A Wide Bandwidth Trans-impedance Amplifier for Picosecond-Scale SiPM Characterization in a Wide Temperature Range

TL;DR

This work addresses the challenge of accurately characterizing Silicon PhotoMultipliers (SiPMs) across a wide temperature range, including cryogenic operation at $80\ \mathrm{K}$, by designing a wide-band, low-noise transimpedance amplifier that preserves single-photon timing. The authors compare two configurations, Opamp Dominant Pole (ODP) and Transistor Dominant Pole (TDP), deriving and validating open-loop and loop-gain models, stability criteria, and closed-loop transfer functions, and then selecting a cryogenically robust implementation. They demonstrate a two-stage design (HBT front-end with a CFOA back-end) achieving high gain ($\sim 7500\ \mathrm{V/A}$), sub-ns rise times ($<500\ \mathrm{ps}$), and sub-pA/√Hz input noise, with measurable output jitter as low as a few picoseconds, even when operating SiPMs at low over-voltage. Performance is validated through extensive simulations, board-level tests, and real SiPM measurements at ambient and cryogenic temperatures, showing that the amplifier can faithfully reproduce SiPM signals with minimal noise and timing jitter and enabling accurate time-of-arrival, gain, and recovery-time measurements in challenging environments. The resulting system offers a practical, near state-of-the-art solution for high-precision SiPM readout in radiation-rich, cryogenic detectors used in high-energy physics.

Abstract

Future high-energy physics experiments using SiPMs as photosensitive elements may require operation at low temperatures (down to 80 K) to measure single photons with high time resolution in a highly radioactive environment. This calls for a complete characterization of these sensors over a wide temperature range to find the best compromise between detector performance and cooling requirements. This paper presents the design of a transimpedance amplifier featuring high gain ($\sim 7500$ $\mathrm{V/A}$), very high speed ($ < 500$ $\mathrm{ps}$ rise time) and low input noise ($\lesssim 0.2$ $\mathrm{pA/\sqrt{Hz}}$), able to faithfully reproduce all the features of SiPM signals with very low noise and time jitter. These features make the amplifier suitable for precise measurements of the time-of-arrival of single-photon signals, as well as gain and recovery time. This article provides a detailed and thorough analysis of the circuit. The network was simulated and measured in two configurations that differ in their open-loop gain and dominant pole frequencies. After selecting the best configuration for our purposes, the amplifier was characterized in detail at ambient temperature and at 80 K. Finally, we evaluated the amplifier using a SiPM operated at low over-voltage. While SiPMs are typically characterized at high over-voltage to enhance gain and minimize timing jitter, testing at low over-voltage allowed us to assess the amplifier's performance under more challenging and realistic conditions for single-photon timing.

Figures (19)

• Figure 1: Schematic of the amplifier circuit. The gray components represent the parasitic capacitances.
• Figure 2: Focus on the SiPM structure, with N the number of SiPM cells, $R_q$ the quenching resistor, $C_d$ the single SiPM cell capacitor, $i_{1cell}$ the current signal of one fired cell and $C_g$ the grid parasitic capacitance. Outside of the SiPM structure, we have $R_{HV}$ and $C_{HV}$ which are used for the SiPM bias voltage filtering. $R_{s1}$ and $R_{s2}$ can be used for line termination or amplifier compensation. $Z_{in}$ is the amplifier input impedance and $i_{in}$ is the current flowing into the amplifying system.
• Figure 3: Schematic of the amplifier configurations with all parasitics included. The orange dashed box contains the impedances denoted by $Z_i$, while the blue dashed box highlights those denoted by $Z_b'$.
• Figure 4: Amplifier layout. It features four amplifiers around a SiPM test structure (white square at the center). There is one PT1000 temperature sensor coupled to the PCB ground pad for more precise temperature measurements. In this picture only one amplifier and one SiPM socket are mounted.
• Figure 5: Test signal path from the signal generator to the "Test signal X" input for the four different amplifying channels.