Unexpected results when retracing origin of photons detected in Ge detector

Hi there, I am using a Ge detector covered by a thin sheet of Al and one made of polypropylene (to simulate a marinelli beaker which I have not yet been able to implement fully). I generate the photons, using the particle gun, in a marinelli volume made of Gd2(SO4)3.

I am interested in the initial positions of the photons that deposit all their energy inside the Ge detector. (I use the exampleB1 as template) To do that I define the x,y,z coordinates in the B1PrimaryGeneratorAction and then I call them as extern in the B1EventAction where, inside the EndOfEventAction() function, I compare the energy deposited to the energy of the particle gun.

I would expect the distribution of initial coordinates to resemble the marinelli´s geometry with higher density of points closer to the Ge detector. But I get the folowing :

Is my initial asumption that the distribution of initial coordinates should resemble the marinelli´s geometry corect? If so, what could be causing this effect?

I ran the simulation twice, once with the default random engine and once changing the random engine, but I get similar results.

Thank you for your time.

ps: I don´t know if this is the coorect category to post this issue.

The issue has been solved, the problem was the use of the same angle when defining the momentum and position in the B1PrimaryGeneratorAction file.