Main INDEX
Monthly INDEX
PREV
NEXT
Make New Entry,
Make Followup Entry
User name Bob/Youcai
Log entry time 01:00:11 on November 30, 2008
Entry number 251382
This entry is a followup to: 251164
Followups:
keyword=hit rejection on RICH
Youcai and I have been looking at some recent RICH data.
The number of data rejected in recent runs is, by category
1. Too many hits: rejects 7 to 9% (it was higher for the higher HV)
2. Timeouts -- a negligible loss (ca 0.02%)
At present we're rejecting around 7% of data. Note, it is rejected
by the DAQ --> lost forever.
Before the T/H scan we were losing about 1%. Youcai thinks this is
because we have more noise now. Bob (a non-expert!) thinks it's
hard to believe that the "noise" depends on T/H.
If the experiment is worried about this we might:
Try a few runs with the cuts turned off but the data flagged.
This would tell us:
a) The trade-off of deadtime increase (we already know it will go up,
Youcai thinks it will go up a huge amount).
b) The nature of the rejected events -- are they even analyzable ?
(if they are very noisy they may not be).
No strong opinions from us, but the experts should be informed.
A copy of this log entry has been emailed to: camsonne@jlab.org,iodice@jlab.org