Skip to main content
shopping_basket Basket 0

Building a Cosmic Ray Detector Part 1: Introduction and Planning

Andrew Back

Sensor used in construction of CosmicWatch
Constructing the CosmicWatch desktop muon detector.

CosmicWatch is a project from MIT in the US and Poland’s National Center for Nuclear Research, that enables anyone with basic electronic skills to build a low-cost desktop detector for the muon particles which are created when cosmic rays collide with the earth’s atmosphere.

In this series of posts, we’ll take a look at constructing the CosmicWatch muon detector and where necessary, modifying the design very slightly to make use of more widely available parts.

Primary and secondary

Image showing primary cosmic rays: Image Source CERN
Image source: CERN.

Primary cosmic rays are high-energy protons and atomic nuclei that travel through space at close to the speed of light, with evidence to suggest that a significant fraction originates from the supernova explosion of stars. Upon impact with the earth’s atmosphere, these then produce showers of secondary particles, including muons.

A muon is an elementary particle — which is to say one that is not composed of other particles — that is similar to an electron, but with a much greater mass of around 207 times that of an electron. Thanks to their mass they accelerate more slowly than electrons in electromagnetic fields and hence for a given energy are able to penetrate much deeper into matter.

It is these muons or secondary cosmic rays that we are interested in since they are able to penetrate the atmosphere, reaching land and even into deep mines.

Detecting muons

When we think of particle detectors the first thing that usually comes to mind is a Geiger counter, which makes use of a tube filled with inert gas, to which a high voltage is applied and then we count the ionisation events that result as particles enter the tube. This has a number of limitations and perhaps most notably that the signal output is always of the same magnitude, regardless of the type and energy of the incident radiation.

Scintillator material
Scintillator material.

A somewhat more sophisticated alternative is to use a scintillation detector, whereby instead we have a scintillator material that produces a flash of light upon incident radiation, which is then observed by a light detector. Scintillation detectors offer numerous benefits over Geiger counters and this includes that, through the selection of the scintillation material used they can be constructed to respond to a particular type of radiation. In addition to which they are also faster, much more sensitive, and are able to measure the energy and intensity of radiation.

Image showing a vintage photomultiplier tube-based scintillation detector
A vintage photomultiplier tube-based scintillation detector.

Traditionally this type of detector has utilised a photomultiplier tube (PMT) to measure the light generated by the scintillator material, since these are extremely sensitive, to the point where they can detect even a single photon striking their surface.

Image of a Photomultiplier tube
Photomultiplier tube.

However, there is a downside and PMTs are not only relatively expensive but also require the use of a high voltage power supply.

Image of a SiPM device
SiPM device.

Fortunately, in more recent years we have seen the advent of the silicon photomultiplier (SiPM), a far more compact and eminently convenient semiconductor alternative, that like the PMT, may be sensitive to a single photon. CosmicWatch makes use of a C-Series SiPM sensor from onsemi (185-9609) , which can be seen pictured above in its protecting packaging.

Main componentsSiPM sensor - Component

Above we can see the scintillator and SiPM sensor, plus a small tub of silicone optical coupling compound that will be used at the interface between the two. The scintillator and silicone compound were both obtained from an eBay vendor who was conveniently offering material cut to size and small quantities of the compound, specifically for constructors of the CosmicWatch detector.

Enclosure to house CosmicWatch

So now that we have the physics end of our detector, what else do we need? Well, an enclosure of some sort would be advisable and the CosmicWatch design specifies one from a particular US manufacturer, which unfortunately we were unable to source here in the UK. So instead we settled on an RS Pro one (195-1545) of a similar design and slightly larger, but we can laser cut a mounting plate for and secure the main PCB to this, following which slide the plate into the case PCB guides.

CosmicWatch PCBs

Along with the main PCB we have one for the SiPM sensor and scintillator, and a small board for a Micro SD card socket. The sensor assembly will first need to be wrapped in reflective foil and then black tape to block out ambient light, before being attached to the mainboard. The mainboard integrates a power supply for the sensor, plus amplifier and peak detector circuits for its output.

Arduino Nano Board

The conditioned SiPM sensor signal is then fed into an ADC input on an Arduino Nano (696-1667) , which can be used to send data over USB to an attached computer or record it to a Micro SD card.

I2C display board for Arduino Nano

In addition, to which, the Arduino Nano can also output statistics to an attached I2C display. The CosmicWatch design specifies a generic 0.96” 128x64 OLED part, noting that you need to be careful to ensure that you obtain the variant with the correct pin header ordering. Given that this sort of thing can be a bit hit and miss, particularly where you may only have a stock photo to go on, we decided instead to try using a similar display from Seeed Studio (174-3239) . Since we’re using a slightly longer enclosure this will need to be cabled to the main PCB in any case.

Next steps

At this point, we have a rough understanding of how the cosmic ray detector will work and as we progress with the build in the next post, we’ll explore this in a little more detail. We’ll also take a look at processing the signal and what can be done to distinguish background radiation.

Andrew Back

Open source (hardware and software!) advocate, Treasurer and Director of the Free and Open Source Silicon Foundation, organiser of Wuthering Bytes technology festival and founder of the Open Source Hardware User Group.

Recommended Articles


May 9, 2021 15:09

Hey Andrew, just to let you know that UK freight issue is a 2 way street, I have no idea why but here in the U.S. getting anything from the UK is very costly as I'm sure it would for anyone attempting to have something shipped from here to there. I have been looking at a lot of stuff online lately, and I find that it is also costly to have anything shipped from Canada to the US. It sucks too. Items can often be shipped from China to the US for free which makes very little sense. I have found some items I am in need of in the Russian Federation and freight from there is not quite as bad as from the UK or vice versa. I would assume it is something to do with the production costs along with free trade agreements which is a shaky issue with China and the region where sanctions are an issue. Again it makes little sense when you consider the distances, China to the US vs. US to UK or any other location within the Euro/Russian region then there is the distance from the US to Canada, go figure? It all comes down to how bad you want something and where it can be obtained, next is delivery time, I find myself waiting for as much as a month for many items, simply to save a dollar and much of the time, I have no option. This comment is also intended for Tony and anyone else addressing the same situation. This project looks like a lot of fun and might even be something a person could make a bit of income on to me. I have the opinion if people out there in the rest of the world knew about stuff like this, they would buy it just for the novelty of having it. After all I see people spending much money on far less useful stuff. If it turns out that way, see about getting enough parts to make this and any other neat stuff in more than just 1 for yourself and put them together then see if you can sell them locally at some shop. Many times things like this can be shipped for little on no extra fee when you order 1 or 6 or whatever.

0 Votes

April 19, 2021 09:53

Hi Andrew
Is anybody sourcing the scintillator and SiPM in the UK? As I found with the equivalent project that CERN are doing, these components are expensive in one-offs. I looked at the eBay listing for the scintillator and there is a considerable postage charge from the US.

0 Votes

April 19, 2021 11:47

@Leicester Hackspace You can buy the SiPM from RS: I'm not sure about UK scintillator sources.

April 19, 2021 12:46

@Andrew Back Thanks for the link - I'd forgotten how expensive these specialist SiPM's are. I tried to get a sample for free at an exhibition, but didn't succeed ;-)

April 20, 2021 09:00

@Leicester Hackspace NP! Indeed, not cheap, but a great deal cheaper than a PM tube and particularly when you factor in the HV power supply that they need. And I'm guessing that given their performance, they're not the simplest device to reliably fabricate.

February 16, 2021 08:12

Well done Andrew. Great project, reminds me of the old Scientific American, 'amateur scientist' projects, e.g. build your own Wilson Cloud Chamber.
What is the scintillator material used, and where to get it?

0 Votes

March 10, 2021 15:36

@kiwironnie The material is Bicron BC408 and the same eBay vendor also sells the optical coupling compound:

DesignSpark Electrical Logolinkedin