Dear all, I found a mistake in the calculation of number of p.e. (a counter was not properly reset at the beginning of the event). I have now 12.13 p.e. with 3 parameter fit and 0.59 background events. Marco