User name R. Michaels
Log entry time 14:39:06 on April17,2008
Entry number 232718
keyword=re: online deadtime.The following illustrates our problem with online deadtime
where we only know L1A (accepted trigger rate) and rates
of raw triggers.
This is for production run 3276 where the rates are
L1A = 1120 Hz (obtained by zooming on Y axis in fig 1)
T5 = 710 Hz
T6 = 1090
The other triggers don't matter, to 1st order.
If we assume, as datamon does, that "every T6 is a T5" then
we get a deadtime of 3%. This is what the halog end-run
script shows too.
If, on the other hand, we assume T5 and T6 are completely
uncorrelated the deadtime is 38 % !! (a factor of 13 higher).
The reality is somewhere in between and we need offline calculation.
FIGURE 1
