Employees of NRNU MEPhI, students, and schoolchildren are working on the data acquisition system of the detector for studying auroras

The instrument is to operate as part of the CubeSat 3U space platform from Geoscan, that is, under very strict constraints on dimensions and power consumption. Nevertheless, excellent values have already been achieved for the key parameter—the instrument’s energy resolution.

How the CsI Crystal Works

The heart of the detector is a scintillation crystal of cesium iodide activated with thallium (CsI(Tl)). Its operating principle is based on converting the energy of ionizing radiation into visible light.

When a gamma quantum or charged particle enters the crystal, it transfers its energy to the lattice electrons. The crystal atoms become excited and then, returning to the ground state, emit photons—a flash of scintillation light. The intensity of this flash is proportional to the absorbed energy, which makes it possible to reconstruct the energy of the radiation incident on the crystal.

A cesium iodide-based crystal is one of the most efficient scintillators for detecting gamma radiation and charged particles. Its light yield is about 65,000 photons per MeV, which provides a comparatively bright flash and good energy resolution. The crystal is weakly sensitive to humidity and has a decay time of about 700–900 nanoseconds. The latter is important for instruments operating under high load conditions: too long a flash would limit the count rate and increase dead time, which would inevitably affect statistics.

The light from the crystal is collected by silicon photomultipliers (SiPMs), which convert photons into an electrical signal. The key task of the electronics is to register this signal with minimal losses and noise, since the number of collected photons directly determines the accuracy of energy measurement. However, CsI(Tl) has a feature that determines the complexity of the design: light propagates in the crystal in such a way that a significant portion of the photons reflects from the side faces and exits through the face that is optically coupled to the photodetector. If the SiPMs are poorly positioned, some of the light is lost, and energy resolution drops.

Choosing the Crystal Geometry

This is precisely why the project team considered so many options for SiPM placement. It is necessary to collect the maximum amount of light from the crystal with the minimum number of photodetectors, since each channel requires its own electronics, and space and power consumption in a CubeSat are limited.
Figure 2 shows the SiPM configurations studied: from a simple 2×2 grid to complex schemes with 12–16 elements. Each configuration is a compromise between light collection efficiency, uniformity of response over the crystal volume, and electronic complexity. Options with SiPMs at the edges or in a cross shape can provide better light collection with fewer channels, but require more complex signal processing. The 2×2 configuration is simpler but collects less light.

Electronics

Once the geometry was approved, a new difficulty arose: SiPMs require precise control. The flash of light from an absorbed electron in the crystal lasts nanoseconds, and capturing this signal without losses is not easy.

A custom analog data acquisition board was developed to solve several tasks:

• stabilizing the SiPM supply voltage: temperature fluctuations in orbit shift the breakdown voltage and distort energy measurement;
• low-noise amplification and summing of the signal so that the useful signal is not lost in electronic noise.

And, of course, it was necessary to quickly receive signals from the front-end electronics and process them. An FPGA proved suitable for this role—it is the computing core of the acquisition board, turning weak and fast SiPM signals into scientific data.

What does the FPGA do on the acquisition board?

First, digitization. The SiPM signal is an analog pulse: small in amplitude and very fast in rise time (nanoseconds). The FPGA controls the ADC and “captures” the amplitude and time of each pulse.

Second, real-time selection. The detector can register many events per second, but not all are useful. The FPGA checks each pulse directly in the data stream and decides whether it should be recorded. For a CubeSat this is fundamental: memory capacity and the transmission channel to Earth are limited, and it is this selection that determines which data reach the researchers.

The board was designed from scratch, tested, and refined by laboratory staff with the involvement of young participants, both students and schoolchildren, who for the first time had to become acquainted with such complex electronic devices.

The assembled detector prototype with the new electronics and verified geometry underwent calibration tests. The key parameter to be measured was the energy resolution of the full absorption peak.

The full absorption peak of Cs-137 was approximated by a Gaussian function; its resolution (FWHM) is found as FWHM = w * 1.177 / xc = 1.177 * 55.4 / 661.77 = 0.0985 (9.85%), where xc is the energy of the center of the full absorption peak, w is twice the standard deviation of the approximating Gaussian function.

For electron and soft gamma-ray detectors based on small crystals operating as part of microsatellites, a typical result in world practice is considered to be a resolution of 10–15%. During testing, spectra were accumulated; after processing, the resolution reached a value better than 10%. This is a very good result, which will make it possible not merely to record the presence of electrons but to determine their energy, which is important for the various scientific tasks to be addressed within the mission.