What could explain inconsistent counts?

Hello everyone, I’m simulating gamma spectroscopy using an HPGe (scoring volume) detector. However I’m getting very inconsistent count rates in between small changes to my simulation. For example, I was getting 30 000 counts for a 166keV peak, removed a wall outside the detection chamber and now the peak reaches a maximum of 120 counts (this is for 10^7 events Ba139). Similarly a few weeks ago I was getting a number as high as 400*10^3 counts. I can’t explain why I can’t get a consistent number of counts, though maybe it’s also an issue with my ROOT files.
Thanks in advance.

Additionally I’m getting a peak at 102 keV that I can’t really explain… Any help is appreciated.

The peak at 102 is most likely a backscatter peak, and is a function of your geometry surrounding your detector. As for the counts, there is little we can help with without more details. What specific things did you change, and what was the direct result?