# How to Build a Cloud Chamber

2,672

55

17

I've been building a cloud chamber for the last 2 years on and off. It's finally reached maturity and now i'd like to share a video to explain how it's done.

This video shows some good tips and tricks in home made cloud chamber design. It mainly focuses on such details I would like to have known during the building process. Other information, such as how to build a water cooling system, is already available at many computer hardware sites.

The whole rig costs about 250-300 € not including the tools to build it. During these 2 years, I must have spent well over 800 € on it, trying out different solutions to problems along the way. In the end, this design is very simple, reliable, and easily replicable.

I also have an older video with the early construction process, some failed attempts and the first iteration of a working machine: Youtube link

### Teacher Notes

Teachers! Did you use this instructable in your classroom?
Add a Teacher Note to share how you incorporated it into your lesson.

## Recommendations

• ### 3D Printing With Circuits Class

6,887 Enrolled

## 17 Discussions

Wow, this is fascinating! I was quite intrigued and really enjoyed your video.

Hey, and welcome to the site! I hope you'll share more of your projects here in the future.

Hello David,

I'm currently using a 240P water cooling system made by cooler master with a tec12715 and a tec12710 on top. Temperatures average at -15 degrees celsius, which is obviously insufficient for the diffusion effect to occur. However, after some considerations I discovered that in a normal dry ice cooler, a hot substance was put on top of the chamber to create the temperature gradient. Using the same theory, if I sacrifice my "ion scrubber" on top and put a hot object, wouldn't this create the diffusion effect which can visualise particles? Even though I have -15 at the bottom? How can I achieve lower temperatures otherwise with what I already have?

Hi! I believe you need to use a weaker top-peltier than 12710, as a 12715 may have a hard time transporting all the heat. Perhaps try feeding the 12710 around 5-6 V, or running it in series together with a second 12710.

As for the heat source, it's not a good idea when using peltiers as any excess heat has to be dealt with by your peltiers and thus affect its minimum temperature at some point. With dry-ice (or compression cooling) you have plenty of cooling capacity to spare, so you can afford to have a heater for your alcohol. Personally I don't really think its necessary, as the heat that leaks in through the glass dome is enough to maintain a supersaturation above the cold plate (due to a sufficiently large temperature gradient).

I am not entirely sure how the diffusion process works, but in my experience reaching the -25 C mark is a must; its not just about having a proper gradient. Also, I believe that the sole purpose of the ion scrubber is not only to remove naturally ocurring heavy ions suspended in the gas, that may deflect incomming radioactive particles.

My best guess is the following model: Alcohol molecules, being dipolar, can spontaneously build droplets with some surface tension if the temperature is high enough, because then the kinetic energy is enough to overcome some potential energy barrier related to the surface tension. These droplets look like a constantly moving cloud or mist, on top of the cold plate, and the few radiation trails that are not deflected by ions, mostly drown in that mess. However, below a certain temperature, lets call it T_critical, alcohol molecules need a condensation core or a small perturbation to make droplets. If this is somewhat correct we can draw some conclusions: 1) Naturally occurring free ions may provide condensation cores for making undesired droplets even below T_critical, which means that 2) an ion scrubber, upon removing these ions, leaves a volume of (desired) supersaturated gas ready to condense around a trail of ions left by alpha/beta/muons. 3) An ion scrubber may also align the (dipolar) alcohol molecules making it even harder for them to condense into droplets on their own. 4) Last, but perhaps most relevant to your question, is that if any of this is accurate it suggests that reaching a low enough temperature is a prerequisite more important than the gradient. In other words, no temperature gradient would allow a working cloud chamber if the temperature isn't low enough, since you simply need to get the mean kinetic energy down below some threshold value for the process to occur.

I mostly like this idea due to my experience playing around with the machine and the ion scrubber, but I don't have anything solid to support it.

Hi again!
We have gotten ourseleves basically everything we need.
When testing the TEC's seperatly we get the 12706 to -4 celsius and the 12709 to only 0 degrees, stacking them will only get us to -2 degrees.
What would you reckon the problem is?

5 replies

Hard to say really. How are you cooling them? What voltages are you running them at? What temperatures do you have at the hottest side?

A 12709 will probably have a hard time cooling a 12706; you'd need to lower the 12706 voltage somewhat so it doesn't produce so much heat.

We ran them both at 12V since we only have one PSU. Might need to buy a volt regulator to plug in also.

Firstly we just now have an aluminum heatsink which we placed them on, and the heatsink was in water to transport the heat somewhere. But isn't it wierd that the 12706 got much cooler single than the 12709? We got the best values without thermal paste at all, might have been low quality we got from school. Will test with other thermal paste also.

To lower the voltage i recommend connecting units in series. For instance, to get the 12706 to run at 6V, connect 2x 12706 in series. Of course then you can't stack that pair on top of each other, but you could put each on top of their own 12709 running at full 12V. The problem with voltage regulators is that you lose lots of power on the conversion, and peltiers don't run well on PWM regulators, for instance.

Thermal paste is absolutely essential. You won't get anywhere without it. The fact that you're getting better results without tells me that either you are applying too much / too little of it, or that your surfaces are truly epically flat (unlikely). Even cheap thermal paste should be better than none. Make sure to apply lots of pressure to squeese the paste out, and any trapped air bubbles.

Heatsinks differ quite a lot depending on their inteded usage. Are you using PC heatsinks? You need to use heatsinks comparable to (or better than) Cooler Master Hyper 212 Evo, that can handle 150W or more at the temperature range in question. The rule of thumb is have lots of copper, lots of fins, and plenty of airflow (or water).

Are you getting ice buildup? It takes lots of needless power to build ice. Try some drops of isopropanol if you have, to get rid of frost.

Also, the temperature on the hot side should not be above ~40. Remember that the maximum delta T of these units is around 40-50 C in practice, which is why you need to stack!

Oh, and are you getting any condensation and buildup of ice? that could also impact the performance near 0.

What thermal paste are you using? too much? Too little?

Hi!

The bottom peltier element needs to have enough cooling capacity to transport the heat load that is applied to its cold side. In this case there are two types of heat source: resistive heat generated by the top peltiers, and any heat load applied to the cold side of the top peltiers.

In other words, each peltier unit will generate resistive heat (P = U*I) even before any external heat is applied to them. That heat must also be carried away. Thats why when you stack peltiers, the bottom peltiers must always be stronger than the top ones.

I find it is easiest to convince myself of this fact by trying to imagine a situation with two identical peltiers stacked; in the best case scenario you've just built 1 peltier that is twice as thick that produces twice as much heat, and performs as well or worse than the one you started with.

We are building one for our last high-school project and want it to look good also.
We thought about using 2 similar peltier elements so we only would have 1 PSU. With 2 different peltiers you will need 2 PSU:s, right?

Great choice of project! I suggest you start as early as possible, as the building phase can take considerable time. Make sure to use proper, rigid mounting for the peltiers that apply even pressure so they don't crack.

Using similar peltiers is possible but you will have to lower the voltage to the top peltier somehow, and if your PSU only outputs 12V that can be challenging. Even so, running a peltier at half voltage (half power) is not as efficient as using one running at full voltage at that same power. I suggest you try combining pairs similar to TEC1-12709 and TEC1-12705, both running from a 12V PSU. This will allow you to cool down a surface of about 6x6 cm sufficiently, give or take a centimeter.

Using a PC PSU gives you a 5V rail to run peltiers, but make sure the power output is for 5V and 12V rails are sufficient. In any case I wouldn't recommend a PC PSU, in part because of the efficiency issues i mentioned.

Lycka till!

Thanks for all this useful information. Looking at buying the PSU you linked in description, seems good and would fit good in some sort of a box to make it look good.

Thanks once again.

By the way, have a look here for more inspiration on building the chamber. I wrote about the early stages of my build there as well:

http://elektronikforumet.com/forum/viewtopic.php?f=3&t=66726