SiPM fiber test

[Beni Zihlmann <zihlmann@jlab.org>]

Hi All,
it occurred to me that there is another correction one has to apply in 
order to get
the mean number of photo electrons caused by electrons from a Sr90 
source when
using a SiPM. I have put a draft  of a document I am writing about the 
fiber tests
I do here at Jlab on the doc dtabase 1234-v1 where in section 2.2 I 
describe the results
when I also use a SiPM.

The spectrum of the SIPM with no Sr90 source can be viewed as random 
triggers, noise in the trigger PM
and when a cosmic is going through the fiber. Both cases are independent 
of the Sr90 source trigger. That
is the reason why I think one has to subtract the normalized background 
spectrum (normalized to time) from
the Sr90 spectrum as Andrei did. But the resulting spectrum is still a 
convolution of the the Sr90 electron event with the
dark rate of the SiPM which is to first order the background spectrum 
and hence on has to subtract the mean
of the background Poisson distribution from the Sr90 source Poisson 
distribution.

After subtracting the background spectrum (no Sr90 source) from the main 
Sr90
source spectrum the remaining spectrum is still a convolution of two 
Poisson distributions.
Namely one from the light produced by the electrons from the Sr90 source 
and the other from
the dark rate of the SiPM. Therefore, after extracting the mean number 
of photo electrons from the
background corrected spectrum one has to subtract from this the mean 
number of photo electrons
as obtained from the background spectrum alone.


cheers,
Beni

-- 
***********************************************************
Benedikt Zihlmann
TJNAF
HallD 12C/ F350
Suite 8
12000 Jefferson Av.
Newport News VA, 23606

tel: 757 269 5310 
fax: 757 269 6331                zihlmann@jlab.org
***********************************************************