Difference between revisions of "OWG meeting 11-Apr-2008"
From GlueXWiki
m |
m |
||
| Line 43: | Line 43: | ||
'''Event Size Study''' - Alex S | '''Event Size Study''' - Alex S | ||
| − | * Dave L out of town for this meeting, everything needs to be compared to his original results | + | * see Alex's slides via link in agenda. |
| − | * studied rates and occupancies in all detectors | + | * Dave L out of town for this meeting, everything needs to be compared to his original results. |
| − | * included EM background (100 ns gate) from Geant3 | + | * studied rates and occupancies in all detectors. |
| − | * included hadronic signal and background from Pythia (Eugene's version) | + | * included EM background (100 ns gate) from Geant3. |
| − | * used simple trigger cuts (e.g. FCAL E>0. | + | * included hadronic signal and background from Pythia (Eugene's version). |
| − | * 400 kHz hadronic rate at 10**8 photons/sec, L1 divide by 2 to 200 kHz, L3 divide by 10 to 20 kHz | + | * used simple trigger cuts (e.g. FCAL E>0.4 GeV), will redo using strategy from CNU Genetic Algorithm study. |
| − | * EM supplies major background, more than hadronic | + | * 400 kHz hadronic rate at 10**8 photons/sec, L1 divide by 2 to 200 kHz, L3 divide by 10 to 20 kHz. |
| − | * advantage to tighten gates used in trigger to minimize EM background (Alex used 100 ns) | + | * EM supplies major background, more than hadronic. |
| − | * 10**8 tagged photons corresponds to about 3 10**8 in tagger, collimator removes about factor of 3 | + | * advantage to tighten gates used in trigger to minimize EM background (Alex used 100 ns). |
| − | * | + | * 10**8 tagged photons corresponds to about 3 10**8 in tagger, collimator removes about factor of 3. |
| + | * average number of hits in each detector in rough agreement with Dave L's results. | ||
| + | * most hits in FDC, 475 per event, of which 269 is due to EM background. Cuts are photon energy > 8 Gev, FCAL energy > 0.4 GeV | ||
| + | * most detectors hit count dominated by EM. | ||
| + | * some questions about photon cutoffs in simulation need to be resolved. | ||
| + | * 100 ns may be too large a gate width. | ||
| + | * optimization needed: tighter gates may allow looser trigger, since events cleaned up | ||
| + | * Alex will report again at the next meeting, will include summary tables, use CNU trigger algorithm, etc. | ||
| + | * need to estimate bytes-per-hit for each detector to get final average event size. | ||
Revision as of 12:19, 14 April 2008
Contents
Agenda
- Announcements
- Review of minutes from 4-Apr-2008 meeting
- Event size study - Alex S.
- Pre-brief outlines for May 2008 Subsystem Review - many people
- Progress towards subsystem review - many people
Time/Location
2pm Fri 11-Apr-2008 CC F326
Announcements
Subsystem review likely Wed,Thurs 7,8-May.
Elke's Deadlines
- Final presentations to Elke: 5-May close-of-business
- Pre-brief material to reviewers: 30-April
- Pre-brief material to Hall-D group to read: 23-April
- Pre-brief outlines: 11-April
Next Meeting
2pm Fri 18-Apr-2008 CC F326
New Action Items from this Meeting
Minutes
Attendees: Elliott W, Simon T, Elton S, Alex S, Fernando B, Mark I, Dave D, Chris C, Jim S, Eugene C.
Admin has approved two new sorely-needed hires for the JLab DAQ group. Both will focus on hardware. One will work with Dave A, the other will work with Ed J.
Event Size Study - Alex S
- see Alex's slides via link in agenda.
- Dave L out of town for this meeting, everything needs to be compared to his original results.
- studied rates and occupancies in all detectors.
- included EM background (100 ns gate) from Geant3.
- included hadronic signal and background from Pythia (Eugene's version).
- used simple trigger cuts (e.g. FCAL E>0.4 GeV), will redo using strategy from CNU Genetic Algorithm study.
- 400 kHz hadronic rate at 10**8 photons/sec, L1 divide by 2 to 200 kHz, L3 divide by 10 to 20 kHz.
- EM supplies major background, more than hadronic.
- advantage to tighten gates used in trigger to minimize EM background (Alex used 100 ns).
- 10**8 tagged photons corresponds to about 3 10**8 in tagger, collimator removes about factor of 3.
- average number of hits in each detector in rough agreement with Dave L's results.
- most hits in FDC, 475 per event, of which 269 is due to EM background. Cuts are photon energy > 8 Gev, FCAL energy > 0.4 GeV
- most detectors hit count dominated by EM.
- some questions about photon cutoffs in simulation need to be resolved.
- 100 ns may be too large a gate width.
- optimization needed: tighter gates may allow looser trigger, since events cleaned up
- Alex will report again at the next meeting, will include summary tables, use CNU trigger algorithm, etc.
- need to estimate bytes-per-hit for each detector to get final average event size.
