Main INDEX
Monthly INDEX
PREV
NEXT
User name R. Michaels
Log entry time 15:25:58 on August 7,2006
Entry number 178066
This entry is a followup to: 177995
keyword=helicity less sensitive to timing
I found one problem with the helicity code. It computes "fT0"
using a function TMath::Floor(x) where x is an expression that
involves calibration constants. Replacing x -> x+0.5 makes this far
less sensitive and the warnings occured only 3 times in 1 million
events of run 2273 (a high rate run with 25% DT). I investigated
one of those remaining episodes: It appears that the timestamp
data are ok but a helicity transition occured 12.9 msec too early.
That's a mystery, but if its only 3 in a million I guess it's ok.
We perhaps should assign helicity=0 to such mysterious episodes
but that's not what the code does now. It just issues a warning.
The conclusion is that the experiment will have fun analyzing the
data.