To characterise the entire surface of the CCD we need a criterion to determine the pixel capacity on flat field images. Two quantities can be used: the evolution of intensity (ADU) versus flux and the evolution of variance. The first shows the break on the linearity curve and the second identifies the case where the poisson statistic (average equals variance) is no longer respected. We use the variance to determine pixel capacity because it is more sensitive.
Flat field images are taken with white LED illumination and with different flash durations. For each flash time we compute the variance and the average on a difference of two images and on boxes of 40 by 40 pixels at different positions. For each position we plot the variance and average versus flash duration. On this plot the pixel capacity is the average of the image about which the variance is decreasing.
The result on the entire surface of a CCD is presented in Fig. 2. Each square represents the pixel capacity of that position. This capacity map shows a difference between the center and the edges of the CCD. This difference is observed on five of the ten CCDs tested. The saturation in the center of the CCD occurs during charge transfer because charges spread upward. This non-uniformity is explained by the difference between the well shape during transfer at the edges of the CCD (where phases are injected) and the shape in the center far from the injection point. This difference is due to the resistivity and capacity encountered by phases to reach the center of the CCD.
Was this article helpful?