Picture of Make a Cloud Chamber using Peltier Coolers

Cloud chambers let you see vapor trails left by radioactive particles!

A version of this cloud chamber is now available for purchase (fully assembled and tested) from http://www.nothinglabs.com

Discuss cloud chamber assembly and radioactive samples at http://forum.nothinglabs.com

Cloud chambers work by creating a supersaturation of alcohol vapor.  Vapor trails are then formed when the radiation ionizes the alcohol.

Cloud chambers need to be very cold (about -15f) - so most designs require dry ice.

This design uses inexpensive electrical "Peltier" coolers instead!  No need to buy / store dry ice!

It's fairly easy to construct - and should only cost about $100.

This chamber design cools down very quickly - you can see vapor trails within a few minutes.

On the downside - it's not as cold as a dry ice chamber - so the supersaturated area isn't as thick.

Step 1: Stuff You'll Need (Part 1)

Picture of Stuff You'll Need (Part 1)
Peltier Coolers (need 2)
- Model 12709 (138 watts "max") for "top" cooler ($8 on ebay)
- Model 12710 (168 watts "max") for "bottom" cooler (also $8 on ebay)
- Search eBay -descriptions- to find specific models
- Often listed as "TEC", "Thermoelectric" or "Peltier" coolers and have TEC1 or CP1 as a prefix to the model number
- Other models with similar wattage ratings can likely be used (bottom cooler should have a slightly higher wattage rating than the bottom)
- Coolers look identical - put a piece of tape on a wire of the 12710 so you can identify it later
- Larger coolers are not necessarily better
- Fragile wires tend to break - avoid bending (might want to buy a spare or two)
- Do -NOT- connect a peltier to power backwards or without a heatsink!
- See the bottom of this page for info on a single-Peltier option

Power Supply #1
- 12v and at least 8 amps (good results between about 11v and 15v)
- Used to power 'bottom' peltier cooler
- www.allelectronics.com part # PS-1512 is a good option for $25

Power Supply #2
- 5 volts and at least 2.5 amps (good results between about 4v and 8v)
- Used to power 'top' peltier cooler
- D-Link JTA0302C power supply on ebay for about $10
- Alternative: Use a PC ATX power supply for both 12v and 5v sources (Google for more info)

CPU Cooler / Fan
- Look for "heat pipes" and a 120mm fan
- Design that can be run "upside down" with good airflow
- The Cooler Master Hyper 212 works well and is cheap (under $30 on ebay)
- Cooler Master V8 works well and looks cool (about $50)
- The Corsair H50 (liquid cooling) works a little better - but is more expensive / complicated (about $70)
- See http://www.frostytech.com/ for other possible coolers

Alternative Single Peltier / Single Power Supply Configuration
- This project can also be built using a single dual-stage peltier cooler
- The 19012-5L31-06CQQ from Custom Thermoelectric ($53) has been tested and works well (order with Moisture Sealing option)
- Requires a 12V power supply capable of 6 amps (5V power supply is not needed)
- Other project details stay the same

1-40 of 132Next »
AdnanN11 month ago

Why two separate power supply for peltier and fan while both can run in 12 V ??

BillM75 months ago


I've read the various comments - the solutions are simple

Develop a kit - under 200.00 - add some free tech assistance - I'll buy two

One for home - one for office - makes a great conversational piece.

Have a good presentation - easy to assembly kit - go to work

I'm waiting to buy one now -

ThomasJ18 months ago

Hi Rich,

I finished building mine and I get a good supersaturated layer over a 5x5" black anodized aluminium plate. I suspect the chamber is working well since I get really good "fake trails" whenever I place/remove/recharge my static source (balloon). however, I don't seem to get trails coming from cosmic rays. Is there a trick with the static source regarding cosmic rays?

nothinglabs (author)  ThomasJ18 months ago
Hi Thomas -

It's most likely you aren't actually getting cold enough to see traces.

From my experience - the kind of "ghost trails" that occur when moving the static source around can occur at higher temperatures than needed to see traces from particles.

The peltier setup for this project has fairly modest cooling power - and can really only cool a small area down to the -15f or so needed to see traces.

Using thin plastic for a surface causes a small area to be cooled directly above the peltier (due to low thermal conductivity). Using aluminum will attempt to cool a larger area - and will counter-inuitively produce less cooling.

I had some luck extending the saturation area with an aluminum base-plate - but only using much larger peltiers - and insulating the bottom of the plate.

My general advice to anyone trying this project is to start with a thin plastic base-plate - get that working - and then experiment from there.

Good luck!


Rich, you were absolutely right !

I removed the aluminium plate and used a thin transparent plastic sheet blackened from under with a sharpie. After a few minutes of cooling, I reached between -45F to -50F (compared to -10F previously). I can now see about 5 to 10 trails per seconds above the peltier area.

Thanks for the advice !


nothinglabs (author)  ThomasJ18 months ago

glad it's working for you. now for some real fun - go get some uranium ore on ebay!


PopsicleGhoul8 months ago

Wow! I love your detailed instructions! I also like how you used easy-to-find Peltier coolers instead of difficult-to-acquire dry ice. Hopefully I can try this out soon!

kamhagh10 months ago

mine gets cold on the upper side, but how can i know that the connection to heatsink is ok? i used kinda less grease ! im worreid that overtime it will get hot ! or my thermal connection is not good :|

nothinglabs (author)  kamhagh10 months ago
If things work (ie - you can see traces) then you're getting good contact.

Other way to test is to use an infrared or contact thermometer that can go down to -20f or colder.

In general - less grease is better assuming you get full coverage.

Good luck!


I have a dumb question about powering the CPU fan. I bought the Cooler Master Hyper 212 for my chamber, which has a 4pin output. I tried looking up which pins to connect to my 12V 8A power supply, but I can't seem to get a straight answer online. Did anyone else use this fan that can help me out?

nothinglabs (author)  cmerrill3141 year ago

It's been a while - but I believe this is correct:


short answer - black = gnd, yellow = 12v

julio792 years ago
So I've constructed a cloud chamber using this article and have run into a slight problem. Everything works but I don't think the chamber is getting cold enough. I haven't had a chance to buy a thermometer yet but I'm fairly certain it will only confirm what I think I know. The alcohol is condensing on top of the cooler forming a thin square layer to form on the base, but no mist at all. The voltages are 12 for the bottom and 5.6 for the top cooler, so are there any other things I might need to adjust? I have already removed and reapplied some arctic silver but no dice. There is a good amount of frost buildup underneath the container by the coolers so maybe the container is too thick? I bought one similar to the one used here however. Any troubleshooting advice would be greatly appreciated!
nothinglabs (author)  julio792 years ago
Sorry for taking so long to reply.

What kind of light are you using? (LED light is pretty much mandatory to see the traces)

Also - have you painted / used a magic marker to make the surface black? That's also critical to see anything.

The exact container I used seemed almost custom-made for cloud chamber user with it's thin bottom. It's definitely possible the one your trying is too thick at the base.

One option is to use some thin black plastic (1/16th" or so is best) for the chamber base - and then use a water glass for the chamber itself (don't worry too much if the seal isn't totally perfect). This is the approach I usually use on chambers now.

A local plastics retailer might have something that would work. Can also try something like a plastic picnic plate. Either way - paint or magic-marker it black.

The alcohol pool itself will inhibit traces - so if it forms - wipe it off with a paper towel.

Hope this helps!

I am going the route using the single dual stage peltier cooler you listed. I bought a 12V 6A 72W power supply for it. Just a stan dard AC DC transformer. I am worried about the Power rating on it. Should it be enough for the peltier. It will be used to power the TEC only. I am experienced with electronics but I have not worked with TECs much at all.Thanks for the help in advance!
tacoredneck2 years ago
what kind of plastic do you use on your commercial chamber? id like to know the thickness too. the plastics i have used are about an 1/8th of an inch thick but i think it has insulating properties. if you don't mind me asking,i know your trying to sell them.
nothinglabs (author)  tacoredneck2 years ago
i don't really keep a lot of secrets any more - happy to share.

I use 1/16" delrin plastic (check mcmaster.com) in the commercial cloud chamber. delrin is a bit flexy though - so I support it with some thicker plastic around the edge.

my suggestion for others would be to use 1/16th" garolite. mcmaster.com had this in nice 6"x6" pieces in black last i checked.

I like garolite because it's stiffer than most plastic. iif the material is too flexible - the temperature differential will cause major flex that can be a problem.

for any kind of plastic I recommend getting black if possible to avoid having to paint it.

I wouldnt worry too much about how insulating the plastic is.
tacoredneck2 years ago
i have a problem. im not getting a fine mist but the alcohol is pooling over the peltier. i have abandoned my aluminum plate for now and am using a Tupperware container
nothinglabs (author)  tacoredneck2 years ago
do you have a temperature reading?

Is the surface black? It's pretty much impossible to see the traces / mist against a light background.

What kind of light are you using? If you are using something other than an LED - probably won't work.

Good luck,

tacoredneck2 years ago
what about something like this... The peltier would fit in the square hole. then the aluminum plate would mount to the cpu cooler. insulation would fit between the peltier and the plate. a square plate would fit over the pelteir to protect it from the alchohol
nothinglabs (author)  tacoredneck2 years ago
Maybe? Couldn 't really say for sure.

My suspicion is that you'd only get super-saturation (trace visibility) over the actual peltier.

Might want to try getting everything working with a plastic sheet first - then experiment - but good luck with whatever path you try!

tacoredneck2 years ago
how do you feel about metal base plates for the chamber instead of plastic?
nothinglabs (author)  tacoredneck2 years ago
Certainly can be done - but don't recommend it.

I get about an email a month from people trying to use metal - and not having it work.

Metal tends to conduct heat -too well- - so you end up with a larger cooled area - but doesn't end up being as cold.

IE - 8 sq inches at 0f (no worky) as opposed to 3 sq inches at -30f.

myakka2 years ago
OK. Sorry to bother again. I have an ATX power supply rated at 650 watts, a cooler master V8, one 12709, one 12710 and its all hooked up. I get a reading around -7 through -10 degrees Fahrenheit on the bare top peltier. I did not use arctic silver, I used the thermal paste that came with the cooler master. Do you think i would get much better results with arctic silver? I am also using a laser thermometer same as yours. I have read the trouble shooting section but i feel that I need your opinion...
nothinglabs (author)  myakka2 years ago
suspect the included thermal grease is fine. Generally the stuff included with coolers is reasonable quality.

are you powering the 12709 (top) at 5v and the 12710 (bottom) at 12v?

sometimes the IR thermometers can be a bit flaky (try holding it closer to the surface to see if that gets you a colder reading).

If you think everything else is right - might be worth just trying it out.

thank you. i tried it again this morning and it got to around -20. maybe, like you said, the paste might have to set in.
myakka2 years ago
never mind. i just found a brand new cooler master v8 on eBay for $20.
05leftlj2 years ago
What is the reason the second peltier cooler is at 5V not 12V?
nothinglabs (author)  05leftlj2 years ago
see end of step 18
myakka2 years ago
I bought this from ebay, do you think it will work? http://www.ebay.com/itm/190712915267?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649
nothinglabs (author)  myakka2 years ago
Can't say for sure if that'll work - but it seems reasonable.

Good luck with your fully-automated build! I toyed with that idea - but never got around to it. Seems very do-able though.

myakka2 years ago
I plan on building one of these to be fully automatic. meaning of course that a pump will supply the alcohol to the felt and a hv source to clear out static. i want to just leave it on my desk at work an watch the look on peoples faces as they walk by.
mamzypig2 years ago
What is a 12v 8 amp power supply? I googled it and it came up with a " Ac " or "Dc adapter"
Is that the right thing.
nothinglabs (author)  mamzypig2 years ago
It means a power supply that can produce 12 volts (DC) at 8 amps of current. (from AC wall power).

8 amps is a minimum - more is fine.

This would be an example:

jae10283 years ago
I'm 14years old student in Korea and Im working on building a peltier cloud chamber. Our science teacher asked me to build one for a competition. I ordered tec 12012 and 12010 since 12009 is not available in our country. I also ordered an atx psu. Can I ask what voltage I should put on each module?
Also I was wondering if the heatsink in the link below and 120mm fan can withstand heat. Otherwise I have to buy a cpu cooler.
Thanks for help.^^
nothinglabs (author)  jae10283 years ago
I suspect the 12012 and 12010 will work for you. I would run them at 12v for the bottom (12012) - and around 5v for the top (12010).

Since the 12012 is higher rated than the 12010 typically used for the bottom - you might try a little lower votlage - maybe 10v.

I can't say for sure if that heatsink will work or not. I typically have had best luck higher performance heatsinks including "heatpipes"

If you already have the heatsink - there shouldn't be any harm in trying though.

Good luck!

toblorone3 years ago
i have an excess of 300 v flash charging circuits.(they actually gave them to me for about a month. then the larger stores refused to give me any.) will these be a suitable substitute for the static?
nothinglabs (author)  toblorone3 years ago
I don't think the static is a more-is-better type thing. I've found too much static can disturb the supersaturation layer - and prevent tracing from showing up.

A constant low-level source of static like a balloon has worked really well for me (using a maybe 1.5k-3k DC source is another option).
i would just use a balloon, but i am building one as part of my applied science class, so i aim to reduce as many variables as possible. can you recommend anywhere that would sell a dc source like that?
nothinglabs (author)  toblorone3 years ago
I've used an AC power supply like this one:


But you of course need to rectify it (you can do this using a single diode)

Please be careful with high voltage!

Even though I have the HV supply - I still generally use the balloon as it seems to give a bit better results (sure there are variables I haven't pinned down).
OK, thanks. just one more thing: how long would a design similar to the one here take to build?
1-40 of 132Next »