When closing NOTCam again after the array electronics upgrade (see above) we did not achieve a good vacuum. After the installation of the Science Array NOTCam seemed to cope relatively well during the commissioning night 13/12, but over Christmas the leak was so strong that we had to let it warm up and open to examine the situation. During these tests we did not find the cause of the problem, and the January NOTCam runs had to be done with the pump attached to NOTCam on the telescope, pumping all the time. When finished, NOTCam was again examined and leak tests pointed to the entrance window. First the entrance window o-ring was suspected, but eventually a tiny crack was found in the window. It was exchanged with a spare window.

It had been noted that while NOTCam was in Copenhagen a vacuum better than $10^{-6}$mbars has been achieved, while since the instrument has been at the NOT we have never seen this. Following a review of the literature the whole instrument was baked to a maximum temperature of 40C for several weeks and more care was taken with handling of component when the NOTCam was open, i.e. use latex gloves always and not clean the internal surfaces or the O-rings with any liquids. Since then we have achieved a better vacuum than ever when pumping in a warm state, which made us believe we must have had a tiny micro-crack in the window for a long time. However, when cooling down there were some poorly understood vacuum losses and we only achieved a pressure of $10^{-4}$mbars which is not better than the typical pressure we had throughout 2007. A possible reason why we can not get a better pressure is now believe to be the slow degradation of all the O-rings which have been inside since NOTCam was made. We have purchased new O-rings for the main vessel and for the entrance window. There are various mostly small O-rings which are very hard to reach and remove with the risk that we might make things worse. The current plan is see what happens when we exchange these two main O-rings. If that does not improve things we are thinking about doing some vacuum tests with one of the smaller O-rings that can be easily reached and compare those with tests using a new O-ring to see if there is any significant difference which might explain the relatively poor vacuum.

Thomas Augusteijn 2008-11-14