The varying `dark-current' was thought to be caused by a design error of the printed circuit board (PCB) that sets the `biasgate' voltages. A new PCB was designed and changes to the code of the detector controller were made by Preben Nørregaard (CUO). The new board and code were installed with the engineering grade array and tested to ensure no problems were encountered. After successful tests the detector was swapped for the new science array. Unfortunately the changes have made no difference to the dark current behavior, though it has now provided us with the extra biasgate voltage to adjust, which can be used to get a larger dynamic range at the expense of readout speed.

Because the change of the NOTCam bias voltages did not solve the problem with the dark current, the next idea is to modify the clock drivers. At the moment the clock levels are from 0 to $\sim$3.5V where the specification is for 0 to 5V. To get the higher level intermediate buffers will be mounted on the controller clock drive board that will convert the existing 3.5V to 5V.

Still, a major improvement has been the installation and commissioning of the science array. A web page with the commissioning report can be found at: and the

A web page detailing the main characteristics of the array can be found at:

The new array is about 10% more sensitive than the previous science array and up to 60% more sensitive than the engineering grade array. It is cosmetically good-looking with $<$ 1% bad pixels, of which most are located in one corner. It has one bad feature, about 33x13 pixels, centred at x=150, y=508. It has a very flat quantum efficiency over the whole area, even better than the previous array. Still some tests needs to be done to investigate the level of crosstalk between the different quadrants, determine the precise plate scale with the different cameras and map the photometric scatter across the field of view.

The interference pattern seen since mid 2007 is present also with the new array electronics and the new array, but at a relatively low amplitude.

Thomas Augusteijn 2008-11-14