Compare PCB Prices:
|... 13 more lines|
This project is on Hackaday.io, where I also post project updates and other announcements!
Open hardware for a hackable scintillation counter and multichannel analyzer (MCA) all-in-one device using a popular NaI(Tl) scintillation crystal and silicon photomultipliers (SiPMs). Suitable for (DIY) gamma spectroscopy while being significantly cheaper than any off-the-shelf commercial platform. Uses a standard serial-over-USB connection so that it can be integrated into as many other projects as possible, for example data logging with a Raspberry Pi, weather stations, Arduino projects, etc.
Hardware design has been done with EasyEDA and all the needed files for you to import the project as well as the schematic can be found in the
hardware folder. There is also a Gerber file available for you to go directly to the PCB manufacturing step.
The software aims to be as simple as possible to understand and maintain; to achieve this I decided to use an off-the-shelf microcontroller - the Raspberry Pi Pico. This board can be programmed with the Arduino IDE over micro-USB and is powerful (dual core, good ADC, plenty of memory, ...) enough for the purpose and also exceptionally cheap.
Here are some of the most important key facts:
- Compact design: Total size 120 x 50 mm. 70 x 50 mm area for electronics and additional 50 x 50 mm to mount a scintillator.
- All-in-one detector: No external parts (e.g. sound card) required to record gamma spectra.
- Easily programmable using the standard Arduino IDE or drag-and-drop firmware files.
- Low-voltage device: No HV needed for a photomultiplier tube.
- Can use SiPMs in the voltage range of 28 V to 33 V.
- Simple OLED support out of the box.
- Low power consumption: ~15 mA @ 5V in standard operation.
- Adjustable preamp gain for the SiPM pulses (affects energy range & resolution).
- Customizable energy range with 4096 ADC channels.
- Default Mode: Capable of at least around 60,000 cps while also measuring energy.
- Geiger Mode: Capable of at least around 180,000 cps without energy measurement.
- Additional broken-out power pins and I2C header for displays, etc.
- True Random Number Generator
For quick access and purchase of all the parts (PCB and BOM), you can use Kitspace. Otherwise, use a PCB manufacturer and an electronics distributor of your choice and proceed on your own.
In both cases you will also need to buy a SiPM (e.g. the MICROFC-60035-SMT-TR1) and scintillator (NaI(Tl) recommended) at a distributor of your choice.
This project utilizes a silicon photomultiplier (short SiPM) which is way smaller and more robust than a traditional photomultiplier tube and does not need a high-voltage supply (traditionally ~1000 V). Here are some very helpful in-depth datasheets about the particular MicroFC SiPM recommended here:
- C-Series SiPM Sensors datasheet
- Linearity of the Silicon Photomultiplier
- Introduction to the SiliconPhotomultiplier (SiPM)
- Biasing and Readout of ON Semiconductor SiPM Sensors
The hardware consists of the main detector (
hardware folder) which includes amplification, pulse detection and energy measurement. If you already have a SiPM/crystal assembly compatible with voltages around 30 V, you may use it with the detector board and connecting wires directly to the correct pads. Otherwise, you can use my SiPM carrier board, which holds the SiPM and all the optional decoupling.
The heart of the detector board is the Raspberry Pi Pico which uses its ADC to measure the pulse amplitude (i.e. the energy) immediately after an event occurs starting with an interrupt. I can really recommend you reading the datasheet or maybe also having a look at a deeper analysis of the Pico ADC, if you're interested:
Here are some front and back side renders of the detector PCB. Size is about 12 x 5 cm. If you don't need the additional space to mechanically mount the SiPM/scintillator assembly to the rest of the detector board, you can just cut it off at the white line and you're left with an even smaller detector.
On the back side of the PCB there is place for two optional components:
- a voltage reference for the ADC (LM4040AIM3-3.0+T recommended) to get rid of most power supply related noise and inaccuracy
- and a 0 Ω resistor link to connect the analog ground to the rest of the ground plane.
These can be retrofitted easily and are quite affordable. The voltage reference is highly recommended, because the Pico DC/DC converter isn't the most accurate. Do not solder the 0 Ω resistor, though, except if you know what you're doing.
There are also broken-out pins for the power supply and I2C connections. These can be used to modify the device, e.g. by adding a display or using a battery charger. You can have a look at the great Raspberry Pi Pico datasheet for more info on this.
Here is a helpful image about the potentiometer settings for Rev2.0:
The finished MicroFC- and AFBR- SiPM carrier boards are there to allow for easier packaging with the scintillator as well as to be reusable for different detectors as that's by far the single most expensive part and you'll want to use it as long as possible. You should apply some optical coupling compound between the SiPM and the crystal window to reduce reflections as good as possible (this way the best photon detection efficiency is achieved). There are also special materials for this use case but you can also use standard silicone grease - works great for me. After you applied some, you press both parts together and wrap everything with light-tight tape, again, I'm just using some black electrical tape here. That's essentially it, now you can solder some wires to the pads on the board to connect them together and secure it in place in the free space on the board.
I got all of my scintillators (used NaI(Tl), LYSO, ...) on eBay. Just search for some keywords or specific types, you'll probably find something! Otherwise you can obviously also buy brand-new scintillators, however, these are much more expensive (depends, but a factor of 10x is normal). Just be sure to look out for signs of wear and tear like scratches on the window or yellowing (!) in NaI crystals as these can deteriorate performance significantly.
More assembly instructions can be found on the Hackaday.io project page!
Due to the detector measuring small voltages, energy resolution being limited by noise and a tiny 220 pF capacitor being on board, it is generally pretty sensitive to EMI. In fact, without any shielding and periodically discharging the capacitor, mains interference would slowly charge it until the device gets overwhelmed with noise. To mitigate this effect, the Firmware is programmed to clear the cap every 500 µs by default, which is enough to mostly fix this issue. However, this adds an additional ~4 ms dead time total per second which could be roughly equivalent to 500 missed events in geiger mode (worst case).
For the best performance, you will need to put your detector into a metal enclosure. It doesn't need to be a thick metal case, a tin can will most likely suffice.
You can get a 3D-printable case for the Open Gamma Detector with different styles of covers.
All the STL files to print the two parts (main body and cover) can be found in /enclosure, as well as some more info on the necessary screws and the USB extension cable.
Programming is done using the Arduino IDE. The so-called "sketch" (i.e. the programmed software) can be found in
For convenience there is also a ready-to-use default firmware UF2 located in
/software/, which is targeted at the standard Open Gamma Detector that you can build using the instructions or buy on the makerfabs store.
This is the easiest way to get started! You'll not have to download anything else besides the firmware UF2 file. This is the latest default firmware that you can use if you don't plan to modify any of the hardware aspects (such as the voltage reference, ADC resolution, etc) on the board itself. You can of course also use it with the Open Gamma Detector purchased from the makerfabs store.
To get started with a fresh device, plug in the Raspberry Pi Pico via the micro-USB connection. A file manager window should now pop up. Drag and drop the
opengamma_pico-XXX.uf2 file that you downloaded into this directory. The device will restart automatically once the transmission is completed and is ready to use!
If you want to update or re-flash the firmware, press and hold the
BOOTSEL button on the Raspberry Pi Pico while plugging it in the USB port of your computer. A file manager windows should once again pop up and you can let go of the button. Drag and drop the UF2 file just like above.
To program the Pico you will need the following board configs in the latest release of the Arduino IDE:
The installation and additional documentation can be found in the respective GitHub repo, it's not complicated at all and you only need to do it once. You will also need the following additional libraries:
They can be installed by searching their names using the IDE's built-in library manager.
Please have a look at the
USER SETTINGS in the Arduino sketch. The most important setting here is the
VREF_VOLTAGE. If you soldered in the voltage reference then this probably needs to be set to
3.0, otherwise leave
3.3 as is.
Flash the Pico by choosing the
Raspberry Pi Pico under
Tools/Board/Raspberry Pi RP2040 Boards and then selecting
Flash Size: 2MB (Sketch: 1984KB, FS: 64KB), leaving everything else at the default value. You can then press the big
You can control your spectrometer using the serial interface. The following commands are available, type
help to get a list of commands. Commands are case sensitive. Additional parameters needed for each command are surrounded by
< ... >.
> helpUsage:<command> [options]Commands:read spectrum : Read the spectrum histogram collected since the last reset.read settings : Read the current settings (file).read info : Read misc info about the firmware and state of the device.read fs : Read misc info about the used filesystem.set trng : <toggle> Either 'enable' or 'disable' to enable/disable the true random number generator output.set display : <toggle> Either 'enable' or 'disable' to enable or force disable OLED support.set mode : <mode> Either 'geiger' or 'energy' to disable or enable energy measurements. Geiger mode only counts cps, but has ~3x higher saturation.set int : <mode> Either 'events', 'spectrum' or 'disable'. 'events' prints events as they arrive, 'spectrum' prints the accumulated histogram.set reset : <toggle> Either 'enable' or 'disable' for periodic resets of the P&H circuit. Helps with mains interference to the cap, but adds ~4 ms dead time.set averaging : <number> Number of ADC averages for each energy measurement. Takes ints, minimum is 1.set delay : <number> Delay between trigger and ADC readout of pulses in µs. Set this to ~1/2 of the maximum pulse duration you are expecting. Minimum is 1.clear spectrum : Clear the on-board spectrum hist.clear settings : Clear all the settings and revert them back to default values.reboot : Reboot the device.>
The detector board features a standard I2C header where you can connect any standard SSD1306 OLED display that will be supported with minimal changes to the Arduino IDE sketch. A 128 x 64 px OLED is supported as is and ~will be automatically used upon boot~ [UPDATE: See https://github.com/OpenGammaProject/Open-Gamma-Detector/issues/19] of the device in the current firmware version! If no display is connected the device will work as usual only via a Serial interface. You can also force the device not to use the display even if one is connected via the I2C header.
At the moment the software only draws the overall energy spectrum and the overall mean cps value on the screen. This is sufficient in most (simple) cases and more features will be implemented over time.
Radioactive decay is a great source of entropy for a random number generator of course. This device can output true random numbers meaning they are truly, per definition, random, instead of the pseudorandom numbers a PRNG (i.e., an algorithm) can produce. If you enable the TRNG via the serial command, it will periodically (once they are ready) output random numbers between 0 - 255. This could be further used to, for example, periodically re-seed a pseudorandom number generator to increase the overall amount of available numbers to use.
Note that the TRNG uses the same Serial connection as the gamma spectroscopy output, so you have to disable one of them so that you don't mix the outputs. The numbers of both functions are formatted the same for ease of use, i.e. with a semicolon ";" as some kind of "end of data" delimiter.
Do not use this TRNG for real-life applications or sensitive data. There is always some risk that the implementation is not truly random and some numbers correlate.
To get the data from the detector the serial-over-USB port is used by default. The quickest and easiest way to do this is by using my own web application called Gamma MCA where you can connect straight to the serial port and plot the data live as well as import and export finished spectrum files. You don't even need to install it, it can work out of any Chrome-based browser! Please head to the repository to find more specific info about this project.
You can of course use any other serial monitor or gamma-spectroscopy software that's compatible with serial connections. To my knowledge there isn't much, though, that's why I made one myself.
Here is a small collection of example spectra I could make quickly without putting much effort into the detector settings (gain, threshold, SiPM voltage). In addition, neither the electronics nor the scintillator and sample were shielded whatsoever.
Two hour long background spectrum with no samples:
Spectrum of a tiny (~5 g) LYSO scintillator inside some lead shielding showing all three distinct gamma peaks (88.34, 201.83, 306.78 keV) with an additional ~55 keV X-ray peak (2h measurement):
Spectrum of a standard household ionization smoke detector. Contains roughly 0.9 µCi of Am-241. Gamma peaks at 26.34 and 59.54 keV:
Spectrum of a small tea cup with pure Uraninite (Pitchblende) contents in its glaze. You can see all kinds of isotopes of the Uranium series:
Spectrum of a Na-22 test source, about ~2 µCi. You can see the 511 keV annihilation peak and the small 1275 keV gamma peak (much smaller due to limited NaI efficiency and small scintillator size for this energy):
The Raspberry Pi Pico's ADC has some pretty severe DNL issues that result in four channels being much more sensitive (wider input range) than the rest. For now the simplest solution was to discard all four of them, by printing a
0when any of them comes up in the measurement (to not affect the cps readings). This is by no means perfect or ideal, but it works for now until this gets fixed in a later hardware revision of the RP2040.
It's very important to get the SiPM/scintillator assembly light-tight. Otherwise you'll either run into problems with lower energies where noise dominates or outright not measure anything at all, because the sensor is saturating.
Due to the global parts shortage many chips are much harder to come by, if at all that is. This is most severe for high-quality op amps and other specialized parts such as the DC/DC converters and the tiny peak detector cap. This makes it much harder for me to choose from components and maybe even limiting the performance. Parts that are listed in the BOM should be available easily and with high reliability and stock so that they don't run out quickly. Notably, this also drives the cost of the detector up by a large factor, again mostly due to the op amps and also the SiPMs.
Please have a look at REFERENCE.md.
Using multiple detector boards with updated firmware should be able to deliver some coincidence measurement features. By respectively connecting the
GND and one of the
SDA pins to each other on both boards you have everything you need to get started. The
SDA pin will be used for an interrupt from the child detector to the parent to trigger a pulse if both timings coincide. Since you lose the ability to use I2C with this, future detector boards will likely feature an entire new coincidence pin header.
At the moment, though, I couldn't get a coincidence mode feature running due to misc timing issues. There might be a firmware update in the future to implement this feature.
During operation all the electronics including the photomultiplier naturally heat up ever so slightly. Due to the detector board being connected to the SiPM only by a single pin connector and/or by touching the PCB, its heat shouldn't affect the SiPM much if at all. Air or water cooling alone won't improve performance considerably, because it won't heat up much above ambient temps. However, you could cool the SiPM PCB with a Peltier module to sub-ambient temperatures. According to the datasheet AND9770 (Figure 27) every 10°C reduction in temperature decreases the dark count rate by 50%! But be sure to correct the SiPM voltage (overvoltage) in this case as it also changes with temperature.
Note that the required breakdown voltage of the SiPM increases linearly with 21.5 mV/°C, see the C-Series SiPM Sensors Datasheet. This means that you would also need to temperature correct the PSU voltage if you wanted to use it with considerably different temperatures.
Shielding the ambient background can be done ideally using a wide enough layer of lead (bricks) all around the detector with a thin layer of lower-Z material on the inside (to avoid backscattering) such as copper. The SiPM and the sample can then be put into the structure to get the best measurements possible (low background).
See Wikipedia: Lead Castle
Thanks for reading.