After looking more into bremsstrahlung formulae I found that I could
improve my rate estimates for our upcoming CLAS run by using a better
bremsstrahlung yield curve for the few-GeV energy range. It is an
extreme relativistic formulation with Coulomb correction and
intermediate screening. The formula is given in Koch and Motz,
Rev. of Mod. Phys. 31 (59) 920, together with parameterizations of
various corrections given in Matthews and Owens, NIM 111 (73) 157.
This bremsstrahlung curve has been found to compare well with data at
3 GeV. Compared to the formula in Jackson, the main effect is to
reduce by 30% the rate of untagged low hadronic events, to increase
the rate of good tagged events in our highest energy prescale bin by
20%, and increase our overall rate of tagged events by 5% (for a fixed
total tagger rate). I think all of our estimates are now in very good
agreement.
At the meeting today I presented the prescale factors I suggested
before and showed the distribution of event rate as a function of
photon energy. There were no counter suggestions, so I think we have
a consensus.
>> Use the four Tagger sub-master (SM) divisions as they now are. <<
>> Prescale by factors of 1, 2, 4, and 8. from SM1 to SM4. <<
Note that these are prescale FACTORS, i.e. the numbers you want to
divide by in the DAQ.
______________________________________________________________________________
CLAS Real Photon Event Rate
______________________________________________________________________________
Endpoint energy = 2528.000 MeV
Total Unprescaled tag rate = 1.000E+07 photons/sec
Tagged energy range : 523.296 2409.184 MeV
Target thickness = 1.210 gm/cm2
Radiator thickness = 1.0E-04 rad. lengths
Integration step size = 100.000 MeV
Total hadronic cross section data: realsigma.num
Using Koch and Motz extreme-relativistic photon form
with Coulomb correction and intermediate screening
Electron Beam Current = 2.2 nanoAmps
Range k/E0 k_E0 Eg_lo Eg_hi Prescale Phot/sec Phot/sec Hadr/sec Hadr/sec
Min Max (MeV) (MeV) Factor (x1E6) Prescale Prescale
1) 0.060 0.207 152. 523. -1 10.08 0.00 1876 0
2) 0.207 0.331 523. 837. 8 3.35 0.42 541 67
3) 0.331 0.510 837. 1289. 4 2.80 0.70 424 106
4) 0.510 0.802 1289. 2027. 2 2.75 1.37 329 164
5) 0.802 0.953 2027. 2409. 1 1.06 1.06 106 106
6) 0.953 1.000 2409. 2528. -1 0.27 0.00 25 0
Tagged Photons (prescaled) = 3.55E+06 photons/sec
Total Photons (tagged & untagged) = 2.03E+07 photons/sec
Hadronic Trigger Rate (prescaled) = 445 events/sec
Untagged Hadronic Rate = 1901 events/sec
Total Hadronic Rate (unprescaled) = 3304 events/sec
Now the total hadronic rate is 3304 Hz, which is roughly comparable to
Elton's scaled estimate of 6 kHz, based on the e1 data of the CLAS
"Level 1" rate using the TOF counters. The prescaled hadronic trigger
rate comes out to be 445 Hz. This seems to be within the range of
present capabilities of the DAQ system, though we would have large
deadtime unless DAQ speeds up considerably. If the DAQ system were
running at the design value of 1.5 kHz we could relax the prescaling.
Elton's accidental rate calculation makes sense. Using Elton/Claude's
estimates for a Start counter rate (0.4 MHz), and the expected Tagger
rate (1x10^7Hz) with a 15 nsec coincidence time, gives and accidental
rate of 60 kHz coming out of the Tagger.Start counter coincidence.
Given the 10 kHz cosmic Level 1 rate for the TOF counters, plus my
estimate of about 3.3kHz for the real hardonic rate, and assuming a
100 nsec coincidence window in the Trigger Supervisor, results in an
accidental rate of 80 Hz. The ratio of real to accidental events in
the data we write to disk is then about 5 to 1, i.e. 17% junk on disk.
Regards, Reinhard.