The cause of the varying and large readout overheads is believed to be mostly due to an aging ISA PC-board. Nevertheless, using a `spare' ISA PC board for testing gave no difference. It was suggested to move the whole NOTCam data acquisition system to the old computer `elizabeth' and try to run the software from this computer. It was also noted that a new computer recently installed for SOFIN by Dr. Ilya Ilyin (Potsdam) appears to work as well with ISA PC boards as the old (outdated) system computer. We have been trying to purchase one of these computers, but have not obtained a reaction from the (German) company yet.
In the end, we should get rid of the large readout overheads (unpredictable as well as expected ones) with the new controller. In the mean time, a special exposure command for NOTCam was suggested in which the dither overhead can be folded into the readout and file storage overhead. This could save up to 1.6 hours per night, depending on the type of NOTCam observations being made. Work is on-going and such a new expose command is currently being tested. For the notcam.expose command, which now has 13 seconds overhead, it is expected that this work-around would lower the overhead to around 6-7 seconds per exposure, quite an impact for short exposures. It is not clear whether a similar approach is worth-while for the frame command, since it is not clear exactly where during the exposure command the hanging/overheads are `happening'. The multi-exposure command notcam.mexp is also more complicated, but we will look at this as well.
Thomas Augusteijn 2010-05-27