MDC4 SVT

1
MDC4 - SVT
Most of the work shown here done by Lilian
Martin, Javier Castillo, Boris Hippolyte, Selemon
Bekele
2
SVT Installed
STAR detector gets new silicon heart
CERN Courier
Radii 5,10 15 cm Length 12.4 cm 18.6 cm
21.7cm (-1 lt h lt 1)
91 live 6 of the detector read out if set noise
level to 4 mV
3
In MDC4
Ran Fast Simulator Private Vertex Finder
(grouping technique) Global tracking
(est) Also have raw data reader,cluster-finder,
cluster fitter, alignment code
4
Slow Simulator

• Nearly ready to rock and roll
• Includes
• affects of coulomb repulasion and diffusion
  during drift of electrons
• pasa response and shaping (causes undershoot in
  our signals)

5
Clustering
Anode direction
Drift time
gt95 Efficient with expected signal/noise
Drift direction
Position Resolution lt25 microns Systematic
shift in drift direction due to pasa response
6
Principle of EST

• Take TPC tracks
• Using primary vertex location refit tracks
• Project to SVT layers
• Use several loops sorting by pt to match tracks
  to SVT hits increase search area each time.
• First insist on one hit on each layer then 2 hits
  total
• Using unmatched tracks search again not using
  primary vertex

IMPORTANT SVT DOES NOT CREATE TRACKS
7
Tracking
Impact parameter for global TPC only tracks
Better Discrimination!
Log10Impact
Impact (cm)
Impact parameter for global TPCSVT tracks
Log10Impact
Impact (cm)
8
Primaries
Primary tracking efficiency 3 hits in the SVT
and gt5 in the TPC Efficiency 80 Purity 90
1.5
1.5
-1.5
9
Primaries
Primary tracking efficiency 2 hits in the SVT
and gt5 in the TPC
pt
pt
0
2
0
2
Efficiency 70 Purity 85
h
1.5
-1.5
h
1.5
-1.5
10
Secondaries
Secondary leaves 3 hits in SVT and gt5hits in TPC
pt
Efficiency 40 Purity 90
h
11
Own Primary Vertex Finder
Resolution 65 microns
Resolution 350 microns
Good way to check alignment between SVT and TPC
as function of time
12
Primary Vertex
Resolution 60 microns in x,y , 100 in z
(c.f 100 and 150 for year 1)
13
Primary Vertex
Wiggle gone Except possibly at edges where SVT
isnt
(Mind you it never used to be there it has
developed over the past year. Still not clear
what causes it.)
14
Strangeness
Maximum finding efficiency 3.5 Find 12 of
those after cuts
Find a good signal
X
15
Strangeness (2)
Efficiency 0.9 - BEFORE CUTS!! Efficiency of
the cuts tuned for year 1 is 3 of this 0.9
W
Associated W Mean pt 1.8 GeV Simulated with a
mean pt of 1 GeV
16
Strangeness (3)
X
With at least one track from the SVT
TPC Only
Secondary Vertex Resolution improves
17
Strangeness (4)
Cut variables change so can cut tighter to
improve signal to noise and efficiency
18
Conclusion
Worked better than expected. All the vital
pieces of code are in place Lots of tuning to
do