Minutes of the E00-102 Meeting
Tuesday, November 08 2005

Fissum

KF/JL/BR/AS/LW

o JL is Jeff Lachniet.  Welcome Jeff!  LW announced that his
new PhD student Kurniawan Foe has recently joined the project
and is coming up to speed.  JU and the Madrid group will send
a PhD student to Lund in the first half of 2006 to start work
on building the experiment in software.  KF continues to stage 
for the possibility of an "analysis sabbatical" at JLab during
the second half of 2006. 

o KF presented MA's Lic thesis.  It is available for download
at http://www.jlab.org/~fissum/e00102/figsetc/lic/lic_uslet.pdf
for those who are interested (substitute "a4" for "us_let" for 
the worldwide version).  KF will try have some bound versions 
ready for the next analysis meeting (see below).  AS questioned 
the kinematics-to-kinematics variation in the 1p12-state momentum 
distribution between the HRSh angular settings (see Fig 3.16 in
the Lic).  We attribute this variation to the fact that no effort 
has been made to keep (Q^2,omega) constant in this analysis.  
Further, corrections to the data such as VDC tracking efficiency 
and electronics deadtime (anticipated to be very kinematics 
dependent -- see below) have not been made.  

o JL and LW went through the analysis "todo" list step-by-step 
to outline the progress they have made since JL joined the project.  
Strategically, it has been decided to change analyzers to the 
ROOT analyzer as JL is much more comfortable with this modern 
toolkit.  This of course means that all MCEEP work (see below) 
will have to be "ROOT formatted" for eventual comparisons between 
data and simulation, but this is supposedly painless.  Stay tuned.  

o Before completely recooking the data, time will be spent 
chasing down the remaining questions.  In order to do this, JL 
does a cut-and-paste of the ESPACE database into the ROOT database, 
which is obviously a tedious process.  But this has allowed 
investigations of:

  - problems with the electronics deadtimes.  The EDT algorithm
    predicts very large electronics deadtimes which in turn 
    require gate widths (~700 ns) which are much larger than 
    they were in order for any internal self-consistency checks.  
    Something is clearly not understood yet.  Basically, events 
    coming sooner than the pulser self-timing event "steal" the 
    stop, which results in deadtime.  The ratio of the integral 
    of the grass to the integral of the spike in the deadtime 
    TDC is related to the EDT, but it is not quite obvious as to 
    how.  Work remains.  LW would like to examine the correlation 
    between the EDT TDC and VDC T0 distributions.

  - proton-detection efficiency.  Acceptance cuts are enforced
    on the electron-arm data in parallel kinematics such that
    the proton is kinematically constrained to turn up within the
    proton-arm acceptance.  96% is the result.  At first glance,
    this seems consistent with proton absorption in the trigger
    scintillators (especially S0) and the spectrometer windows,
    but this needs to be checked by simulation.  This can also be 
    cross referenced against the trigger-efficiency information 
    we have from the S1, S2, and S0 scintillator planes.

  - VDC tracking efficiency.  We define this to be "the ability of
    the VDCs to reconstruct a track".  Two schools of thought 
    exist -- aggressive cuts so that one knows for certain
    the track was good at the expense of statistics, or passive 
    cuts so that the track may or may not be good, but many more 
    events are accepted.  BR strongly suggests that we adopt the
    latter school of thought based upon the work of Paul Ulmer
    in the deuterium analysis.  This is because checks of the
    results obtained from the different schools of thought have
    demonstrated that the passive cuts produce much more stable
    results.  That is, if the cuts are varied slightly, passive
    cuts yield about the same tracking efficiency while aggressive
    cuts yield dramatically different results.  There is clearly
    much more work to be done here. 

  - white-spectra issues.  Recall that we operated the electron 
    spectrometer at an extreme excitation.  As a result, we were
    not able to step the spectrometer excitation higher than the
    default excitation as is the common practice in a white-spectrum 
    measurement.  The consequence is that our white spectra do not 
    resemble white spectra at all for the electron spectrometer.  
    Recovering the relative focal-plane efficiencies will thus 
    require further work.  Algorithm to be determined.  Again,
    stay tuned.

  - BCM calibrations.  These have been questioned.  AS agreed to
    investigate.

o JL requested that KF add a tarball of MA's MySQL work to the
  E00-102 website.  This will allow the various people working
  on the data analysis to have their own individual copies of
  the database as we are having a difficult time finding a host
  who will allow us to leave the database live and online.  We 
  will try ODU again, but in the meantime BR offered a JLab PC 
  as a host.  LW requested that KF speak with the ODU sysadmin
  re. a web-based logbook so that we can better coordinate the
  data analysis and simulation work on either side of the Atlantic.  
  KF has been using the Oak Ridge logbook at MAX-lab with reasonable 
  results.  But again, we are having trouble finding a host for 
  the project.  Any ideas or offers of hosting would be appreciated.

o next meeting is Tuesday, November 15 at 14:00 in B101.