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

Discuss cloud chamber assembly and radioactive samples at

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.

Remove these adsRemove these ads by Signing Up

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
- 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 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

ThomasJ18 days 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)  ThomasJ16 days 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)  ThomasJ13 days ago

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


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!

kamhagh2 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)  kamhagh2 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!

cmerrill3147 months ago

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)  cmerrill3147 months ago

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

short answer - black = gnd, yellow = 12v

julio791 year 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)  julio791 year 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!
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)  tacoredneck1 year ago
i don't really keep a lot of secrets any more - happy to share.

I use 1/16" delrin plastic (check 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. 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.
(removed by author or community request)
They're typically rated in watts - but that doesn't really tell you how much heat they'll pump (because the wattage number includes heat generated by the peltier).

For more info I'd look at

Thanks, that's a lot more than I knew. I'll start re-reading all that for all future reference.
Of course, I mean BTU/Hr rating!
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)  tacoredneck1 year 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,

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)  tacoredneck1 year 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!

how do you feel about metal base plates for the chamber instead of plastic?
nothinglabs (author)  tacoredneck1 year 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.

myakka1 year 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)  myakka1 year 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.
myakka1 year ago
never mind. i just found a brand new cooler master v8 on eBay for $20.
05leftlj1 year ago
What is the reason the second peltier cooler is at 5V not 12V?
nothinglabs (author)  05leftlj1 year ago
see end of step 18
myakka1 year ago
I bought this from ebay, do you think it will work?
nothinglabs (author)  myakka1 year 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.

myakka1 year 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.
mamzypig1 year 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)  mamzypig1 year 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:
jae10282 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)  jae10282 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!

toblorone2 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)  toblorone2 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)  toblorone2 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?
nothinglabs (author)  toblorone2 years ago
i'd figure on a half-day to day to build the chamber.
Why do you need the static electricity? I have never understood these things.

nothinglabs (author)  spark master2 years ago
The static clears out ions and dust - which would provide other nucleation points besides the ionized trails left by the particles (preventing you from seeing the traces).

It may also help "push down" the ionized trails into the super-saturated area - so that more are visible.
Ahh, you might consider an ion fountain destatic tool made from a piezo electrostatic discharger. Back when vinyl records were still king and cd's were just emerging. I had a gizmo that had a piezo xtal and a squeezer that caused th eXtal to discharge through a pin in the front of the barrel. You held it over the cleaned record and pulled the trigger. It let out a plume of ions that counter acted any on the record. Hence less or no static pops or hiss.I used to use mione to give myself electropunture shots , before it was accepted, (electrical accupuncture). If I can find the name or the unit and pictures I will bring it back herre if you think it might be useful
it was called Zero Stat and was red, now they're blue
Radpro3 years ago
I apoligize for the continued questions, I have not recieved my parts yet so it may be easier to understand when I get them. Does the cord, the one similar to a computer cord, all you need? How does this power both units with only one output. Im assuming the peltier and the coolermaster hook together and the cord plugs into the coolermaster. I also see a box in the regulare instructable(the picture between the cord and the peltier) Is this part eliminated due to the single stage peltier. I am sure I could figure all this out when I receive the parts, but I need to make sure I have all the parts ordered. Thanks again
nothinglabs (author)  Radpro3 years ago
you should be able to power both the fan and peltier cooler off the single 12v / 6amp power supply (the one that looks like it's for a laptop).

you will need to cut the end of the cable off the power supply's cable / solder it up to both the fan's power inputs and the peltier's. both the heatsink's fan and the peltier take 12v.

I can't tell you off the top of my head if the wires on the power supply are color-coded for polarity. You might want to use a multimeter to verify which is positive and which is negative.

I think the "box" your talking about is a big 12v power supply (the "cord" pictured is a 5v power supply). With the setup you'll be using - you'll just need the smaller 12v power supply I mentioned in the prior post.

hope this helps!

Radpro3 years ago
I ordered the single stage peltier and was wondering what type of power supply you used, one that powers both the fan and the single cooler and also looks clean. Where can I get one? Thanks again for your help
nothinglabs (author)  Radpro3 years ago
I use one of these (there are similar models that work as well):

just look for anything 12v / 6amp.

If you want to add a power switch - one of these will plug right into the power supply:

(you can find these cheaper on ebay - but they ship from china / take a long time to show up - also be aware there's s similar / but different power switch - avoid it - as it's junky)
Radpro3 years ago
Hi, I am a current student pursuing a degree in radiation protection. I have been very facinated with the cloud chamber and have decided to build one for my project at school. I love the one that you have for sale but cannot afford it at the current time. The project that I am building it for is actually for my autocad class, not my radiation protection class. Today I went to speak with one of my potential employers, as they were showing me around the lab I ask them if they had ever seen a cloud chamber. I was stunned that they had not heard of one(being that they study and test sources for a liveing) and they asked if I would bring it down to display when I get done. I was wondering if I could get your advice and maybe a materials for constructing one like the one for sale since I will be displaying it in a proffesional setting. If you would be so kind to help me out with this, and as a thank you I could send you a file of a complete blueprint of the construction and assembly to include with your instruction manuel, and display nothing labs-with your name or names of the people who developed this design when I display the chamber. Thank you so much for your time, and I understand and will hold no hard feelings if you feel uncomfortable sharing your design. I will in no way attemp to sell this or any thing similar. My email is Thanks again, and thanks to all who have sharred their time and information on this site.
nothinglabs (author)  Radpro3 years ago
hi radpro -

happy to share some more info on the commercial chamber (there have been two versions - but both work about the same). glad to offer some general guidance - getting too specific probably won't be that helpfull - there is a large amount of "nuance" in how things go together.

thanks for the offer on the blueprint - would be curious to see what you come up with (I probably wouldn't include it with the manual though)

first - it uses a special (little more expensive) dual- stage peltier - this eliminates the 5v power supply (and lets you use a cheaper 6amp for the 12v) - and simplifies wiring.

for a link - look at the 2nd page of the instructable under "Alternative Single Peltier / Single Power Supply Configuration"

the cooler is a Cooler Master V8. it was chosen mostly for looks - but does work well.

the "base plate" of the current commercial chamber is made of delrin (aka acytel) - which is held in place using some fairly complex mounts I engineered.

the earlier version of the chamber uses a "garolite" fiberglass base (i don't remember if it's g-9 or g-10 - its whichever comes in black). this version was held in place using arctic silver thermal epoxy (it also had some supports for strength).

(delrin doesn't epoxy well)

I would not epoxy anything in place until you've tested the chamber using regular thermal paste.

both materials are black and 1/16th thick and available from

the chamber's glass is a high quality food storage container (have only found these locally - but I'm sure theres something simlar near you). the top of the chamber has a furniture glide silicon-sealed to it to absorb the alcohol.

For more hints on how the commercial chamber works - go to - then the page for the commercial chamber - and download the manual.

hope this helps! feel free to ask any other questions.


I really appreciate your input and help. I am ordering the parts this weekend and will let you know how it goes. As for the blueprints I will b required to show many views, Isometric included and it will be created in autocad. There will be realistic views, assembly views, and schematic view. If there is anything in particular you would like for me to add I would be glad to do so, any view or any assembly methods or dissassembly.Once I get the parts configured manipulating them to show different things is easy. It will be complete around nov.5 and ill shoot you a copy. Thanks again for you help and wish me luck!
samol4 years ago
Hi nothinglabs,
Are the cooling heatpipes efficient even though they are mounted upside-down ?
I mean, there's still the capillarity effect, but the gravity is in the wrong way.

I'm planning to build a cloud chamber but I'm not yet sure if I use your type of cooler or if I need to bend the heat pipe (a single sided headpipe version) in order to make them horizontal.

Thanks for the great tutorial !
Samuel from France
nothinglabs (author)  samol4 years ago
the heatpipes seem to be working fine - if they weren't I would be getting temperatures below -30f!

not sure if performance would be better with things "rightside up"

I also have a more power Corsair liquid-cooling system - which should run fine upside down - which only gets a few degrees colder than the heatpipes do (also indicating the heatpipes are working pretty well).

I highly recommend using the Kingwin HTC XT-1264 for this project. I've tested a lot of coolers - and it seems to provide the best bang for the buck.

The Cooler Master V8 is a little better - but more like $60.

Good luck - feel free to ask questions.

Thanks for the help,
I'll use the Cooler Master hyper 212+ since the kingwin is not easily available in France.
My conctruction (almost done, only need the 4 peltier) is inspired by this cloud chamber :
I'll make a little overview when it's working.

nothinglabs (author)  samol4 years ago
that cooler should work fine - although you may need 2 of them if you're planning on doing the 2x2 peltier setup.

I've been in contact with the guy who made that chamber - he's done some very nice work! (seems very committed to doing research)

here's my latest video - featuring a Lead 210 alpha source:

good luck!
Hi, I am a current student pursuing a degree in radiation protection. I have been very facinated with the cloud chamber and have decided to build one for my project at school. I love the one that you have for sale but cannot afford it at the current time. The project that I am building it for is actually for my autocad class, not my radiation protection class. Today I went to speak with one of my potential employers, as they were showing me around the lab I ask them if they had ever seen a cloud chamber. I was stunned that they had not heard of one(being that they study and test sources for a liveing) and they asked if I would bring it down to display when I get done. I was wondering if I could get your advice and maybe a materials for constructing one like the one for sale since I will be displaying it in a proffesional setting. If you would be so kind to help me out with this, and as a thank you I could send you a file of a complete blueprint of the construction and assembly to include with your instruction manuel, and display nothing labs-with your name or names of the people who developed this design when I display the chamber. Thank you so much for your time, and I understand and will hold no hard feelings if you feel uncomfortable sharing your design. I will in no way attemp to sell this or any thing similar. My email is Thanks again, and thanks to all who have sharred their time and information on this site.

Your actual setup is very neat and gives excellent results ! I almost completed my chamber :
Let's hope it's succesfull !

Edit : It works !!!
Here's a little video of the first experiment :
still need a little bit of tweaking (the artic ceramic is not well applyed on one element) but it's very promising !

Thanks again for your great tutorial

nothinglabs (author)  samol4 years ago
that looks great!!

>Here's a little video of the first experiment :
Thanks nothinglabs!
I spent last day trying to tweak the machine in order to visualize alpha rays.
In the end, I discovered that the americium button was on the wrong side :D !
It don't even seems to need a HV or electrostatic source.

nothinglabs (author)  samol4 years ago
if it makes you feel any better - the first time I took apart a smoke detector for americium - I got confused - and thought the piezo speaker was the source.

it seems sometimes I can get good alpha results without static - but other times not.

I suspect how cold the chamber gets and humidity (effecting background static charge) play a roll...

also seems less intense alpha sources do better without static. I think too strong alpha sources can ionize the entire bottom of the chamber - so the static source is needed to clear things out.

glad you're getting good results!
creatrope3 years ago
Is there anything special about those cpu coolers or will a $1 special from a computer surplus store (a cpu cooler with a fan) do the trick?

The inexpensive ones don't have the heat pipe or large fan - are they important or are you just looking for something with the proper orientation?
nothinglabs (author)  creatrope3 years ago
the CPU coolers specified for this project are pretty high performance.

they need to dissipate well over 100 watts. at very least you'll need something of decent size.

I've had some luck with larger non-heatpipe coolers - but I've generally gotten better results with the heatpipes.

there's probably no harm in trying a cheap random one (assuming it's not too small) - but if doesn't work - I'd suspect the cooler.

creatrope3 years ago
The 12710 on ebay all appear to be 106w, not 168w. Is this model number correct or is there more than one 12710?

is there an alternative source or model number?

thanks - creatrope
nothinglabs (author)  creatrope3 years ago
there are different ways to rate the coolers.

i took a look at those - they are rated at 15.2v max @ 10amps - which comes out to 152 watts consumed - so they're about the same.

i think they are probably fine.

picking coolers has some flexibility. ie - you'd probably also be fine with a 12706 on top and a 12709 on bottom

good luck!

KenRocks4 years ago
First off, thanks so much for the awesome build. I've been messing around on the net for a day or so trying to find a non-dry ice cloud chamber and yours is the first set up I've found that anyone can confirm function of...

One thing I'm curious about is generating the vapor for the chamber. Most folk seem to do it the same way as you: soaking a sponge/cloth/etc with solvent and attaching it to the top of the chamber. I was wondering if you had ever experimented with actually warming the solvent in another container and piping it over.

My goal with this would be to make a chamber I could run more or less indefinitely...anyone ever tried this?
nothinglabs (author)  KenRocks4 years ago
i generally use felt or sponge for the alcohol

yes - i've spent a fair amount of time experimenting with heating the alcohol. i've never had it produce noticeably better results though. (dont let this stop you from trying though)

as for having the chamber run longer - it seems the limiting factors / possible soutions:

the static source (balloon) - can last as short as 10 minutes - but as long as 60 minutes or more depending on the balloon and how much alphas the sample is emitting. a high voltage source would address this.

after about 60 minutes - you may get too much alcohol on the chambers bottom. maybe some felt to whick it back up from the bottom - or some kind of drainage system? this problem can be reduced by not over-doing the amount of alcohol on the sponge (more is not always better)

finally - you might run out if alcohol at somepoint - but you could probably drip more from on top of the chamber via a small hole.

i haven't intentionally run the chamber as long as possible that many times - but i've had at least several times i've run it an hour or longer without tweaking anything and still got good results.

good luck!
Ugifer4 years ago
Surely the plan would be to connect the power for a second or two at a time until you can detect whether the top is getting hot or cold. That way it never need go about about 40C and there's no scramble to disconnect it if it's getting hot. Great project BTW.
 I was totally waiting for a KABOOOOOOOM! *mushroom cloud*
Been having fun with the chamber for a few weeks now.  Have tried a variety of sources, including exposure to a dentist's x-ray machine:

Great little unit.  I have made a few changes, but overall the basic unit here will work fine. 
neimad4 years ago
 So far I have not read anywhere a mention of what type of alcohol to use. I'm sure this is probably an important detail. Some people may not know exactly what to use.  Can you please include what type of alcohol to use for best results?  

One other thing you can try is to "salt out" standard isopropyl alcohol to get 99% purity:

I did this today and it works as the alcohol separates to the top while the salt stays on the bottom.  Only drawback is that the resulting alcohol may contain traces of salt. 
nothinglabs (author)  neimad4 years ago
check step 2 - it's in there.

if you can't find 99% isopropyl - I'd try "denatured alcohol" (which is ethyl)

joseph seems to be getting pretty solid results with 91%.

I wouldn't bother trying 70% (tried it sometime ago - didn't work well at all)

you can also try "everclear" - if they have that at your liquour store.
(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago

nice looking video!
(removed by author or community request)
nothinglabs (author)  simplyscience4 years ago
I've switched to a balloon for static.

I've also switched to a 1/32nd fiberglass (garolite brand) base plate and am using a water glass for the chamber itself.

There isn't really a seal - but it doesn't seem to matter.

You can see the setup in a video at this blog post:   (also has some slow motion footage)

(there's also some other stuff at related to the cloud chamber)

As others have mentioned - if you get things really cold the static is less fussy.

I've been able to drop my temperature a few degrees by carefully re-apply all my thermal paste using a razor blade to both sides of each surface.  This has gotten me down to -40f!

(removed by author or community request)
nothinglabs (author)  simplyscience4 years ago
You ask a simple question - and there's a long answer...

The brand name of the fiberglass I'm using is "garolite" - 1/32nd inch thick.

The Garolite brand name is also used for other composite materials. has a bunch of "grades" - I'm using G-10.  This is the same material many circuit boards are made from.  Stuff is fairly cheap (under $10 for 1'x'1' sheet).

Unfortunately they don't have any of the glass-based versions in 1/32nd that are black (it's light green).  So - I've had to paint mine (krylon fusion spray paint works pretty well - but can come off).

They do have 1/16th" G-10 which is black.  I've used this with good results - but it seems thinner would be better.

They do have grade "XX" garolite which is 1/32nd and black.  Hower "XX" isn't glass-based - it's paper-based!  My research suggests it will work well - but I haven't tried it yet.

By using the razor blades I mean use them to apply a super-thin and even coat.  The razor blade helps because it's super-flat / sharp / hard.

I basically apply a bit of paste to the surface - and then use the razor blade as a putty knife - getting the coating as thin as possible.  Apply the past to both surfaces help further assure there aren't any gaps.

(removed by author or community request)
nothinglabs (author)  simplyscience4 years ago
I just got my garolite xx today (1/32nd)

performance seems comparable to the 1/32nd glass - and better than the 1/16th glass (I've got so many variables going on - it's hard to quote meaningful numbers).

one big thing I've figured out - sometimes I test in the mornings - when the house is still 65 degrees - other times I'll test in the evening when it's 72 degrees.

as you can imagine - that's probably resulting in about a 7 degrees final temp difference.

it's quite possible things are getting colder for you - but it's not enough to visibly make a big difference.

I just got some of these lights:

they work great!
(removed by author or community request)
nothinglabs (author)  simplyscience4 years ago
Since the bottom of the garolite is flat perfectly flat (unlike most containers) - I'm able to user a thin coating of thermal paste.

I've moved to the Arctic Silver Ceramique - it seems to work just as well as the Arctic Silver 5 for less money.

joniverson4 years ago
I built this device also a few weeks back.  I found that, although the fan cooling does a good job, you get slightly better performance (thicker, fuller trails) if the heatsink is immersed in ice water with the water being circulated by a fountain pump.  I know this takes away portability, but it's worth the extra effort if you plan on using it long term because the tracks also stick around longer.  Not as dependent on charge either (I found the extra cooling really helped with alpha sources without much additional charge).

One other thing you can do which helps somewhat, although not as much as ice water immersion, is to place a "blue ice" pack in front of the fan (just don't allow the pack to reduce fan circulation too much or you'll burn out the fan).  This also cools the chamber more, but obviously not for long term.  In a pinch however, it is as effective as ice water for a short period of time.  

Good luck to all who try it.  As mentioned, with the project in its current form, proper charge is the key to keeping everything working smoothly.  

Jon I 
My latest, best results using fairly strong uraninite:

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
that looks great! - nice radioactive sample..

think you've got a cosmic ray at 0:08!

I'll pass on a few more tricks I've come across recently.

For my most recently posted video - I've milled the bottom of the container flat - and colored the inside with permanent marker.

The milling added several advantages:

- It gave it a texture - so the black marker would more permanently mark - increasing contrast (when the inside of the container is smooth / shiny - it's more prone to rub off - especially once alcohol gets on it).

- It thinned the surface a bit - allowing better cold conduction

- I think a flat surface in general is a bit more optimal for supersaturation depth

I've also been playing around with using black fiberglass (any kind of thin black plastic would also wrok) as the base surface - and then placing a glass cup on top of it.  This leaves the bottom of the container unsealed - but produces surprisingly nice results.


(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
I used an actual mill to mill the bottom flat.  Not sure how I'd do it otherwise.

I've also been experimenting with the bottom of the chamber being just a flat piece of fiberglass or plastic - which is easy to spray paint black (Krylon Fusion spray paint works well).

The chamber is then just a drinking glass sitting on top.  Surprisingly the seal at the bottom of the container doesn't need to be that good either.

I'm using a Casio EX-FC100 - kind of mixed on the camera - but it does have a cool high speed (up to 1000fps) mode.
(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
nice video! - I'll add blacking-out the inside with a sharpie to the "tips" page (reason I moved away from the approach is the alcohol inevitably dissolves it making a mess)

the smoke detector americium is a major alpha emitter - so I think it "eats up" the un-ionized alcohol quicker than other sources.

may just try adding more static - multiple styrofoam cups - or maybe a balloon.

could also add a "baffle" like I did in this video:
(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
Do you have a manual focus option for your camera?  Once I figured out to use manual focus - the quality of my videos got much better.  (I can now see individual alcohol droplets)

With my setup I'm able to go a good long time without recharging - if varies a lot - but sometimes it seems to work almost indefinitely - 15+ minutes at least.

I think the balloon will give you more static - it's what I've switched too.

I've had similar issues when things weren't cold enough.  May also be too much or not enough alcohol on the sponge.  I've found with the top off - you need more alcohol (maybe a larger sponge)

markford4 years ago

Thought I would share my experiences with this. I use a coolermaster V8 heat sink, a top & bottom peltier as described (i.e. 5v and 12v from a MicroATX power supply) and 99% IPA and Arctic 5 compound. I used a clear acrylic square draw from a makeup container for the chamber, it is crystal clear and just the right size, I use 1mm matt black Polystyrene sheet (strengthened by sticking extra plastic underneath so it doesn't distort when cooled!!) as the base. then a piece of foam inside for the IPA.
I drilled a hole in the side for a bolt to hold an AM241 source from a smoke alarm. It Works really well, but one thing is Vital and people should be aware that static charges make a big difference to this chamber working! I use a small square of expanded polystyrene which I charge up by rubbing in my hair (several times in use) and place this on top, it only works if I do this, without it you get no trails. -  I would like to find a way of doing this automatically so will try an HV supply next. I think a small hole in the top of the chamber with a solder tag on the outside and washer inside bolted through to act as a small 'plate' should work.

Once again, those of you that have trouble, play with polystyrene or other plastic charged up by rubbing, it really helps, I think careful atention to the material used in the chamber construction may help this too, some plastics are better than others.


Absolutely!  It took me several days before I realized just how important the static field was.  I'm also trying to work this out with a high voltage supply.  One arrived today, but I can't get the tracks to respond to it so far.


I have some experience with Ion Chambers, and the Bias voltage for those is usually 30-100v it depends on the chamber volume and the radiation dose rate, too much voltage and you might even end up deflecting the paricles due to the field strength. All the field is doing is mopping up the electrons/charge that the source is creating and thus returning the charge to neutral.  The polarity of the voltage is all important, it has to be right otherwise you will make the problem worse of course.

Ion chambers normally have their bias voltage on a plate inside the chamber isolated from the outside of the chamber, so I would suggest a small metal plate (foil?) stuck inside the chamber with an insulated wire going to it, but try to avoid charge on the outside of the chamber try to get it only on the plate. - Theory is one thing of course!

some people claim success by just wrapping foil on the outside, this might also work, but the idea is to create a localized charge field inside, (that is not effected by external stray charges), the reason the polystyrene cup works is because it is simply blatting the whole setup with a massive electric field of many KV. I found too much charge on the polystyrene actually stops it working too).

(I'd love to know how you get on, feel free to email me at meteoritemark at maybe we can share info on this and crack this HV thing!)


(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
at least some AC HV power supplies have a "hot" wire and a ground - make sure you've got the right one!

(sorry if this is obvious - it wasn't to me just a few weeks ago!)

I get pretty decent results just adding a diode to a 2kv neon tube power supply (it's high frequency AC - not sure of the exact HZ off the top of my head)

I think once things get to a high enough frequency - a diode is enough to make it "look like" DC.

I wonder if one of those cheap electronic kitchen lighters would work, with a dioide on the end?

I Just tried a +150v battery made no difference on my setup, tried grounding the battery to the case and also just using the + side only ,  all with no effect on the trails.

Either the voltage needs to be higher, the polarity is wrong, or I need to make the plate nearer to the source, my chamber is quite big I suppose (10" square)

Has anyone tried connecting the bias voltage to the source? that might work too....

nothinglabs (author)  markford4 years ago
distance seems to make a big difference with the high voltage.

with a chamber that big (I'm guessing it's tall?) - might want to put the voltage source lower.

I've had good results with having the voltage at the top of a 4 inch tall container - but not with a 10 inch tall container.

What voltage are you using?

Yes it has to be DC, (and a stable reasonably rectified DC too not too ripply). From memory the old fashioned Griffin cloud chambers sometimes used +150v, (least they had a socket for 150v on the side) however I am not sure if this was for 'normal use' or just one paticular experiment, there is a chance it is just used to bend the beams.

As it happens I have a 150v battery, so am in the process of adding that to my set up. it will not draw any current so a bias battery should last for decades.

The voltage will depend on the chamber size, (bigger chamber should require a higher voltage field, one would assume!). We will probably find it works over a fairly wide range though, as it also depends on how much radiation you have in your chamber (18V is enough for a commercial air Ion chamber instrument like the RO-2, and that is seeing up to 500mSv/Hr).

Presumably the ground connection should go to a small plate on the base?
nothinglabs (author)  markford4 years ago
 > some people claim success by just wrapping foil on the outside, this might also work,  This is what I've found to work best with my 2kv power supply.  I tried adding a "ground" inside the chamber - didn't seem to help or hurt.  I've had some luck with as low as 26v from a DC power supply.  No luck with just using a diode + 120v / 60hz.  I'm guessing 60hz is "slow" enough where it needs to be better rectified to really act as DC. 
zmoney4 years ago
Do magnets affect the trail?
nothinglabs (author)  zmoney4 years ago
yes - but you'll need a rare-earth (strong) magnet - and the effect is somewhat subtle.
Yes.  See:
markford zmoney4 years ago
Haven't seen much of an effect  on Alpha's but  Beta's get skewed.
markford4 years ago
One observation I just made, it's the Alpha's that seem to need the bias voltage, I can still see Beta's and cosmic rays ect without an charge on the styrene cup!  (It's weird like the alphas arn't there, all the other tracks are though!)

Interesting that beta's dont seem to need any bias voltage?, , I wonder if they would act as a bias in their own right (they are electrons!), I wonder if adding a small beta source with the alpha would counteract the effect and keep the chamber 'sensitive' without a HV supply?

Worth a try..

(removed by author or community request)
nothinglabs (author)  simplyscience4 years ago
can you give me the full rundown of the hardware you're using?

(Any variation from the plans at all?)

Common thing I've run into is that peltiers are kind of fragile.  They can fail real easily if they don't have a good heatsink connection - or the wires get broken.

Other common issue is power supplies / wiring not being able to source enough current.

I would re-test your peltiers - make sure they're both providing some cooling when powered individually (make sure they have heatsinking when you test them...)

I've rebuilt things several times - and it's always worked for me  - definitely want to make sure this works for other people. (I'm not keeping any secrets!)


(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
I doubt that's the issue - I'm just using a single wire to connect to power (although using two or more for at least 12v source is probably best).

I've used both the 3.3v and 5v sources for the top peltier with good results - may want to try swapping that around.

I've seen the result your getting when things aren't getting cold enough.  I think what happens is it's just barely cold enough for traces at first - but once a little alcohol builds up on the bottom of the chamber - things warm up a little.

Is the heatsink getting warm?  If so - it's probably not being effective enough.

Have you re-tested both peltiers individually? (you don't need to take things apart - just power them one at a time - and verify you get some cooling for each)

If top pelt is bad - but the bottom is good - you'll get enough for clouding / some trails - but not very good performance.

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
some uneveness in cooling is typical - but that sounds like a lot.  thermal paste application could be the issue.

I recently tried just applying large dabs of paste - then pushing everything together to even it out - did -not- get good results.  smoothly applying thermal paste to the entire surface seems mandatory.

it sounds like you're probably running a lot warmer than things should - but I'll also mention I've learned to not trust thermometers.

for best reliability I'd try using a contact thermometer - and then use thermal grease to assure it has good contact with the peltier surface when testing (pushing it down with a bit of styrofoam helps further).

I've had some luck with cheaper thermal paste - so it's worth trying - but I've become a big fan of arctic silver ceramique - much cheaper than the "5" - but results seem as good.

did you switch to the 99% alcohol?  I have used 91% - but 99% works better.

is it at all possible the coolers are switched?

is the heatsink warm?  try measuring the pipes on either side of the platform.  anything warmer than 77f or so would be concerning.

I'm wondering if you've got a bad heatsink... (ie - maybe the heat pipes aren't properly filled..)

I've exchanged email with someone getting good results - and someplace in this thread - so I think the issue is just nuances.

this guy has been doing some variations on the design with good luck:

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
really glad to hear it's working for you!!!

I suspect now that you've gotten a taste for it working - you'll figure out a bunch of little tweaks for even better results (I tried to include all the big stuff in the instructions - but there's a bunch of intuition you develop after playing with stuff for a while)

I'll update the instructions to re-iterate the importance of the styrofoam cup - it really does work well.  plastic-wrap or a balloon are other good options (putting the cup "right-side up" with a static-charged balloon on-top seems to produce good long-lasting results.

I've (and another experimenter) have played with hi-voltage a bit - it does seem to be a bit more reliable - with voltages as low as 30v working at least to some degree.

are you sure it's DC not AC?

I'm using this 2000v power supply - just 6.50!

just add a diode to the "hot" wire to make it DC.


I think it was a DC source I was using, but I don't believe I had it set up properly to establish a proper "clearing field".  With the supply you recommend, how are you coupling the hot side to the unit?  I might as well get the diode at the same time, what number recommended?

I have to say, the reason I was so determined to try and get this working was because it is the only design I know of that does not require ice of any kind- be it dry or water ice!  This makes set-up and operation extremely portable and fast.  I did notice even more and/or thicker tracks appeared when I placed the heatsink in water before and I'll have to see what happens now that I've solved the cup issue.   Of course, adding a water container starts to sacrifice portability which would be getting away from my goal.  I've also thought of even a more powerful fan, if one exists in this size, or perhaps adding another fan somehow. 

I have yet to see any tracks from an internally place Americium source, but I think I know why.  Unlike the uraninite (which showed plenty of tracks), the Am was hung into the chamber through the sponge and I used a small piece of wire with alligator clip holding the Am.  When the cup comes near the wire, it's probably energizing the Am clearing away all tracks.  I'll simply try an internal placement with self support.    
nothinglabs (author)  joseph.raul14 years ago
I connect the hot side to a sheet of aluminum foil and place it over the unit (trying to add the ground to the bottom didn't seem to make a difference)

I'm using a 50v 6amp diode from radioshack - (276-1611) - but it's just occurring to me now that a bunch of the AC might be jumping right across it considering it's so under-rated for the voltage.  It seems to work fine though...

Considering that I've seen some results as low as 30v - maybe just rectifying wall power would work? (could also add a resistor to reduce safety concerns)

I have found that getting colder seems to make things work better in general / less need to have the static field just right.

I don't have numbers - but I've experimented with the following - and  feel pretty safe saying they will get things a bit colder:

1. Switch to a Corsair H50 liquid cooler (about $70) - this is fully contained - so there's no additional mess

2. Add an additional peltier (Tec1-12709) - powered at 3.3v

Note - if you do just #2 and not #1 - you may not see much of a gain - in that I think the current setup is already getting close to a traditional heatsink's cooling limits..

I'll also mention that I've never gotten any improvement from increasing the size of the bottom peltier.

Tip on americium - I've found it's really easy for the surface of the sample to get covered with alcohol - which blocks all the alphas...

Good luck!

nothinglabs (author)  simplyscience4 years ago
Joseph just got his working well - see later in the comments for details.

(make sure you're using the styrofoam cup!)

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
looks nice! 

I've made a special cover for my chamber with about a 1" hole in the top just for shooting video. 

The camera lens goes right into the hole - so it's almost as if it's closed - produces pretty nice results:

Using the head from one of those cheap multi-LED flashlights as lighting - but I'm running it off 5v from the power supply.

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
10f would be about the temperature I've seen when just the bottom peltier is running - not the top.

I'm currently running my chamber of an ATX power supply using the 3.3v tap as opposed to the 5v - because my particular power supply makes terrible noise when using the 5v.

5v gets things a few degrees (-28f vs -33f) colder - but I really can't see a difference in chamber performance.

I mention this because you might want to try it on the assumption your 5v tap is bad.

Good luck!  Happy to help with further troubleshooting..

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
if you're getting the alcohol precipitation - things are pretty cold - so you're probably not too far off.

the fact that the heatsink is getting quite warm indicates to me that it might not be adequate (a bit warm is probably OK - but if the pipes are over 90f - that probably indicates its a bunch over 100f under the pelts - which might be to hot.)

for reference - the pipes on my heatsink only get to about 80 degrees or so.

one option is to ditch the fan on the heatsink - and put the whole thing in a small container of cold water (maybe even toss in some ice).  it'll eventually heat up - but until then you should have very good cooling.

on the other hand - I've gotten things to work with a pretty low-end heatsink.

other stuff to check:

How's the thermal paste?  Too much - and it won't conduct well.  not enough - and you'll get air pockets which act as insulation.  this may be the #1 issue.

What kind of container are you using?

Are you using the Arctic Silver 5 or Ceramique - or something else?  I've gotten things working using really cheap stuff - but the good stuff does seem better.

are any of the wires getting warm?  if so - they may not be thick enough.

are you using 99% isopropyl alcohol?  (91% also generally seems to be ok)

Is the bottom of the chamber getting wet with alchohol?  If so - dry it off with a paper towel.

also - if your source gets covered with alchohol - that can block the particles.  (I've had this problem a lot with americium from smoke detectors)

one more thing - have you tried using the styrofoam cup / static source to "clear" any ions?  This can help a bunch.

Let me know what you find out - I'm sure you'll get this to work with a bit of tweaking.

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
glad to hear you're at least getting some results!

I'll mention to try out watercooling in the troubleshooting section.

which kingwin cooler are you using?

I specifically use the XT-1264 - but initially left the model out of the instructions - since I was having trouble locating them on ebay.  They seem to be available again - so I've added it to the description.

The XT-1264 has a 120mm fan - I've also noticed some of the Kingwin HTC models are smaller / have smaller fans.  If that's your cooler - that might be the prob.

Which peltiers are you using?  I've updated the descriptions a few times.  Any of the configs I mentioned should work - but the currently listed models are probably the best.

it's still possible the underlying reason for the heatsink getting warm isn't the heatsink itself - but it may be getting too heavily loaded (ie the container is too massive / too thermally conductive).

It's also still possible the warm heatsink itself isn't the issue - but just an inefficiency in the setup (ie thermal paste issues) - this might be overcome when adding the water cooling.

I've taken things apart and put the back together a dozen times or so - I've definitely had once or twice where the thermal paste wasn't quite right - and things didn't work as a result.

Are you using the suggested container - or something similar?

If possible - you may want to try -reducing- the voltage to either / both of the peltiers.  One easy way to do this is just add a beefy diode (will drop the voltage 0.7v)

Let me know - want to keep the instructable updated to reflect issues people run into / etc.
(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
I suspect the fan may be the issue - it would definitely explain the problem (I'd hold off on messing with the thermal paste).

If you stall these fans for for even a second or so - it can kill them (I've done this).  You can also damage them - so they end up running too slow (probably what's going on for you).

good news is you can get a replacement at radio shack for about $10 - "Thermaltake 120mm Fan" (they stock this in-store where I am).

I'm able to see tracks pretty much indefinitely with my heatsink setup.  If too much alcohol accumulates in the bottom - it may hurt results after a while (this can happen in maybe 15 minutes - but if the sponge isn't too wet - it's not a problem).  I know I've left it running for at least an hour - and possibly several hours - heat dissipation doesn't become an issue.

I'm using a multi-led 12v module I got off ebay (although the one I've got runs way too hot - so I won't recommend it specifically).  I've actually had really good luck with the multi-led flashlights - going brighter may help a bit - but not radically.  You may just want to try connecting your flashlight to the 5v on your power supply - little extra brightness / avoid eating batteries.

I think the light is about 60% "how you use it" (aiming it and your viewing angle) and about 40% power / size.

For diodes - I'd go with Radio Shack Model: 276-1661

(but again - I'd bet the issue is the fan)

Good luck!

(removed by author or community request)
(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
glad to hear your making progress!

the setup in the videos should match what's in the instructable exactly - I don't think I'm leaving anything out.

that said - I've definitely had times where I took it apart and put it back together - and things worked better / worse.  think this is mostly due to the quality of thermal paste application.

do you know how strong the pitchblende is?  it could be that it's not that strong a source.

have you played with the styrofoam cup for static yet?  this can make a big difference - especially with the americium (I think this is due to the large number of alphas it emits)

also - I've found it's really easy for the americium sample to get completely covered with alcohol due to it's small surface area (even if the chamber is mostly dry).  would definitely double check that (I usually just blow it off)

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
it's definitely possible to get some results off just one peltier - so maybe 1 out of the 2 is bad.  if you're getting clouding - I'd suspect the top one is bad.

(the bottom one had enough cooling to get to clouding on it's own - but probably not the top one)

I've found peltiers can be pretty reliable once they're "in-place" - but if you're moving them around / plug them in backwards / without a heatsink -or look at them funny - they may fail.  When experimenting - I regularly kill peltiers.

I don't know exactly specs - but the pelts should probably read well under 10ohms resistance - if they don't - they're probably bad.

it's really easy to damage the lead wires... (feel free to try to resolder them - may not be time well spent though)

I'm not 100% sure about the model details on the CP-12710 vs Tec1-12710.  I think they're a little different (CP1 being a bit more powerful) - but they should both work fine though (the exact peltier specs are pretty flexible).

Are you getting significant mist / precipitation?  (I've found you'll get at some precipitation before it's cold enough for tracks - but if you're getting a lot - maybe the problem is the temperature)

Do you have a thermometer?

"Cen-tech Pocket Thermometer" goes for about $20 on ebay - not saying it's actually good - but I think it's enough to tell you if you're getting close temperature wise.

(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
-18f is about when stuff starts to work - don't worry too much that some areas are warmer.

also - try measuring with contact thermometer by pushing it against the bottom of the chamber.

try putting a phone book or something else heavy on top of the entire setup to help push the thermal paste together.

have you played with the styrofoam yet?  this can really make a big difference.

may also want to try adding more alcohol than you already are (soak the sponge).  are you using 99% isopropyl?

does the clouding stay - or does it go away in a few minutes?  in general - when things aren't cold enough - the clouding will mostly go away in a few minutes.
(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
if you can't find 99% isopropyl - I'd try "Denatured Alcohol" (usually refers to ethyl) - hardware store probably has this.

It's commonly 95% ethyl alcohol - but the other 5% is probably mostly methyl alcohol (which is at least better than water).

You shouldn't have any problems ordering 99% iso online though - check amazon - they seem to have several brands under $5.

I've at times though the denatured alcohol was as good as 99% iso - but I've settled on the iso overtime - think it is slightly better.

I've had decent results with 91% iso - but the 9% water can't be helping.
nothinglabs (author)  joseph.raul14 years ago
wanted to toss some more thoughts at ya:

I just "simplified" my setup back to whats described in the instructable - this time using an ATX power supply for 5v and 12v.

things started working right away for me - so this brings down the documented differences to the CP-12710 vs Tec1-12710 - but I honestly don't think that's the issue.  I've had trouble pinning down the differences between the units - but I've done this general setup with a fair variety of peltiers.

sounds silly - but is the fan mounted "forwards" - I think the air should be blowing into the heatsink?

is your heatsink still getting warm?   For reference - I just measured a bunch of places on the heatsink - warmest I could find was 75 degrees.

are any of your wires warm?  what does the voltage measure for the different pelts under load?  (try to measure as close to the pelt as possible)

early on I had problems with too thin wires eating all my power / voltage.

also - if your sources are -really- strong - you might want to try something weaker.  it's possible your source is so strong - you can't (at some points I've seen this with americium)

another option is to put the source -outside- the container.  you won't see alphas - but you might see other stuff.
robotman34 years ago sells radioactive sources.
joniverson4 years ago
(removed by author or community request)
nothinglabs (author)  joniverson4 years ago
the biggest problem I've had with physically larger pelts is having a heatsink setup that covers their full surface area and can dissipate enough heat.

It seems like most of the nice / inexpensive heatsinks are about right for the small pelts.  The heatsinks with larger surface areas seem like they're actually lower-end - and can't keep up with the heat (I have - not- researched this extensively - there may be good options I haven't seen)

I have found if you don't more than say 85% of the peltier with a heatsink - you get poor results.

I've been in contact with someone working on a larger one:

I know he's gotten it working well using 4 peltiers (2 stacks of 2).

He's also playing around with water cooling.

Things get expensive / complicated pretty quickly...

I'll also mention that I've played around with the idea of attaching the top peltier to a copper (or aluminum) plate on the bottom of a cloud chamber.  This has -not- worked for me yet - but I think it shows promise.  (think part of the trick may be insulating the outside of the copper plate / chamber - so I'm not trying to cool the entire room)

jam BD4 years ago
nice use of two peltiers =D
(removed by author or community request)
nothinglabs (author)  joseph.raul14 years ago
the thermal paste provides a little adhesion - you might be able to move it in one piece of you're careful (I've done this)

I would probably try to find a box just the right size (or pad one with some packing).

The top shouldn't just "pop off" - the bigger concern is that it'll slide off (if it's not all the way off - you can probably just scoot it back in position).

Also - not a huge deal to put things back together if they come apart - I'd just have a little extra thermal paste on hand...

grenadier4 years ago
just bought some coolers, little bigger than you suggested. 138W and 168W. I'll tell you how they work when i get them.
nothinglabs (author)  grenadier4 years ago
I bet those will work fine (in fact - they may be the same ones - the sellers are always messing around with the ratings).

good luck!!!!!!
They definitely work. video coming in a week or so (im lazy)
nothinglabs (author)  grenadier4 years ago
Jayefuu4 years ago
Nice! It looks great. You appear to have used a LOT of thermal paste though....
nothinglabs (author)  Jayefuu4 years ago
yup - have to in order to make good contact with the bottom of the container.

however - if you have a container with a perfectly flat bottom - you'll be able to use a lot less.

I have used cheaper stuff than Arctic Silver 5 - and gotten good results. 

However - once I had some cheap stuff that didn't work well - so I'd figure I'd recommend the "gold standard"
This looks like a really great experiment.
One question: For me, a lot of things a are spent, to cool down
the hot side of the peltier elements.
Would it be possible to simply put them in a jar of water, after isolate them
electrically. The water will heat up, but maybe not that fast,
and also normal ice could be added or the water could be exchanged by new cool water.
What do you think of this idea?

All the best and thank you for your idea.

Steve Mc Craft
nothinglabs (author)  SteveMcCraft4 years ago
I (and many others) have used water cooling by running water over the heatsinks - and that works.

not sure if your idea will or not - sounds plausible though.

give it a try!
do you think, it is possible to drown the peltier element in water, without
isolation. The voltage is not too high, therefore it might work.
But it depends how the peltier itsself is made.
I could "suck" the water, then it would not work.
Does anybody has an suggestion or experience?

If this works, this could cheapen the whole thing immensly - I think...

All the best

endolith4 years ago
Can alpha particles get through the cover?
I don't think so. Alpha particles are basically He nucleuses (i.e. atoms stripped of their electrones), which are quite large, therefore even a plain sheet of paper can stop them. Despite that the thick and short tracks indicate alpha particles, so it MIGHT be possible for them to get through nevertheless. The longer tracks are beta particles (a plain electron or its positively charged counterpart, a positron).
 so all those tiny wee specs are alpha particles, and the large ones, (like in the last pic) are beta ? and gamma would not be seen at all as its a wave ?
Actually it's the opposite I think (short thick ones=alpha, long thin=beta). Gamma radiation can't be seen as it's not a particle in its classical sense (but a high energy photon=light wave instead) and a cloud chamber is used only for radiactive particle detection. So you'll have to get a Geiger-Müller counter for gamma rays, sorry :P
I think the distinction is that cloud chambers only see charged particles.  If you put a big magnet nearby, they should go in spirals, too.
Yeah, but that's kinda logical as the CRT devices use the same principle to deflect the electron beam to form the picture (using a sawtooth signal for horizontal sync). But only if you have the particles to begin with :P
Oh wait.  The sample is inside the plastic, so that's irrelevant.  My bad.  :)  Tobacco has polonium in it, which emits alpha particles.  I wonder if it's enough to be visible.
I bet it is. You see if it is able to "display" background (cosmic) radiation then the tobacco would certainly do as well. And also: every single living organism (and even the dead ones) contains carbon-14 isotopes (an unstable isotope of carbon), which decays through beta radiation, although at a very slow rate (a half-life of ~5730 years) so I'm not sure whether those would be visible as well :P
BTW you can see the background radiation by your TV as well. All you have to do is to pull out the TV cable or tune out from any TV channel and voila! You have the background radiation visible on your screen (AKA static :P).
Yeah, but if it's really minor you won't be able to tell the polonium from the background radiation.  Even the leaves will probably shield alpha particles from the inner leaves.

The background radiation on the TV is just RF noise.  It's not ionizing radiation from decaying nuclei.  Your body emits infrared radiation, but that's not very interesting.  :)
I didn't say that background radiation is particle radiation. But the background radiation is almost like white noise i.e. it has a more or less constant amplitude from DC to the gamma range. But there's also beta-radiation coming from the outer space (and yes, nulei DO decay in the distant areas at the same rate as they do over here :P).
And yeah, you're right. If the radiation is small enough you won't be able to tell it from the background radiation (unless you shield the testing chamber appropriately with lead bricks or something).
MaXoR4 years ago
Umm, I am interested in something, why do you say "Do not use 12V to power the Top peltier?

What did you experience with the 12V power, and why did you decide on 5V when they are 12V peltier's?
This is a very valid point, and I would like to add to it (ATTENTION AUTHOR):
Peltiers rated to run at 12v are quite happy to run from anything between about 4 and 18v, with differences in their current draw and efficiency through this range. Actually, they are far more efficient running at around 15v.
Running at 5v will increase waste heat output, decrease heat transfer (cold side doesn't get as cold) and generally make stuff work less well.
However - there is a reason why this configuration may work better.
Peltier cascading is a well known, frequently used technique to cool stuff to very low temperatures. Temps down to -200C can be achieved (-328F) by cascading 6 or 7 different ones. However there is one vital key to making this work: the peltiers must be cascaded in power output also. (Usually, size scales with power, so you get a pyramid of cascaded peltiers)
To clarify: You can start with a 500w one on the bottom, then go down to a 400w, 300w, so on (with whatever is available), right down to teeny tiny little 4x4mm ones. The reason that these temperatures can't be achieved in the first place is that peltiers have a maximum temperature differential they can sustain - e.g. if the hot side is kept to 30 degrees, the cold side cannot go any lower than 0 degrees, but if the hot side is kept to 0 degrees then the cold side can get to -30. (31 degrees Celsius is the usual differential figure for cheap peltiers)

I would also like to add another point to this: Reversing power to a peltier will not damage or harm it in any way. It simply runs backwards. (I.e. hot and cold sides are swapped)

One thing to remember: a peltier is simply a heat pump (and not a very efficient one). It doesn't "make" cold; nothing can "make" cold as cold is simply the absence of heat. All it does is transfer heat from one side to the other. It does this in exactly the same way no matter which way it is connected - it just swaps sides. You also have to remember though - the hot side heatsink needs to not only dissipate the heat transferred from the cold side, but the 'waste heat' generated by the inefficiency of the device. (And in this case, the transferred and waste heat from the second peltier as well.)
(This leads back to cascading; the reason for this is that as you move downwards (larger and larger), each successive peltier not only has to move the heat from the ones above it, but their waste heat also - which accumulates with each layer.)

Hope this is useful :)
nothinglabs (author)  BlueFusion4 years ago
yup - I agree with everything you shared 97% - I added a bit of this info to the final step in the article

I will say that stacking is a bit harder than it may seem though - the loses end up getting big really quick

> I would also like to add another point to this: Reversing power to a peltier will not damage or harm it in any way. It simply runs backwards

however - if you have the heatsink on the "cold" side instead of the "hot" - things will get really hot really quick (could be bad...) - that's the reason I don't want people getting polarity reversed.  however - polarity reverse + upside-down - that's fine.

I will also say I've read peltiers are more effecient at lower voltages - but I can't say that definitively.

there's also a big pragmatic side to this project - dealing with much more than 200 watts of heat quickly gets outside the range of over-the-counter PC cooling hardware!

>I will say that stacking is a bit harder than it may seem though - the loses end up getting big really quick

I agree 100%. Peltier cascading is a really, really inefficient way to get stuff very cold - even so, it is often the simplest and cheapest way to supercool small objects.

>however - if you have the heatsink on the "cold" side instead of the "hot" - things will get really hot really quick (could be bad...) - that's the reason I don't want people getting polarity reversed.  however - polarity reverse + upside-down - that's fine.

Yup. I did suspect that that may be your intention. I might add though that I've run peltiers like this accidentally for several hours with no ill effects though YMMV.
I have also managed to turn a 130w P4 Prescott heatsink into a solid lump of ice while connected to a big chunky amplifier heatsink :D

Great follow up comment to my question. I think you actually answered my question in your statement.

I've wanted to attempt to design a PC cooler for my CPU with one of these, and your explanation will allow me to finish my design!!! I want to directly mate one of these between the heat sink, and CPU, then hook her up to the 12V line, and now I realize I can also add the 5V line to up it's output even more! (Was your statement about it's heat difference(30C), at maximum voltage?)
Unfortunately trying to do that will blow up your PSU. All PC power supplies these days share a common ground for all voltages - that is, the ground lines for 5v and 12v are connected. Trying to series connect the 5v and 12v lines like yuo describe would constitute directly shorting the 12v line.
Also: Depending on your processor, you will need a fairly big peltier to shift its heat. You will also need an even bigger heatsink to deal with the waste heat created in the process.
There is a far better (IMHO) way to do this though: dual-loop water cooling + peltier.
To summarise: You set up two separate water cooling loops. Loop 1 connects to a waterblock on the CPU, via a pump & reservoir, to a waterblock attached to the cold side of the peltier. This loop has no radiator.
Loop 2 connects to a waterblock on the hot side of the peltier, via a pump & reservoir, to a radiator.
This also has the added benefit of allowing you to physically shift your heat dissipation (radiator) well away from the rest of the PC; allowing more efficient air cooling for the rest of the components.
grenadier MaXoR4 years ago
you can't hook the 5v in line with the 12v in an atx supply.
vidyadhara4 years ago

I missed this link on the previous one. 
vidyadhara4 years ago
I saw a video on you tube that suggests using a gas lantern mantle. Apparently this thing contains thorium oxide. And thorium is a radioactive material.
The whole thing is harmless and widely used. This was a most popular variety of lighting until the coming of the electric bulb. 
Calorie4 years ago
It's a fun setup. And its nice that you don't need to deal with dry ice.

Just keep in mind:


Any alcohol has a flash point. It's the temperature at which alcohol vapor will ignite without a point source. Although unlikely, the other side of the cooler could be hot enough. More likely is some sort of arc. But you don't have a lot so it wouldn't be some sort of humongous whosh.

Emitter Source:
All of the sources you are using are weak decayers, and emit few particles per unit, so that limits any potential problems.

An alpha emitter (the helium nucleus) from these sources are weak. What that means is that they leave their decaying element at relatively low speeds. They can't get through jack. Look how far they can travel in alcohol vapor. That's mostly air, and the alcohol is present for a condensation trail.

A beta emitter is an electron. They are a bit more energetic by nature.

If anyone is curious to how a radioactive source does damage, well it's where it ends up. If you look at your handy dandy periodic chart you'll notice that there are columns. Within those columns each element below the other in the same column behaves roughly the same way chemically.

Alpha emitters do more damage inside of a body because they bounce around until they have stolen enough electrons to become stable. That's bad.

Beta emitters aren't as bad in the body. The electron zips along until it thumps into something. At that point the protein or fat or whatever is damaged. However, there are certain classes of chemicals, known as anti-oxidants, that deal with this sort of damage. Vitamin C (Ascorbic Acid) is a notable example.

You'll also find it in many of your food items as a preseravative to delay spoilage due to oxidative damage. But in that case the damage comes from oxygen itself.

And God knows, if anything being sold was remotely strong the NRC (Nuclear Regulatory Agency) would be on the seller like white on rice.

Have fun and good job!
Can you see trails from radiation of granite or other igneous rocks.
nothinglabs (author)  build a BOOM4 years ago
it's hard to say - it sounds like some granite is very slightly radioactive - you may see some tracers - but if it's infrequent enough - you might have trouble figuring out if it's background radiation / cosmic rays or the actual sample.

can't hurt to try!
granite tends to emit MUCH more gamma than beta/alpha
Could you provide more details on this step?

Does the bolt and sponge need to be air or particle tight, or at least close to slow the evaporation of the alcohol down and aid the process in the chamber? ... Or is it to allow fresh air to sneak in around the threads?

Just a little nervousness regarding nuclear leaks and vaporization reactions in a small apartment downtown Metropolis... or the mention of radiation puts a fear of zombies into the night air.
nothinglabs (author)  fightthemachine4 years ago
does not need to be perfect / air-tight.  leaks at the bottom of the chamber are bad - leaks at the top aren't really a problem (can even run the chamber with the lid a little off)

I'm not going to suggest you ignore concerns about radiation - but if you're buying something on ebay - it's probably not that dangerous (just wash your hands after touching - and don't eat it)
Will it affect/harm little kids, kittens or houseplants if used too long or allowed to radiate?
nothinglabs (author)  fightthemachine4 years ago
you can always build it without using a radiation source - and then see the periodic cosmic ray.

from a radiation standpoint - this project isn't any better / worse than the radioactive source. 

biggest safety hazard is probably plugging the coolers in backwards - then starting a fire. 
I'm not concerned for radiation, more air quality in a small stuffy apartment or in a garage downtown, I would love to show off the reaction, just without combustion of alc or sparking reactions getting the landlord involved, he gets more nervous than I am regarding particle physics lol.

Great detail in the steps. I built a Holly Batchelor cloud chamber. Have you considered an aluminum plate to replace the plastic bottom in the dish? This is definately on my buld list. Thanks for sharing! 

nothinglabs (author)  RainyDayMaker4 years ago
I have tried the aluminum bottom - but when I did I was clearly being too ambitious and hoping to cool a larger chamber than the peltiers could.

I've also tried cutting out the bottom - and exposing the peltier directly to the inside of the chamber - this worked pretty well.  However - it didn't work any better than the approach of connecting the peltier to the outside of the container.

Getting the seal perfect on the bottom of the chamber seems pretty critical - and has been a problem for me in the past - so I'm pretty happy with leaving the bottom of the container intact.
jeremydeath4 years ago
If you are willing to spend a little more money, there are some great radiation sources over at They have some really cool stuff there.
westfw4 years ago
Really nice update of an "old standard" to modern technology.  Very nice job!

kelseymh4 years ago
Whoa.  This is most excellent, and particle physics, too!

It looks like you can build this for under $40, not counting the sources.  Is that about right?  That's a really great price for a junior high or high school science class project.
nothinglabs (author)  kelseymh4 years ago
yup - if you already have the power supplies - you can probably build this for in the $40 range.
kretzlord4 years ago
Awesome 'ible!  its amazing the things you can do at home with a little ingenuity and time.  be careful with using too much arctic silver, i burnt out a CPU once because i had it applied too thick and it acted as an insulator. Great job though!
lemonie4 years ago
That's great, pretty easy to do and effective.