Determination of the multiplier gain of a photomultiplier

Abstract

Two means for measuring multiplier gain are described. The determination of the total number of single electron pulses, N, enters into both methods. Use of a multichannel analyser of calibrated input sensitivity reveals that the error in determining (g) arises mainly from an inability to assess whether all contributions from the low energy region of the pulse height distribution have been included in N. It is concluded that mean gain measurements with better than 10% accuracy are unlikely for most photomultipliers. This has serious implications for the absolute measurement of light intensities through the need to know F, the collection efficiency, the the determination of which can be sensitive to the assumed values of (g) and var (g).