New detector

One issue to consider when changing to a new CCD camera (see above) is that the pixel size is slightly larger (15 vs 13.5$\mu m$). This results in a slightly larger spatial size and wavelength step per pixel. In particular, we will have a slightly larger field-of-view and wavelength range (for the higher resolution grisms). It was noted that the old CCD for ALFOSC that was replaced in 2005 also had 15$\mu m$ pixels, and when it was replaced by the current CCD the instrument camera optics were replaced, keeping the projected size of the pixels practically the same. However, measurements made at the time indicated that the old optics matched to the 15$\mu m$ pixels appeared to have a higher throughput than the current optics, in particular in the red where the new CCD will be especially attractive because of its high QE and low fringing. The optics consistent of the camera barrel, that can be used to focus the instrument, and the entrance window of the dewar. To see if there is truly such an advantage in going back to the old optics the idea was to do measurements with the same set-up and the current CCD, with the current and the old optics. We have been able to locate the old barrel with the optics in it, but we have not been able to locate the old dewar entrance window. Still, we have tried to measure the difference only changing the barrel, but as the current dewar entrance window is not matched to the optics in the old barrel we have not been able to properly compare the measurements. However, we are looking at doing some new tests in the coming months.

There were some questions about the exposure level to which the ALFOSC CCD can be used safely with the new controller. In particular, it was noted that the documentation was not so clear. An important thing to note is that the relevant level for observing is actually the level where non-linearity effects become important which is significantly below the level where true saturation of the signal occurs (which is way below the maximum level where the controller becomes digitally saturated which was an issue with the old controller). This was clarified in the documentation.