Multi Channel Analyzer for Gamma Spectroscopy With Arduino & Theremino




Introduction: Multi Channel Analyzer for Gamma Spectroscopy With Arduino & Theremino


I'd like to show you my homemade multi-channel-analyzer for gamma spectroscopy using a.) the freeware theremino and b.) an arduino.

First of all I have to say some words about gamma spectroscopy at all. To detect radioactive decays you need a detector like a geiger-Counter. For gamma-spectroscopy you have to take a photomultiplier combined with a scintillation crystal made of sodium Iodide. You can get both things on ebay for less than 100 USD.

The gamma-rays from the radioactive source is going through the scintillation crystal and produce very faint light-flashes. Those light-flashes are recorded by the photomultiplier, who converts them is small voltage-pulses at the Output. For this the photomultiplier and the Crystal have to be in a absolutely light-tight housing. I use shrinkable tubing for that.

The Job of the multi-channel-analyzer (MCA) is to measure the height of those pulses coming from the photomultiplier. This is because every gamma-photon produces a voltage-pulse with a height depending on it's energy. The higher the energy of the gamma-photon, the higher the voltage pulse.

When the pulse-height has been measured, the number pulses with this recorded amplitude can be increased by one. At the end you get a spectrum of number of pulses versus pulse-height, which is characteristic for each radioactive element. For example cesium-137 emits gama-rays with an energy of 662 keV. They will produce pulses with a certain pulse-height. With a MCA you'll get a spectrum with a peak at a certain voltage corresponding to the pulse-height. Therefore you can identify radioactive sources with a MCA.

You don't get just the so called photo-peak, but also characteristic structures like the compton-edge, which is produced by the scattered electrons.

Step 1: High Voltage Power Supply

For the photomultiplier you'll need a high-voltage power supply. Most of the PMT (photomultiplier-Tubes) Need voltages between 800 V and 1500 V. I use a negative voltage, because then you'll get the signal from the PMT in an easier way.

There are a lot of different methods producing the high-voltage. In my case I use a cheap CCFL-Inverter combined with a simple electronic stabilization.

Step 2: The MCA Freeware Theremino

With the MCA-Freeware theremino ( you're able to record gamma-spectra in a very easy way. You'll just need the detector (photomultiplier + scintillation-crystal), the high voltage power supply an a cheap USB-Sound Card (

Step 3: The DIY-MCA With Arduino

Though the Freeware Theremino is a nice and well functioning Software I tried to realize an MCA with an arduino. The principle is quite simple:

First the storage capacitor of the Peak-detector is discharged. Then I close the Switch of the Peak-detector and wait until I get a pulse from the monoflop. Then I open the Switch and read in the voltage of the capacitor, which is equal to the maximum voltage of the incoming pulse. After this I refresh the graphics by adding one at the matching column.

To see whether the reading of the peak-voltage is fast enough, I inserted a test-pulse immediately after a variable delay and after I've read the voltage of the storage-capacitor. This confirmed my guess that without a delay I really measure the Peak-voltage and the capacitor isn't noticeably unloaded.

To make another test I fed my MCA with pulses with just one and not varying amplitude. You can see the result in the picture. I got a real spectrum with just one line as it should be.

Step 4: The Peak Detector

For the MCA you need a circuit, which stores the height of an incoming pulse. I use a very simply method with a transistor, a diode and a capacitor. In the pictures you can see the pulses from the photomultiplier and the voltage of the peak-detector. The voltage of the peak-detector gets exponentially lower with time because the capacitor looses his charge over the resistor.

Step 5: The Complete Device

The complete device consists of the following In- and outputs:

  • signal In: Here the signals from the photomultiplier are coming in
  • gain: to vary the signal-heights
  • offset: to vary the offset of the pulses
  • signal out: to check the Signal with an oscilloscope
  • comparator-level: to vary the level from which a signal is sent to the monoflop
  • to counter: the significantly longer signal from the monoflop for external counters or something else

Step 6: Results

Here are some spectra taken with a Thorium-mantle (), an autunite and a cesium-spark-gap-tube. You can clearly see the photopeak at 662 keV...

Maybe you'd like to take a look at my YouTube-channel:

more physics projects:

Explore Science Contest 2017

Participated in the
Explore Science Contest 2017

Be the First to Share


    • Make It Bridge

      Make It Bridge
    • Game Design: Student Design Challenge

      Game Design: Student Design Challenge
    • For the Home Contest

      For the Home Contest



    1 year ago on Step 1

    how many amps is the power supply? I am assuming it must be in microamps? I noticed it must be very very low current . Please advise...


    Reply 1 year ago

    You are right. The photomultiplier needs less than 1 mA because of the high resistance of the voltage divider for the dynodes. But the power supply is able to deliver higher currents.
    Cheers stoppi


    Reply 1 year ago

    Thanks for the quick reply!
    I was experimenting and wanted to see if the power output would light a flashlamp or fluorescent lamp but it didn't. The only lamp that would illuminate was a neon lamp. when i adjust the pot i can see the the neon illumination increase or decrease so i noticed it must be working but at a very low amps so I am assuming it must be working. Wanted to know before I hook it up the circuit.


    5 years ago

    Thanks for sharing! Never thought of it - such a simple and elegant way to determine radiation source! I have an old USSR-times scintillator laying around, that approach will make a great use of it :)
    What sensitivity can I expect from it? At how many Bq signal is strong enough to reliably determine source type?