An analysis of the orbit period
Date: Wed, 13 Oct 2004 12:04:41
From: Darin Acosta
Dear colleagues,
We have performed an analysis of the orbit period during last week's
structured beam test using the CSC Track-Finder at the H2 beam line. We
had installed special firmware developed by our engineer Lev Uvarov in the
Track-Finder to log any occurence of an orbit period that did not equal
924 BX. The bottom line is that when we use the "WE" signal for the start
of our experiment and analysis (the signal that comes after the RF
re-synch according to information provided to beam test users), we observe
NO cases of an anomalous orbit length. It is always 924 BX until the end
of spill signal.
Last June at X5A we did observe anomalous orbit lengths; however, it is
now suspected that we inadvertantly used the "WWE" signal to start the
experiment each spill, and this signal comes before the RF re-synch. To
see if we again observe anomalous orbit lengths using this start signal,
we repeated our measurements at H2 starting from the WWE signal. Indeed we
do find occurences of orbits not equal to 924 BX.
Attached is Fig. 1 with 2 plots on the
orbit analysis when triggered from WWE. The top plot shows the orbit # since
the WWE signal when a case of orbit length = 925 BX occurs. You can see
several spills. The bottom plot re-plots this data in terms of a normalized
orbit frequency (relative to the correct one) vs. time in sec from the WWE
signal. You can see that after approximately 700 ms since the WWE signal, no
anomalous orbits are observed.
Finally, the last two plots show
the BX structure for the triggers identified by the CSC Track-Finder (the
first plot is on a linear scale, Fig. 2a ,
the second, Fig. 2b is on log). You
see a nice 48 BX wide region, with some tails earlier and later. You also
see a very tiny fraction of random triggers consistent with the measured
cosmic rate of our trigger. The integrated time was 80 spills for these BX
plots.