Modern VLSI technology allows the development of new class X-ray imaging detectors capable of capturing an image in various energy ranges in one shot. Such spectroscopic imaging detectors have a high demand for the spatial and energy resolution of individual photons. With decreasing size of pixels, the charge cloud generated by the primary photon interaction, and in high-Z materials also by the fluorescent photon interaction, is shared across several pixels. That limits both spatial and spectroscopy resolution. This paper presents a novel charge summing algorithm called Winner-Master-Slave and describes the functionality, implementation, and simulation of the proposed algorithm in Verilog-AMS.
Several charge summing algorithms have been developed and implemented in ASIC. Unfortunately, most of the algorithms sum charge just from the four neighboring pixels. That prevents the possibility of collecting charge from more considerable distance which is generated by long-travel fluorescent photons. That limits spatial resolution as well as spectroscopic capabilities of detectors.
In this work, a new fully asynchronous algorithm called Winner-Master-Slave (WMS), which deals with charge sharing effects as well as with long-travel fluorescent photons, is presented. WMS is designed to work fully asynchronous, and it is capable of charge summing over 25 pixels (cluster size 5 x 5). Every interaction of a photon with the detector which generates a charge cloud is called an event. WMS is event-driven; thus whenever an event occurs WMS automatically creates a cluster and sums total charge in the cluster.
Described functionality of the novel algorithm has been implemented in Verilog-AMS and simulated with AMS simulator in Cadence® Virtuoso® environment. The algorithm has been systematically tested using randomly generated clusters. The simulation outcome proved that WMS algorithm solves several problems of existing solutions and is suitable for implementation in a front-end ASIC for a spectroscopic imaging detector.