Log entry time 20:32:42 on February 20, 2011

Entry number 347854

Took a controlled access to check the trigger timing at the TS.  There was
some evidence in data that T3 was stealing the time of T5 at the TS.  
However when I checked the timing with the EDTM, I found that

T5 came before T1 by about 22 ns
T5 came before T3 by about 14 ns

This seemed fine.  I added about 6 ns of delay to T3 before the TS in case
the real T3 has some time jitter.  So now T5 comes before T3 by about 20
ns at least in the EDTM world.