Test Beam Characterization of a Digital Silicon Photomultiplier
Finn King, Inge Diehl, Ono Feyens, Ingrid-Maria Gregor, Karsten Hansen, Stephan Lachnit, Frauke Poblotzki, Daniil Rastorguev, Simon Spannagel, Tomas Vanat, Gianpiero Vignola · 2024
Read Test Beam Characterization of a Digital Silicon Photomultiplier by Finn King, Inge Diehl, Ono Feyens, Ingrid-Maria Gregor, Karsten Hansen, Stephan Lachnit, Frauke Poblotzki, Daniil Rastorguev, Simon Spannagel, Tomas Vanat, Gianpiero Vignola (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-06687.txt.
Introduction
Finn King and co-authors characterized a digital silicon photomultiplier in an electron test beam, in a 2024 paper. These light sensors detect single photons and serve physics, medicine, and commerce. Making their avalanche diodes in standard chip processes lets a sensor grid and its circuit share one die, a digital design whose potential is still being mapped.
The prototype, built in a 150-nanometer process, holds a grid of 32 by 32 pixels. Each pixel packs four diodes and a digital front end, and the chip adds timing converters and data links. At a test beam facility, the team measured how well it catches passing particles, to judge it as a tracking and timing detector.
Catalog record
| CREATOR | Finn King, Inge Diehl, Ono Feyens, Ingrid-Maria Gregor, Karsten Hansen, Stephan Lachnit, Frauke Poblotzki, Daniil Rastorguev, Simon Spannagel, Tomas Vanat, Gianpiero Vignola |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-06687
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Test Beam Characterization of a Digital Silicon Photomultiplier Finn King, Inge Diehl, Ono Feyens, Ingrid-Maria Gregor, Karsten Hansen, Stephan Lachnit, Frauke Poblotzki, Daniil Rastorguev, Simon Spannagel, Tomas Vanat, Gianpiero Vignola Abstract Conventional silicon photomultipliers (SiPMs) are well established as light detectors with single-photon-detection capability and used throughout high energy physics, medical, and commercial applications. The possibility to produce single photon avalanche diodes (SPADs) in commercial CMOS processes creates the opportunity to combine a matrix of SPADs and an application-specific integrated circuit in the same die. The potential of such digital SiPMs (dSiPMs) is still being explored, while it already is an established technology in certain applications, like light detection and ranging (LiDAR). A prototype dSiPM, produced in the LFoundry 150-nm CMOS technology, was designed and tested at DESY. The dSiPM central part is a matrix of 32 by 32 pixels. Each pixel contains four SPADs, a digital front-end, and has an area of 69.6 × 76 µ m 2 69.6\times$76\text{\,}{\mathrm{\SIUnitSymbolMicro m}}^{2}$ . The chip has four time-to-digital converters and includes further circuitry for data serialization and data links. This work focuses on the characterization of the prototype in an electron beam at the DESY II Test Beam facility, to study its capability as a tracking and timing detector for minimum ionizing particles (MIPs). The MIP detection efficiency is found to be dominated by the fill factor and on the order of 31 % 31\text{\,}\mathrm{\%} . The position of the impinging MIPs can be measured with a precision of about 20 µ m 20\text{\,}\mathrm{\SIUnitSymbolMicro m} , and the time of the interaction can be measured with a precision better than 50 ps 50\text{\,}\mathrm{ps} for about 85 % 85\text{\,}\mathrm{\%} of the detected events. In addition, laboratory studies on the breakdown voltage, dark count rate, and crosstalk probability, as well as the experimental methods required for the characterization of such a sensor type in a particle beam are presented. Journal: Nucl. Instrum. Methods Phys. Res. A 1 Introduction Silicon photomultipliers (SiPMs) contain an array of single-photon avalanche diodes (SPADs). SPADs are operated in Geiger-mode and sensitive to ionizing radiation and even single photons. One key feature of the Geiger-mode operation are fast-rising signal edges, facilitating single photon time resolutions on the order of 10\text{\,}\mathrm{ps} [ 1 ] . Structure, operation principle, and applications of SiPMs are discussed, for example, in [ 2 , 3 , 4 ] . An effect of the Geiger-mode operation is that the SPAD signal contains no information about the energy deposition by the interacting quanta. This inherent digital nature of the SPAD signal makes thinking of a SiPM as a digital device seem natural. As SPAD layouts in CMOS processes are available, CMOS circuitry can be embedded in the SiPM design, which allows profiting from digital signal processing. Introduced as digital SiPMs (dSiPMs) in [ 5 ] , these devices come with several advantages compared to analog SiPMs. The circuitry allows storing information about the position of a firing SPAD (order of 10\text{\,}\mathrm{\SIUnitSymbolMicro m} precision), the masking of noisy SPADs, and on-chip pattern recognition. Parallel connection of the whole SPAD array can be avoided, which reduces the input capacitance to the readout circuit, and thus improves signal rise times. Finally, it can ease system integration, as no separate device for signal digitization is required. One clear disadvantage is the quality of the SPADs, yielding generally higher dark count rates (DCRs) in dSiPMs [ 6 ] , which will certainly improve once more sophisticated SPAD designs become available in CMOS processes. Another is the reduction of the fill factor, as a part of the surface area is needed for the CMOS circuitry, with increasing circuit complexity. Possible applications have to find a trade-off between these advantages and disadvantages. Especially, the availability of position information is a key point to consider. One example is to use dSiPMs for the readout of scintillating-fiber bundles [ 7 ] , where the available position information is used to identify individual fibers, potentially reducing complexity and cost of fiber-readout solutions. Another application would be 4D particle tracking [ 8 ] , where the combination of position information on the order of 10\text{\,}\mathrm{\SIUnitSymbolMicro m} and time information on the order of 10\text{\,}\mathrm{ps} might be beneficial properties. With this perspective, this work reports on the minimum ionizing particle (MIP) detection capabilities found for a new dSiPM prototype designed at DESY [ 9 ] , and described in section 2 . Laboratory measurements, focusing on breakdown voltage, DCR and crosstalk, are discussed in section 3 . Setup, analysis method and results of the MIP-detection studies are presented in section 4 , 5 , and 6 , respectively. Figure 1: Microscopic image of a pixel. The four-SPAD structure with shared electronics in the lower part of the pixel is visible. The total area of the pixel is 69.6\times$76\text{\,}{\mathrm{\SIUnitSymbolMicro m}}^{2}$ . 2 The DESY dSiPM The DESY dSiPM prototype is fabricated in the LFoundry standard 150- \text{\,}\mathrm{nm} CMOS technology. The basic building blocks of this dSiPM are pixels containing four SPADs with a shared quenching circuit, discriminator and 2-bit counter. A microscopic image of the pixel is shown in figure 1 . The pixel area is 69.6\times$76\text{\,}{\mathrm{\SIUnitSymbolMicro m}}^{2}$ , where the area occupied by the SPADs amounts to 30\text{\,}\mathrm{\%} , determining the fill factor. A group of 16\times 16 pixels forms a quadrant with one 12-bit time-to-digital converter (TDC), generating a time stamp for the first hit (firing SPAD) in each quadrant and readout frame. The readout frames are defined by a 3\text{\,}\mathrm{MHz} clock, and the time and hit position information of each quadrant are buffered and read out in the following frame, if the chip is operated in 1-bit mode. In 2-bit mode, the in-pixel counter is used to store up to three hits per pixel and frame. Consequently, the data transmission takes two frames in this case. For the studies presented hereafter, the 1-bit mode is used. Individual pixels can be disabled (masked) using a transistor switch in series with the quenching transistor, which helps to reduce the DCR in the presence of noisy pixels. A temperature diode in the periphery of the chip allows for precise temperature monitoring during measurements and operation. Further features of the DESY dSiPM, like a validation logic for topology-based event selection, external test circuits, and LED trigger pulse generation, are detailed in [ 9 ] , along with a detailed report on the ASIC design and properties. 2.1 Data Acquisition System To facilitate data acquisition (DAQ) and save development time, the DESY dSiPM is integrated into the Caribou DAQ system, an open-source hardware, software (Peary) and firmware project dedicated to pixel detector development [ 10 , 11 ] . Its key component is a system-on-chip (SoC) board (AMD [ 12 ] , Zynq 7000 SoC ZC706) containing an FPGA and a CPU. The CPU is running a fully-featured Yocto-based Linux distribution and Peary, while the FPGA executes custom written firmware blocks for configuration, control, and readout of the dSiPM. The SoC board is connected to a mezzanine board, the CaR board, that provides supplies like current and voltage sources, and a physical interface to the dSiPM. The dSiPM itself is placed on a custom chip board, connected to the