Minutes:
Attending EG,AS,YX,BT,MK,TA
EG:
-Earlier reported factor 2 yield reduction by wire chamber veto is
not on final event sample but rather at an earlier stage in the
cutting sequence.
-Showing histograms of wire chamber hit multiplicity. On final event
sample after all other cuts, 2.xx % of the candidate events have wire
chamber hits, distributed such that the multiplicity peaks around 20.
Deploying the wire chamber veto by requiring "no reconstructed
additional track" leaves a hit distribution that covers 0.57 % of the
65k surviving events.
-Fraction of candidate e,e'n events that have a charged track and TOF
hit in the neutral sector (=trig 1) in addition to a neutron hit does
not exceed 15% (pulled out from DST).
-Event selection from flr and DST now gives consistent results.
-About random coincidences: longer discussion on how to handle random
coincidences. A random coincidence is an uncorrelated event between a
"good" electron and a "good" neutron not originating from the same
process. In this case the neutron time-of-flight spectrum (Tnn)
exhibits a broad (time-uncorrelated) distribution, extending into the
unphysical region of beta_n>1. These randoms are supposedly diluting
the asymmetries (not necessarily). Up to 6% of the total e,e'n yield
in average could be attributed to randoms. The asymmetry of these
randoms could in average be very similar to the one of the real events
since we still have a quasi-elastic electron and no proton. The "true"
neutron may have been eventually missed (happens to 70% of the
neutrons!).
-TA suggests the method of event-mixing to determine how random hits
are distributed in dynamical variables like pmiss or Mmiss or Tnn.
After applying all cuts except for the ones depending on Tnn, one
winds up in an event list for sector-related (or PID-related)
information, like electron-left and neutron-right. If one now
combines electron-left information of the i.-th event with
neutron-right information of the i+1.-th event, one can construct a
neutron Tnn spectrum based on random coincidences (Tnn=tnn-twL+Twl,
where tnn is the offset-subtracted and ns-converted neutron-TDC, twl
the corresponding electron-TDC, and Twl is the electron time of
flight calculated from track reconstruction). In principle, the start
time of the TDC measurements may vary event by event, but this should
be only 5ns at most. On the other hand for randoms, the TDC
differences of neutron and electron may vary up to 100ns (the width
of the coincidence gate). So, Tnn for mixed events should represent
the distribution of randoms, convoluted with the variation of the
starttime which will not have a big effect though. By comparing the
yield in the unphysical region of Tnn for the mixed events with the
corresponding one for the original events, one can determine a scale
factor for the fraction of randoms in the physical region of the
original events. Not only this, one can also subtract the then-scaled
random distribution from the yield of the original events in order to
get background-subtracted yields. In principle, this can be done for
each spin state separately such that even polarized backgrounds could
be handled this way. For unpolarized background, it is equivalent to
applying a background dilution factor to the original asymmetry.
Note that the background correction due to empty target or, in case
of e,e'n due to hydrogen target is totally based on real
coincidences, not on randoms.
Regards,
Michael
--+-------------------------------------+--------------------------+ | Office: | Home: | |-------------------------------------|--------------------------| | Dr. Michael Kohl | Michael Kohl | | Laboratory for Nuclear Science | 5 Ibbetson Street | | MIT-Bates Linear Accelerator Center | Somerville, MA 02143 | | Middleton, MA 01949 | U.S.A. | | U.S.A. | | | - - - - - - - - - - - - | - - - - - - - - -| | Email: kohlm@mit.edu | K.Michael.Kohl@gmx.de | | Work: +1-617-253-9207 | Home: +1-617-629-3147 | | Fax: +1-617-253-9599 | Mobile: +1-978-580-4190 | | http://blast.lns.mit.edu | | +-------------------------------------+--------------------------+
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:32 EST