VELO Beamtest Trigger

1
VELO Beamtest Trigger
VELO meeting 08/12/2005 Jianchun (JC)
Wang Syracuse University
2
Event Category
Signal
Background

• The testbeam data will be used to align the
  telescope.
• The 450 GeV/c proton beam hits the detector
  sensor instead of the center of circle.
• We need interaction events, which benefit the
  alignment.

3
Scintillation Trigger
A
B
C
4
Simulation On Scintillation Trigger

• MC sample of 120 GeV/c p beam at (30, 10) mm (by
  Tomas Lastovicka).
• Scintillation counter B (veto) size20?20mm2, at
  2m downstream from first station.
• The beam is assumed to be perfect 0 in size, no
  dispersion. The scintillation efficiency is 100.
  
• The trigger efficiency is about 50. With 450
  GeV/c proton beam the efficiency will be less.
• The size of scintillation counter will be decided
  once we have more realistic beam condition. The
  rejection rate will be estimated then. The
  hardware will be provided by Syracuse group.
• Advantage hardware, at L0 trigger.
  Disadvantage very strict on beam condition.

5
Alternative Trigger

• The number of hits can be used in L1 trigger.
• Example Sum of maximum number of R hits per
  station and maximum number of f hits per station
  is required to be gt 20.
• Same MC sample used. The noise level is about
  0.1. All 21 stations are included. The
  efficiency is more than 50.
• With 450 GeV/c proton beam the effect will be
  more significant.
• We can also use hits per sector, or total number
  of hits of all station and so on.
• Advantage simple, not rely on alignment.
  Disadvantage strict on common mode noise.

6
Vertex Trigger

• Use vertex trigger, similar to that of official
  L1 trigger ( see Tomas presentation on
  07/22/05).
• The pattern recognition is complicated due to
  that the tracks are not from the designed beam
  axis. ? DR/DZ ? constant
• Tomas is working on the algorithm.
• I will implement the algorithm. Currently I am
  still on the learning curve of LHCb software.
• Advantage more close to final trigger The
  pattern recognition is needed anyway for
  analysis. Disadvantage CPU time,

!!!