We experimented with operating the output amplifier of the CCD42-90 in a state which lowers the gain by a factor of 5 and increases the full-well capacity by a similar factor. This is accomplished by raising OG2 to roughly 12 V above the substrate voltage (see the manufacturer's data sheet). Unfortunately, limitations in our electronics only permitted us to reach 10 V above substrate. With this value of OG2 we realized a factor of 3 reduction in gain, but also observed a noticeable amount of non-linearity. We are not currently supporting this mode for observers.
In normal operation we use a gain of 3.5e-/ADU, and a 200kHz pixel clock. In this mode, we measure a read noise of 4-5e-. The output amplifier of the CCD saturates around 60,000 ADU, giving us nearly optimal dynamic range. We have measured no channel-to-channel crosstalk down to the 10-5 level.
Was this article helpful?