Main INDEX
Monthly INDEX
PREV
NEXT
User name R. Michaels
Log entry time 15:52:27 on September 17, 2006
Entry number 184292
keyword=T3 is clean, definitions of T4,T6
I made an halog about T4 earlier today. Here I add that: 1) Albert & Bogdan replaced S1
PMTs on L-HRS and it has no after-pulsing now. S2 still double pulses. The trigger T3
is, predictably, quite clean: no after-pulsing. 2) The MLU files for T4 and T6 were re-made
as follows. Let a "1" mean a bit is required, a "0" mean a bit must be absent, and "x" means
the bit may be a 0 or 1. The MLU has inputs (1-5): S1,S2,S0A,S0B,Cer where S0 has 2 PMTs
S0A and S0B. Then T4 is defined as the bitwise OR of 1xx01 and 1xx10 for input 54321.
And T6 is defined as x1111.
That is, T4 = ((Cer.and.S1).or.(Cer.and.S2).and.NOT.(S1.and.S2) -- with no regard to S0.
and T6 = (S0.and.S1.and.S2).
The trigsetup still asks if S0 is in place, but that is only for record-keeping now.
What was wrong since Aug 21 was that two obscure cases polluted T4 due to logical
flaws in the MLU file ("new" on Aug 21):
1) Cer.and.S0A.and.S0B.and.NOT.S1.and.NOT.S2
2) S1.and.S2.and.Cer.and.S0A.and.NOT.S0B
In addition there were 3 cases missing in which either S0A or S0B (but not both) are 0.
These logical flaws probably did not affect T4 much because they probably didn't happen.
In any case, we re-made the MLU file using tried & true code that does not miss any cases.
Finally I remind that T3 does NOT use an MLU (whew!). It is just a trivial overlap S1.and.S2.
A copy of this log entry has been emailed to: doug@jlab.org,gilman@jlab.org,mpaolone@jlab.org,camsonne@jlab.org