CdZnTe spectra improvement through digital pulse amplitude correction using the linear sliding method

Abstract

This paper presents innovative algorithms for optimum filtering, rise-time discrimination and proper correction of solid-state detector pulses based on digital pulse processing (DPP) techniques and on a Linear Sliding Method. These techniques allow a significant improvement on the spectral resolution by taking into account both the electronic noise and the pulse rise time fluctuations mainly due to the incomplete charge collection in the CZT crystal. In order to develop these algorithms pre-amplified pulses coming from a planar XR100-T detector (Amptek) were digitized by a 12-bit PCI acquisition card (125 MHz) hosted by a PC platform. The presented results show that, due to the possibility of fine adjustments on the discrimination and correction parameters during the processing stage, this flexible DPP implementation leads to significantly better results than the analog hardware discrimination solution. Comparative analysis of Am-241 spectra show a significant reduction of the low-energy "tail" on the 59.5 keV photo-peak.