Signal Analysis for Radiation Event Identification

Abstract

An analysis program has been developed to identify scintillation pulses using the digitized signals from two photomultiplier tubes viewing the output from a common source. The analysis is directed towards identifying the noise pulses in the background signal level. In this work a neutron detector test system is constructed and methods incorporated into a program for processing digitized signals from the detector so as to determine the effectiveness of amplitude discrimination for identifying neutron pulses in the presence of spurious events. Two methods are used for identifying pulses, one based on locating a monotonic increase in the signal amplitude exceeding some threshold value and the other on a locating a number of sequential signal data points exceeding the mean and standard deviation of the baseline of the record. The software written for controlling the digital oscilloscope is flexible and allows the selection of records to be taken such that statistically useful properties can be calculated as to the loss of efficiency that results from using a discriminator to remove noise. Experimental results are presented for a lithiated sol-gel based neutron scintillator tested using this system.