Recovery of coincident frequency domain multiplexed detector pulses using sequential deconvolution

TL;DR

This paper introduces a sequential deconvolution method to recover multiple detector pulses in frequency domain multiplexed signals, enabling improved signal analysis in radiation detection systems.

Contribution

The paper presents a novel sequential deconvolution technique for recovering multiple coincident detector pulses in frequency domain multiplexing systems.

Findings

Accurate estimation of charge and arrival time for non-overlapping pulses.

Partial recovery of first pulse when pulses overlap in time.

Limited recovery of second pulse parameters when pulses overlap.

Abstract

Multiplexing of radiation detector signals into a single channel significantly reduces the need for a large number of digitizer channels, which reduces the cost and the power consumption of a data acquisition system. We previously demonstrated frequency domain multiplexing by convolution using a prototype system that multiplexed two EJ-309 organic scintillators signals into a single channel. Each detector pulse was converted to a damped sinusoid which was then combined into a single channel. The combined signal was digitized and the original detector signal was recovered from the damped sinusoid by deconvolution. In this paper, we demonstrate the recovery of multiple detector signals that arrive during the same digitized record via a new sequential deconvolution method. When two detectors produce signals in the same digitized record and their pulses do not overlap in time, we found that the charge, arrival time, and particle type can be estimated fairly precisely for the first pulse, but the second pulse exhibits substantial degradation in the precision of the estimated charge and arrival time. When the pulses overlap in time, we demonstrate both theoretically and experimentally that the part of the first pulse that does not overlap with the second can be recovered accurately, so the arrival time and amplitude of the first pulse can be estimated fairly precisely, but not the charge or particle type. None of these quantities can be estimated precisely for the second pulse when the two pulses overlap.