Distributed and dynamic routing logic for high-speed data extraction from single photon avalanche diode arrays

Time-resolved imaging by means of Time Correlated Single Photon Counting (TCSPC) is the enabling technology for several powerful analytical techniques, such as Fluorescence Lifetime Imaging Microscopy (FLIM), F¨orster Resonance Energy Transfer (FRET) and Fluorescence Lifetime Correlation Spectroscopy (FLCS). Even more, TCSPC could further extend its application also to other research fields, for example in-vivo analysis, if only its relatively long acquisition time could be reduced. To achieve such a challenging result, multi-channel integrated systems have been developed to increase the measurement speed, but posing at the same time tight constraints on power dissipation, area occupation and interconnection complexity inside the chip. Moreover, the high rate of events potentially produced by a huge number of detectors can easily saturate the transfer bandwidth. To address this problem, we designed a dynamic router optimized to use the transfer bandwidth of the system at its best. The router is based on a fully-integrated, 150nm-technology, distributed routing logic that permits to interconnect different parts of the system, each one optimized by exploiting a different technology node selected on purpose. In this work, we present a 32-channel prototype of the router, but we'll show that it has been designed to be easily up scalable and to be the building block of a densely integrated system.