From naranjo Fri Nov  5 05:53:51 2004
Date: Fri, 5 Nov 2004 05:53:51 +0000
To: mjloughlin@natural.demo.co.uk, APX077@coventry.ac.uk,
        nigel.hawkes@npl.co.uk, colin.murray.01@bbc.co.uk
Cc: puherman@ritva.physics.ucla.edu, camara@physics.ucla.edu,
        shopkins@scs.uiuc.edu
Subject: AmBe neutron detection efficiency
Reply-To: Brian Naranjo <naranjo@physics.ucla.edu>

Hi All,

Michael Loughlin suggested I confirm that the number of neutrons detected
in BBC 03 is consistent with a 37 MBq AmBe source.


MEASURED AMBE DETECTION EFFICIENCY
----------------------------------
In BBC 03, 2262 8 ms traces were recorded for a total live time of 18.1 s.
During this interval, 406 neutrons were detected, giving a net detection
rate of 22.4 neutrons per second.

According to _Neutron Sources for Basic Physics and Applications_
(Pergamon Press, 1983), the AmBe source gives off 6 10^-5 neutrons per
Am decay.  So, the net neutron output is
  (37 10^6) (6 10^-5) s^-1 = 2200 per second.

Therefore, the measured AmBe detection efficiency is 22.4/2200 = 1.0%


CALCULATED AMBE DETECTION EFFICIENCY
------------------------------------
I modified my BBC Monte Carlo to use an AmBe energy spectrum instead
of the monochromatic 2.45 MeV source used earlier.  I also placed the
source approximately where it was during the run (see 'bbc_ambe_top.gif
 and 'bbc_ambe_side.gif').

Out of one million neutrons emitted, 12404 neutrons caused scintillations
in the acceptance range between 300 keV and 1050 keV (see 'bbc_ambe.pdf').
Remembering that 10% of neutrons aren't detected because they are below or
above the proton PSD region, the estimated detection efficiency is
  (12404/1000000) * 0.9 = 1.1%


The discrepancy between 1.0% and 1.1% may be due to the source's
unknown calibration.

Brian
