Warm LC DAQ Tom Markiewicz SLAC

1
Warm LC DAQTom MarkiewiczSLAC

• LCWS Victoria
• 30 July 2004

2
Introduction Caveats

• I am not qualified to give this talk
• Version of this talk given in Cornell03 DAQ
• Assumptions
• No QSR backgrounds (dominant SLC background)
• No trigger problems
• 0.1 Hz trigger in 1995 lead to 10 deadtime
• Dead-timeless 120 Hz trigger in 2015
• IP Backgrounds dominate data load
• Muon backgrounds not yet included
• At the appropriate time real DAQ experts will
  design system
• Channel counts, resolution, range, mean
  occupancy, occupancy fluctuations, buffer sizes
• Smart readout (front-end intelligence)

3
Snowmass 2001 LD Background Occupancies _at_ 500 GeV
for LD
4
Detector Occupanciesfrom ee- Pairs _at_ 500
GeVfcn(bunch structure, integration time)
TESLA
NLC
5
LD Data Rates from ee- Pairs _at_ 500 GeV presented
at Cornell03 ALCPG DAQ Session
BytesHits192 120
6
Todays Talk

• Update expected front-end data load
• SiD Detector
• Hits based on Toshi Abes GEANT simulations
• Beamstrahlung photon interactions producing
• ee- incoherent pairs
• Hadrons
• mm-
• Assume one Z H event per train crossing

7
ee- Pairs _at_ 500 GeV
8
Hadronic 2-photon events at NLC/GLC
9
Detector Occupancies
Study by T. Abe
10
Tracking Detector Hits per Train(T. Abe)
11
Calorimeter Hits per Train (T. Abe)
12
Muon Hits per Train (T. Abe)
13
Bytes/hit per Detector Brute Force
14
Silicon Detector Data Load
15
Conclusions

• Largest change with respect to the exceedingly
  crude 2003 estimate is TPC vs. Si-Tracker and
  dumb readout assumption of 100 bytes/hit
• Total front end data load for SiD seems modest,
  even by SLD standard
• Wire systems (CDC, CRID) dominated SLD data
• Would like to better understand integration of
  machine info with detector