With last week's curious results about residuals, I decided to
look at what happened when I opened up the road window from .02 to .03.
I added 81 more 4-hit tracks (from 3098 to 3179), but the distributions
look the same. Opening up to .05 makes no difference.
So, I look at the 3-hit stubs. The results
are interesting. The first plot shows the impact parameter of the
3-hit stub with a simple chi-squared cut for the cases of the missing
hit being layer 0, 1, 2, and 3. The second plot is of the missing hit.
It is known and understood that the efficiencies are somewhat lower
for the cells 0 and 3, because the pulse height, not strengthened by
reflection from the far end, is lower. However, there are twice as
many missing hits in the second two wires as in the first two. The
plots missing 0 and 3 are substantially broader, as one might expect,
but also double-peaked. This may be partly explicable from geometric
considerations. If the hit in layer 0 is missing because the track
missed the cell, then if the track is perfectly radial it will also miss
cell 2 and thus not be picked up at all.
Bob has revised his previous plots, using better time corrections.
The East Front Bsu timing plots,
the West Front Bsu timing plots, and
the Tsu or Rear Bsu timing plots. These
are noticably tighter than before, and the resulting resolutions
are now consistent with those from cosmic ray tests. Also notice
the z distribution plot. The first line
marks the center of the overlap region between front and rear Bsu's,
and the second marks roughly the end of the front. The dip is
probably due to the eta gap moving from front Bsu to Tsu coverage,
and the second peak is in the right place to be due to the 30 degree crack.
Tom LeCompte showed some plots of eta distributions in the BMU which
did not make sense, especially the one that was supposed to be the
eta distribution of the plot that is of muons with both a CMX and BMU
stub. I still don't understand these. I thought at first there
might be a sign error in the offset correction, but changing that only makes
things worse. (Had to fight with
talk MuonLinkerModule
firstExtrapolator set M
secondExtrapolator set M
exit
