(by Paul Breitenstein from the Pascal-Gymnasium in Münster)
Since two months the problems with the focus are solved. We successfully use a linear fit of the 4,646 best images between Aug 2012 and Jul 2013 with an updated offset foff:
f(Ta,T1,T2) = -0,01751 * Ta - 0,00334 * T1 – 0,00456 * T2 + foff .
The focus depends on the temperature of the environment Ta, the primary mirror T1 und the secondary mirror T2. The fitting has been done by the AiM-Projectgroup of the Pascal-Gymnasium in Münster. According to this group of students this model predicts the focus with an accuracy of +/-0,05mm (s=0,025).