As seen from the alpha particle testing, the spectrometry performance of the CPS32 is limited by peripheral hits and pixel-to-pixel variability, issues unrelated to radiation effects. It is, therefore, not surprising that there was no clear degradation of spectrometry performance with radiation. We will therefore assess the radiation effects, not against spectrometry performance, but against signal-to-noise ratio.

First, it is important to note that, up to the 10 krad dose tested, the CPS32 remained entirely functional. The dominant total dose effect was an increase in the 25ºC dark current of 0.58 nA/cm²/krad. This is an order of magnitude larger than the increase expected due to displacement damage, confirming the assertion that the total dose effects dominate. Although the dark current at 10 krad is 20 times its initial value, it still amounts to only 4500 e-/ pixel for a 100 ms integration time. The shot noise on this value is
65 e-, comparable to the noise floor of the device. Thus, the dark current would not be expected to degrade the device in any case.

The dark current FPN, which would be relevant to imaging devices, is 300 e- for the same conditions. Furthermore, this FPN is proportional to integration time, whereas the shot noise increase with the square root of integration time. Thus, for imaging, the dark current FPN is likely to be the main performance limitation.

Dark current has the additional effect of reducing dynamic range by leading to saturation. At the 10 krad dark current, saturation will occur in 13 sec.

Because the dark current is thermally activated, the dark current problems will be aggravated at higher temperatures and ameliorated at lower temperatures. In particular, for the STRV-2 application, the dark current will be 8 times greater than the stated values at +60ºC, while at –40ºC it will be 200 times less. Clearly, the usability of these devices will remain to much higher doses at the cooler temperatures.