Computer correction for shifts in gain and baseline in gamma-ray scintillation pulse height spectra

Abstract

In the quantitative analysis of multicomponent pulse height spectral data, precision depends upon 1. the statistical variation inherent in the radioactive decay process, 2. statistical variations in the detection process and 3. instrumental variations. The nature of the statistical variations in categories 1 and 2 are well understood and easily considered in the estimation problem. Category 3 variations are more difficult to evaluate and correct. The main variations observed in this category are of gain and baseline. In order to determine the instrumental conditions that prevail at the time of measurement, the spectrum of a standard ( 44Ti) was obtained immediately before experimental spectra were obtained. This standard spectrum was then compared to a reference standard spectrum to obtain estimates of gain and baseline status, and appropriate corrections were applied to the experimental data. This technique, which can correct for shifts in both baseline and gain in excess of one channel (with respect to channel 256 of a 256-channel system), is capable of making corrections of as little as 1 20 of a channel. Improvement in the resultant quantitative estimation of the constituents of multicomponent pulse height spectra is demonstrated. The method should provide the basis of a good quality control procedure for the scintillation gamma spectrometer, provided the spectrometer data can be conveniently entered into a computer.