Due to self-absorption of the sample matrix, some of the particles have less energy then what it started. As a result, the pulse height spectrum displays a continuum of pulses throughout the voltage scale, most falling in the alpha window and some attenuated particles counted in the beta window. This tailing of the alpha window pulses is a result of greater number of pulses in the lower voltage region, accounting for the alpha spillover into the beta window. It should be expected that as the operating voltage is increased, the majority of the alpha pulse would begin to migrate fully in the alpha window thus minimizing spillover. Yet in the following figures, at voltages over 650 VDC, there is upturn to the net count rate, outlined by the shaded box with the figures.
This observation was investigated and it was determined that the small contribution of the pulses in the o window of the alpha spillover may be secondary peaks of high energy pulses. This explains the high spillover data at lower operating voltages during the figure of merit discussion. The secondary peaks are not seen at the lower voltages because they are discriminated out by the lower threshold voltage. So when the voltage goes up, the peaks migrate past the threshold and are counted. Adjustment to the lower window threshold will help mitigate its