Log entry time 17:49:12 on March 09, 2012

Entry number 365115

This entry is a followup to: 365063

As seen in the previous log, this morning Ryan and I thought we had fixed
the rate problem with T2.  However, after taking a cosmic run I found
that there was a significant number of events (~1%) that corresponded to
both T1 and T2; that is, DR.evtypebits==6.

I took a closer look at the strobe signal compared to the s1 "or" signal,
and found that the strobe was coming too late.  I also found that the
strobe was more narrow on the RHRS (~40ns) compared to the LHRS (~80ns).

This afternoon Vince and I took a detailed look at the timing of the LHRS
and found that the strobe signal came ~64ns after the logic signals of
s1, s2m and the Cherenkov.  On the RHRS we first moved the strobe signal
from a 754 logic unit to a 755 logic unit, which allowed for the signal
to be made ~80ns wide.  We also found that the delay which had been added
to the strobe signal this morning was much more than was necessary.  For
some reason the programmable delay module in slot 21 of crate 1 adds much
more delay than the module is actually programmed for.  We moved the
strobe to a different delay module and added an appropriate amount of
delay so that the strobe came between 30-80ns after the logic signals of
s1, s2m and the Cherenkov.

The T2 rate is now ~1-2 Hz for cosmics, comparable to what we found in
December. There is also no longer an overlap between T1 and T2.  We will
continue to monitor when the beam comes back.