Observing cosmic rays and radioactive particles using parts from Aliexpress
At some point in physics class I learned about ionizing radiation and found myself staring at this image:
Somehow, people were able to take photos of radioactive particles over 100 years ago using this apparatus:
Surely with advances in modern technology (and AliExpress) I could achieve the same at home?
Before building anything, I needed to understand how a cloud chamber works: supersaturated vapor is produced in a chamber, so that when a charged radioactive particle passes through the vapor it ionizes gas molecules along its path. These ions act as nucleation points where the vapor condenses, creating a misty trail (superficially similar to airplane contrails).
To create the supersaturated vapor, a vaporization liquid is boiled at the top of a chamber and cooled at the bottom of a chamber. This temperature gradient is what causes supersaturation: the warmer air can hold more vapor than the cooler air, so as the air from the top of the chamber falls toward the cold end of the volume, it reaches a state where it holds more vapor than it can ordinarily hold at that temperature - it becomes supersaturated.
Due to the process of vapor being continuously evaporated at the top of the chamber, diffusing downwards (due to gravity) and progressively cooling until supersaturated, this apparatus is known as a “diffusion cloud chamber”.
The vaporization liquid which is most commonly used is isopropyl alcohol, due to its reasonable boiling point (82C), ease of purchase, and low cost. I used the same.
To achieve the cooling, a common method is to place dry ice underneath the chamber, creating a very cold floor. This is cheap and simple, but is a short-term, one-time solution. The dry ice is a consumable and lasts a few hours at most. There is a much more interesting approach to this problem which can run continuously and on-demand: thermoelectric cooling with Peltier coolers.
By applying a voltage across a stack of n-type and p-type semiconductors, a temperature gradient is created. With a sufficiently large stack and adequate cooling of the hot side (to extract as much heat as possible), a cold floor can be achieved beneath the chamber.
With these fundamentals figured out, my process was the following:
The chamber needs to achieve a temperature below -26°C (the temperature at which alcohol vapor supersaturation occurs), which means the floor of the chamber needs to be cooled even more than that (likely under -35°C). To achieve this, I used multi-stage cooling.
As always I began by surveying other successful projects, and this one provided a useful diagram of a thermoelectric cooling system:
I took a similar approach with my cooling system: a stack of 3 thermoelectric cooling stages (Peltier modules), one water cooling stage, and one air cooling stage. In this arrangement, heat is “pumped” from the metal floor of the chamber through two layers of weaker thermoelectric coolers before being pumped through a more powerful thermoelectric cooling layer - preventing any individual Peltier module from being overwhelmed. The watercooling layer removes moves most of the heat away from the cooling stack (to an air-cooled radiator), and an air-cooled heatsink handles any residual heat below the watercooling block.
The cooling system begins with the metal floor of the chamber. Wherever metal was required in the cooling system, I used copper due to its high thermal conductivity (one of the highest among commonly-used materials). The only drawback to copper is that it is relatively expensive.
To maximize the particle viewing area, I used two 1/4" thick copper blanks side-by-side to achieve a 4" x 4" floor:
To maximize contrast when viewing particles, I made the chamber floor as dark as possible by placing black aluminum foil tape on top of the copper:
Every metal surface which interfaced with another layer of the cooling stack needed to be as flat as possible for maximum thermal transfer. The brilliant Main Presenter of the Tech Ingredients YouTube channel describes a classic lensmaker grinding process to achieve this in his video on thermal interface materials:
As a final step to maximize thermal transfer, I used Noctua NT-H2 thermal compound between all thermal interfaces:
The active elements of the thermoelectric cooling system are cheap and readily-available Peltier cooling modules. These come in a standard size of 40mm x 40mm, (roughly 1.5" x 1.5") so a 2 x 2 arrangement worked nicely under the 4" x 4" copper floor. I arranged stacks of 3 modules each, with the most powerful module at the bottom:
Each layer of the Peltier stack requires a different voltage, provided by multiple DC power supplies:
The watercooling system is a simple design:
Heat is taken out of the thermoelectric cooling system via two copper cooling blocks underneath the Peltier stack:
The hot water (and cool water on the other side of the loop) passes through two PVC impellers to measure flow rate (with temperature sensors installed to monitor water temps):
The hot water passes through a large radiator, which has 3 industrial fans attached to it:
The water is pumped from a large cylindrical reservoir using a generic pump:
Everything is connected using soft silicone tubing, backed up with small hose clamps:
The assembled watercooling setup:
I used a large copper heatsink to radiate any residual heat from the Peltier stack:
I built a 3D-printed base to house the entire chamber cooling stack, with small 12V fans blowing across the copper heatsink final stage:
With the cooling system completed, I moved on to measuring the temperature of the chamber floor.
The target temperature at the chamber floor is below -35°C, which is too low for most common temperature sensors (commonly-available temperature sensors have reduced accuracy below -20°C to 0°C). Affixing a Type-T thermocouple to a hole drilled into the chamber floor was the ideal solution to this problem. With a range of −200°C to +350°C, it provides accurate readings at any temperature the chamber floor can reach.
A Type-T thermocouple was easily obtained:
Measuring temperatures with the thermocouple is a much harder (read: expensive) problem - digital thermometers compatible with Type-T thermocouples aren’t cheap. Building my own digital thermometer was a quick and inexpensive alternative.
For the microcontroller I used my favorite, the Arduino Pro Micro:
To interface with the thermocouple I used a Universal Thermocouple Amplifier from Adafruit:
And for the display I used the ubiquitous backlit 1602 (16 characters, 2 lines) LCD display:
Everything was wired up on a protoboard:
And I designed a 3D-printed case:
The final product:
I then moved on to designing the diffusion chamber.
The diffusion chamber needed to enclose the copper floor within a sealed transparent cover. The top of the chamber needed a heating element to evaporate alcohol. The copper floor needed side-lighting to enhance contrast when viewing particle trails. And lastly, a high-voltage ion scrubber was needed to improve the performance of the cloud chamber.
I started with a mock-up of the assembled chamber:
The cover was an easy problem to solve: 5 sheets of 6"x6" acrylic (2mm thick), bonded together using Weld-On #4:
Wherever the cover touched the floor, I used neoprene foam to ensure an air-tight seal:
Side-lighting with white LEDs was used to maximize contrast when viewing particle trails. Two rows of 5mm super-bright white LEDs did the trick:
I designed a 3D-printed mount for the LEDs so they would snap into place on the chamber floor:
Evaporating alcohol at the top of the chamber was simple. Felt squares were soaked in alcohol and a simple loop of resistors heated the felt. The felt was placed on a wire mesh, which is raised off the chamber floor using delrin rods:
The rubber stoppers holding the wire mesh in place are an unusual item: cattle castration rings.
Applying a high-voltage electric field across the chamber volume provides better viewing of particle trails due to a variety of effects:
To achieve this, I harvested an electric fly-swatter for its high-voltage power supply and metal grid:
I cut out the metal grid and placed it above the viewing area. The copper floor of the chamber is the ground reference.
With all of the components ready I assembled the full cloud chamber:
The watercooling system worked as expected. I added ice packs to the air-cooling system for even better performance:
I tested the lowest temperature it can achieve:
-37.72°C is more than sufficient for supersaturation. The cloud chamber was ready to reveal particle trails.
Background radiation:
Thoriated tungsten welding rods:
Being able to see the path of radioactive particles is a luxury, usually we are only able to detect their presence through the horrifying click of a Geiger counter (or similar detector). What if I combined the two? By placing a Geiger counter close to the viewing area, I could observe the particle trail and hear a confirmation of the particle hitting the Geiger–Müller tube. This was a ludicrous idea and I do not believe the correlation is so simple (e.g. there are radioactive particles coming from every direction at all times), but it was a fun experiment to try.
As usual, I started by making my own Geiger counter.
Thanks to the Chernobyl incident, eBay is littered with surplus Soviet SBM-20 Geiger–Müller tubes:
Using the SBM-20 as a detector, there were countless examples of simple geiger counter circuits. I used this one as inspiration. The circuit is based on a 555, and pulses from the SBM-20 tube are used to drive an LED and a speaker.
I built the circuit on stripboard, and lasercut a clear acrylic case:
The final product:
Placing the Geiger counter next to the viewing area of the cloud chamber, I achieved my goal: capturing the trail of a particle before it hit the Geiger–Müller tube: