Figure 7: Gain, raw and light-level corrected QE, and dark count rate
for a 0.5 PE threshold, for all characterized channels at T=6.5 K and
U=6.5 V
Figure 8: Sample mean and rms for dark current, gain, and QE as a
function of VLPC bias at T=6.5 K, and as a function of operating
temperature at U=6.5 V.
Figure 7 shows the distributions for gain, raw and corrected (for light level nonuniformity in the test cassette) relative quantum efficiency, and dark count rate for the whole sample at the nominal operating point. The accumulation of entries at half the nominal gain is due to defective preamplifier channels.
Figure 8 shows the dependance of the observed means and root mean square widths of these distributions on temperature and bias. One sees the strong dependance of noise and gain on bias voltage, as the latter approaches the breakdown voltage, whereas the quantum efficiency approaches a plateau. As shown in [8], the optimum balance in this tradeoff between QE and dark count rate depends on the light levels to be detected in the particular application.
