pour it from the Dewar flask? If you placed a copper tube to the bottom of the flask with a pressure tight fitting at the top of the flask and a valve, shouldn't the off gassing at the top of the flask pressurize the fluid LN2 and cause it to rise up the tube to the heat exchanger? Just a thought.</p>
me likes. yeah ok, so many noobs out there that are 'leet haxor overclockers' brag about having supah dupah cooling systems at like, -30C n crap. silicon junctions work optimally around 40C .so cpu says DO NOT WANT. your electronits will get rather unhappy at cold temps especially any caps that happen to get that cold, their mystical powers of filtering disappear! so cooling to maybe, 20C would be ideal, to allow for core temps being higher then ambient and also not cold enough to cause condensation.
Really? I thought the electrons flowed better at cold temperatures, meaning better conductivity, meaning better performance from any and all devices. No?
Just an interesting detail about N2 (or any inert gas I suppose) is that if it displaces most the air in the room you'll suffocate without warning and you'll die even if brought to fresh air unless the N2 is forced from your lungs. It's the CO2 that triggers your breathing reflex; you won't start breathing without it. A matainance guy died just like this at a local IBM facility when he was working in the sub-floor and there was an N2 leak. Sorry to start the urban myth, but I don't have any more details...
ok, yes you can suffocate if you don't have ample o2. However, it is not true that it needs to be forced from your lungs. That is an urban myth; you can purge even the heaviest of gases by just breathing deeply. No need to turn upside down, or any of that crap. If someone is suffocating, standard CPR rules apply for resurrection. IANAL YMMV good luck. :)
yup. the condition is called hypoxia, it happens when you do not get enough 02, and too much n2. you will not be able to tell when you get hypoxia until its too late.
actually, having been in a hypobaric chamber (for the reason of identifying my own personal symptoms of hypoxia, since they vary from person to person), i can say that the most common symptoms are blue fingernails, dizzyness/loopiness, and losing the ability to write. after about 3 of the 5 mins at a simulated 20k ft, i definitely knew that something was up biologically. therefore, the person educated on hypoxia can know when it's time to use an emergency breathing aid.
my knowledge of hypoxia is limited to what I learned form my Nitrox diving course, it has to do with hypoxia underwater, which is diffrent
ah, right. hyperbaric hypoxia. i do not have diving experience, so that scenario had eluded me when i commented. (oO)
hyperbaric hypoxia? its just hypoxia.
yeah, it is just hypoxia. the five hour class that i took on the physiology of humans in a microgravity environment had a teacher who described underwater hypoxia with the term i used. <br/><br/><a rel="nofollow" href="http://www.google.com/search?hl=en&amp;client=firefox-a&amp;rls=org.mozilla%3Aen-US%3Aofficial&amp;hs=l0V&amp;q=%22hyperbaric+hypoxia%22&amp;btnG=Search">http://www.google.com/search?hl=en&amp;client=firefox-a&amp;rls=org.mozilla%3Aen-US%3Aofficial&amp;hs=l0V&amp;q=%22hyperbaric+hypoxia%22&amp;btnG=Search</a><br/><br/>you may call it whatever you like, silly. 8)<br/>
i wonder if this thing is cold enough to cool my 360...
dude, if you try doing that you're gonna screw up your 360
FUN THINGS: I knew the issue of "suffocation" with "simple aspyxiants" - in low oxygen atmospheres - was a "REAL" issue, I just never understood the mechanisim. It's simple. Person climbs into tank that had nitrogen, or argon as it's covering gas (inert) over say something like food oil, or wine - and the tank is say 20' deep and the layer of gas may be 10" deep or what ever. The person gets to the bottom, feels dizzy, collapses, and then may die fairly quickly.... As all too frequently happens, some one sees the collapsed person, in the bottom of the tank, ASSUMES they must have fallen or had a stroke or something, and then climbs into rescue them.... ' They too collapse. The third person comes along...... same thing happens...... Three people died like this at either a winery in a HUGE wine tank, or it was an empty holding tank for pig manure -- filled with methane. The way (usually) "INERT" and "NON TOXIC" gases, that are almost always COLORLESS and ODOURLESS kill people is really really neat..... it's just so subtle and frequently fast. The lungs and brain are hard wired to EXHALE when the build up of Carbon Di-Oxide gets to a certain point in the lungs...... It's the CO2 coming out of your blood that makes you breath....... If your not taking any or too little oxygen into your blood, then there is no or not enough CO2 coming back into your lungs to register the breathing mechanisim. It's not the pressing need to exhale or inhale that is so much the problem, its just we have NO mechanisim, to detect and respond to any gas, other than CO2...... That's why people who climb into tanks filled with nitrogen, methane, helium or argon etc.... they never panic, they don't feel alarmed, they just feel woozy as the last dregs of oxygen in their blood gets used up and "poof" out like a light. They collapse and occassionally die - very quick and very easy.
(citation needed)
I noticed the person pouring the liquid nitrogen wasn't wearing gloves. Was that safety precaution just for the person holding the funnel where one might get splashed?
As -200 degrees celcius might sound like something that's very dangerous it really isn't. Accidental splashes are harmless and well, we have even tried pouring some of it just on the hand. Nothing happens as it is quickly evaporating when it gets into contact with skin.
Other than the fact that you're risking 4th degree burns. you can do what you will, but I don't think it's a good idea to encourage others to neglect basic safety precautions.
Better to use no gloves than ones that are not fit for the purpose. LN2 will vaporize easily in contact with skin but if a porous surface is presented it can easily absorb into the material which subsequently acts as a more efficient heat exchanger. That said, handling dewars is no problem but filling them is a different story and appropriate measures should be taken. The greatest risks of using LN2 are burns not from the liquid itself but from objects that have been super-cooled. Also wise to avoid confined spaces as 1L LN2 ~ 1 cubic meter of displaced air.
I'll agree to that. but, a pair of appropriate rubber coated cotton elbow length gloves are a drop in the bucket compared to the cost of a dewar. I was merely pointing out that it is better that the audience have a fair amount of reverence for the materials they might be working with.
100,000 ft via balloon. There is a group that routinely launches many small projects to that altitude at no charge.<br/><br/>Check out PONGSAT at <a rel="nofollow" href="http://www.jpaerospace.com/pongsat/index.htm">http://www.jpaerospace.com/pongsat/index.htm</a><br/><br/>Myc<br/>
Very Nice, but the question must be asked, how did you get liquid nitorgen? and why would you need to test electronics for going to space? I like your instructable.
Thanks! Check out page one for more info on where to source LN2. As for testing - - it gets really, really cold up there. resisters have different values. CPUs crash. Wires become brittle. oil-based Lube freezes. Many sensors behave very differently when that cold; CCDs in fact have much less noise on them, but many other sensors (gyros, accels) just start acting wonky. Also, batteries have much less life when they are cold.
I think he's asking - "are your electronics actually going into space?"
Into the 'edge of space', yes. we are targeting about 100k feet, or about 20 miles up. <br/><br/>Here's a handy video from last year: <br/><br/><a rel="nofollow" href="http://blog.makezine.com/archive/2007/04/ahab_story_pdf_cast.html">http://blog.makezine.com/archive/2007/04/ahab_story_pdf_cast.html</a><br/>
Oh, a <em>weather balloon</em> - I was hoping your were involved in some university rocket-shot.<br/>
Weather balloons are the unsung heroes of space exploration. They do what a lot of NEO satellites can, at a pittance. They can stay aloft for weeks, can be brought down on command (no needles needed) and the instruments can be reused again.
bah, uni's can go suckit. We are a group of indy folks doing it for fun. :)
oooh... hippy rocketry? srry, conservative roots equate indie to hippy...
??? Respectfully your roots must be of a hybrid stock. Conservatives have long championed private, independant initiative. Or is that only the initiatives they support? :(
Sorry, by "conservative" I meant "stick-in-the-bum"... You know the type that still believes that anything that goes against the norm is wrong, piercings and tattoos automatically mean you listen to heavy metal, have sex before you're married and you will get aids... conservative. I guess I should've said traditional or something. Getting off-topic, srry...
plz don't confuse "non corporate driven enthusiasts" with fixie tot'n latte sipping ipod warezing hipsters. :) I hope I'm not violating the be nice policy; I'm really just playing around.
LOL You don't think the university folks aren't doing it for fun? Just don't tell anyone the truth :) Many independent groups partner up with university groups. As long there is an educational value the science depart can and do supply things perhaps even the test fixture featured here freeing up the in dependants funds for other things.
Nice project. You made it to hackaday, few instructables do, congradulations What do you mean when you say &quot;press&quot; fitting for the coper pipe, is press short for compression? In the event you are speaking of the common ferrule commpression fitting used with copper tubing, I never found them diffucult to work with. I found them to be very mechanically secure, but I wouldn't use them to support weight. I have known them to with stand up to 1,000 PSI without failure.<br/><br/>Why such a large fan, is there a need for speed and volume when circulating the lack of heat inside a dead air space?<br/><br/>I have been monitoring the telemetry of payloads launch other groups for over 20 years now, is your group reviewing their experiences? I'm physically locate between the launches of the EOSS <a rel="nofollow" href="http://www.eoss.org/">http://www.eoss.org/</a> and of HABITAT <a rel="nofollow" href="http://habitat.netlab.org/index.shtml">http://habitat.netlab.org/index.shtml</a><br/>Good luck on your future flights.<br/>
yay for hackaday
yes, I was talking about compression fittings. The 'problem' with this very simple design is that the coil is supported by the compression fittings. It's not all that heavy, but the fittings, as you pointed out, are not designed to support weight. I have tried much smaller fans, and found that it takes MUCH longer to lower the temp inside of the case. Using a 3" CPU case fan it took over 2 hours to reach -40c. You need to move a large volume of air over the heat exchange & dry ice to col the air temp quickly. In addition, it provides a much closer to 'real world' testing up at 100k, as it's rather windy. The next chamber we build will also likely be a vacuum/thermal chamber. I've been involved in other ham launches and private edge of space launches for years. There is lots of good (and lots of crap :) information on the net for doing this sort of project.
Knowing where this was made, ventilation really isint a problem in a warehouse space. I doubt someone would be building an edge-of-space payload, much less testing it in a small room.
sssh, it's a "secret non disclosed location". :P
Ah yes. Everyones favorite butcher shop. With pantless people walking around.
I hope you're testing the materials used in the glider at the same temperatures. Getting the plastic components well below their glass transition point (Tg) may embrittle them, as will many metal components. It'd be a shame to lose the project because a control horn or mounting bolt snapped or the wing skin fractured.
If you want to test the electronics at those temperatures, why not just dunk the electronics in LN2? The issue is thermal capacity. The upper atmosphere will be almost infinity cold, so any object up there exposed should be tested 100% at those temperatures. If you are measuring -55 away from the cold tube, then you can assure that the parts are not being truly tested at -70. If you are testing insulation, etc, then the same test holds. If the object is mechanical, and can not be submerged, then try putting just above a pool of LN2. Enclose the chamber and blow the fan. Also, those plastic coolers are fun, but they tend to become too brittle with cryogens. They also conduct too much heat, as they are designed for 0C, not -70C. I would recommend those cheap Styrofoam coolers from the dime store. They are cheap, work down to 4K, and are easily cut to make forms.
We are testing insulation, battery run times, and full packaged payloads. From 0 to 100k, in the seattle area where we are launching - - it doesn't get any colder than -70c.
That vent fan you're using is probably adding quite a few watts of heat to your chamber that you could live without - my thought is have the motor OUTSIDE, with a drive-shaft to a fan of sorts passing THROUGH the foam layer - you get the agitation you need without adding the heat...
agreed in theory... that fan draws 50 watts and cost $35 bucks new. The cost/labor of an external fan exceeds the cost of just adding more LN2. Now, if we ran this thing all the time, then yes, every little watt counts.. and it would make more sense to use a better chamber as well. But this is on a budget, and will not see that much use...but by all means, if you can find a cheap externally powered fan for this use, it will be more effective. Thanks for your comment!
3rikj - I think the problem comes when you lose conscious and don't receive CPR. I think the IBM guy died because he was brought to fresh air unconscious and people though he would begin to breathe on his own.
umm i have 1 question. where the heck do you get liquid nitrogen and how much does it cost ?? lol sorry if it was listed and i missed it =/ . soo how much? lol<br/>
Costs vary. Less than petrol at your local petrol mart. Purchased from local welding supply shops. Deposit might be required on the dewar. Local schools also have access to LN2.
Wowzerz. Nicely done Instructable. Images are great, details are nice, everything is awesome. Nice job!
Thanks! We end up working on all kinds of random projects down at our hackerspace, and it's about time we publish some of the stuff we've learned/developed online. Watch this space for more projects.
Nice Instructable. I like the idea of using this site to document projects done at school. One extra safety point to be aware of - ventilation. Nitrigen is not actually toxic at normal pressure and temperature, but a leak can displace the breathable air in the room. Unfortunately, the only sign of this happening is that you feel rather euphoric (in a very good mood) and then die. Fortunately, it is very easy to avoid this fate - leave a door or a couple of windows open or switch the air-con to pull air in from outside.

About This Instructable


38 favorites


More by 3ricj: Thermal test chamber for edge of space testing.
Add instructable to: