Differential-readout: The technique to optimise timing in a monolithic MPPC array

A study of the single photon time resolution (SPTR) and the coincidence time resolution (CTR) of the 4×4 MPPC array of Hamamatsu (S11827) has been performed to characterise the difference in signal response of an array compared to a single cell MPPC device. The monolithic 16-channel array has outputs from the 16 individual anodes and a common cathode. The common cathode has the disadvantage that a signal in any one of the 16 cells is coupled to the others. This is a source of cross-talk and also generates noise in the common ground affecting the timing performance of the array. The dependence of the SPTR on the applied bias voltage, the threshold and the laser intensity was first measured with a single 3×3 mm2 MPPC (10963-50P) since it is the basic element of the 4×4 MPPC array. The results show that when only a single cell of the array is connected (with all other cells ‘off’, i.e. with the anodes disconnected) the MPPC array response is much like the single-cell MPPC with an SPTR of 110 ps (sigma). We then tested the performance of the 4×4 array, incrementing the number of active cells in the array. Increasing the number of MPPC cells decreases the signal (measured by time-over-threshold) leading to a worsening of the SPTR proportional to the number of active MPPC cells. The merit of the differential readout for timing becomes evident when we compare the observed SPTR value of the MPPC-matrix with that of the single-ended readout. This improvement is also confirmed by the CTR measurement results. We measured the timing characteristics of the array using a 4×4 LYSO crystal matrix (3×3×15 mm3 per cell). The improvement in CTR (FWHM) from 283 ps to 226 ps is obtained by implementing the differential readout scheme. The results suggest that for critical timing applications, a differential output from each individual cell is an advantage.