There have been attempts last night to use the extended histograms
provided by Online Monitor for estimating the tagging efficiency. The
outcome they got was quite meaningless, today I have found the bug and
hopefully fixed it. I apologize for that: the nearly zero efficiency was
a flaw in the analysis of the data by Monitor, not a fault in the
apparatus. Let me explain how it happened.
Under item 7 of the extended set of monitor histograms you will find a
plot that says "Estimated Tagging Efficiency". What I define by tagging
efficiency in that context is: the ratio between the counts in the TAC
that are in time with the tagger, by the total counts in the TAC. The
ratio is performed 61 times for each bin of energy in the tagged
Bremsstrahlung spectrum. This definition would work approximately well
if you analyze last September's commissioning runs with MOR alone in the
trigger. Last night's runs (10077 to 10093), instead, had TAC*MOR in the
trigger, so you would expect 100% of the so-called tagging efficiency in
the plot for all bins, but not actually a significant information about
the real tagging efficiency.
Anyways, the program was giving values close to zero instead of 100%.
The reason for that is that the TAC timing changed since last fall. In
September analysis the peak of TAC TDC was about 1000 channels, and "in
time" was defined by a cut at 500 through 1500 channels. Now the same
spectrum turned out to be peaked about 2000 and Monitor was missing most
of the counts at the numerator, except the very tail.
I have now set a new time cut at 500-2500 channels, and committed the
change to the main repository. After the first recompilation of Monitor
the change will take place. I have also run Monitor on last night's runs
and put a summary in the following webpage
http://phpch.phys.cua.edu/~longhi/efficiency.html
Runs prior to number 10077 had no raw data banks in the data stream, so
I couldn't readily analyze them because the efficiency feature is, as of
today, implemented only in the raw data part of the package.
My apologies again to all those people who were mislead and confused by
the faulty Monitor output.
Regards
Alberto