Log entry time 10:31:47 on March26,2005

Entry number 142464

Followups:

142465 03/26/05 10:48 R. Michaels Efficiency of R-arm too

keyword=Left arm scint. efficiency

The T4 rate is slightly higher than the T3 rate. Since T4 measures
efficiency, usually and to 1st order, this raises a red flag. I'm
not aware if anyone checked it, so here it goes. The "bottom line"
is that the scintillator efficiency is acceptably high (98-99%) and
for some reason a lot of T4 are junk. Note, the T1/T2 picture is
different.



Using a randomly chosen production run (2948)...



Examine the trigger efficiency for Left Arm using a plot like this:

===================================================================

T->Draw("L.vdc.u1.nhit+L.vdc.v1.nhit+L.vdc.u2.nhit+L.vdc.v2.nhit","fEvtHdr.fEvtType==3&&(L.vdc.u1.nhit+L.vdc.v1.nhit+L.vdc.u2.nhit+L.vdc.v2.nhit)>0");



And requiring either event type type 3 or 4, and judging if the event is "good" or "bad" from the sum of VDC hits >0  or  ==0  respectively.



From Raw Scaler Rates:

T3 raw rate = 174 Hz

T4 raw rate = 194 Hz   (--> implies naive inefficiency 53% !!)



  Fraction of T3 which are good = 23665 / (3518+23665) = 87%

  Fraction of T4 which are good = 70 / (70+2327)  = 2.9 %



  Estimated Trigger Inefficiency = (0.029 * 194) / (0.87 * 174) = 3.7 %

  ==> Single scintillator plane Inefficiency = 1.8 %

                                           



This is about a factor of 2 higher than experience.  But one can do the same cleanup using L.tr.n (# of tracks).  Result is



  Fraction good T3 = 84 %  (have >= 1 track)

  Fraction good T4 = 1.5% 

  Trigger Ineff = 2.0%

  Single-plane Ineffic = 1.0%

                     

A  1% single-plane inefficiency is about what I expect.

User name R. Michaels

Log entry time 10:31:47 on March26,2005

Entry number 142464

Followups: