Search

The paper presents an implementation of fully differential readout method for Silicon Photomultipliers (SiPM). Front-end electronics consists of a fast and slow path. The former creates the trigger signal while the latter produces a pulse of width proportional to the input charge. The fast shaper generates unipolar pulse and utilizes the pole-zero cancelation circuit. The peaking time for single photoelectron is equal to 3.6 ns and the FWHM is 3.8 ns. The pulse width of the Charge to Time Converter (QTC) depends on the number of photons entering the SiPM at the moment of measurement. The QTC response is nonlinear but it allows us to work with signals in a wide dynamic range. The proposed readout method is effective in measurements of random signals where frequent events tend to pile-up. Thermal generation and afterpulses have a strong influence on the width of pulses from QTC. The proposed method enables us to distinguish those overlapping signals and get the reliable information on the number of detected photons. [ABSTRACT FROM AUTHOR]

Proposed algorithm compensates the gain by changing the bias voltage of Silicon Photomultipliers (SiPM). The signal from SiPM is amplified in fully differential preamplifier then is formed in time by the fully differential fast shaper. The compensation method was tested with four channels common cathode multi-pixel photon counter from Hamamatsu. The measurement system requires only one high voltage power supply. The polarization voltage is adjusted individually in each channel indirectly by tuning the output common mode voltage (VOCM) of fully differential amplifier. The changes of VOCM affect the input voltage through the feedback network. Actual gain of the SiPM is calculated by measuring the mean amplitude of the signal resulting from detection of single photoelectron. The VOCM is adjusted by DAC so as to reach the desired value of gain by each channel individually. The advantage of the algorithm is the possibility to set the bias of each SiPM in the array independently so they all could operate in very similar conditions (have similar gain and dark count rate). The algorithm can compensate the variations of gain of SiPM by using thermally generated pulses. There is no need to use additional current to voltage conversion which could introduce extra noises. [ABSTRACT FROM AUTHOR]