Subject: Re: Tech note

Dear Carol Anne:

Here is the technical note.

niels


****


JEM-X instrument parameters

JEM-X tech. note. Niels Lund, 24/5 2004

Using a large set of Crab observations from revolutions 43, 44, 
102 and 170, and intruducing four new parameters describing 
imperfections in the collimator I have been able to reduce the 
rms-deviations of the derived Crab fluxes by about 30 %. 

Still, the dependence of the derived fluxes on the pointing 
direction is very significant, and this points to problems in
other areas of the instrument modelling, I suspect in
particular the modelling of the distribution of assigned photon
positions relative to the position of the photon at the level
of the detector window.

Another issue which may be important is whether there remains
some residual errors in our pixel position data. Certainly, the
data we have is a good approximation, but I strongly recommend
that we use the collimator shadows in a systemmatic way to 
check our existing transformation from pixel number to pixel
coordinates.

I have divided the pointing directions into four groups according
to the off-axis angle. The parameters heve been optimized for 
the group of pointings up to 5 degrees off-axis angles.

          Pointings inside    3 deg   4 deg   5 deg   6.6 deg
          No. of pointings     392     528     764      836

JEM-X1
rms deviation of Crab flux:
old parameter set:            4.09%    5.09%   6.17%   6.52%                 
new parameter set:            2.86%    3.31%   4.32%   4.50%

w. CBJ coll. rot. value       2.92%    3.42%   4.53%   4.68% (not yet optimized)

JEM-X2
rms deviation of Crab flux:
old parameter set:            5.46%    6.20%   6.71%   6.89%                 
new parameter set:            2.64%    3.36%   4.27%   4.45%

w. CBJ coll. rot. value       2.90%    3.83%   4.55%   4.68% (not yet optimized)

The statistical uncertainty on the individual determinations are
insignificant, they are much less than 1% for those near the axis to 2% 
for the those at the edge  of the field.

PS: Later adjustments of the parameters, keeping the collimator
rotation fixed at the values found by Carl have yielded equally
low values for the rms values.