Alberto Oliva

1
2008 AMS-02 Cosmic Muons TestA Tracker
Perspective
GSTTCSCalibrationEfficiencySignal
AnalysisHold Time Scan
2
Cosmic Rays Test

• 6 months of Tracker pre-integration data acquired
• 2 months Full Tracker GSTTCS
• 357 Calibrations
• 384 Tracker related Runs
• 274M Events (? exp. 30M tracks), 886 GB

Trigger rate test
n. Events Acquired

• Various tests performed
• hold time scan
• DSP code tests
• several thresholds
• dynamic pedestals
• Rotation test
• Acquisition in mix mode

TRD 72 h test
Nominal Conditions
Rotation test
Hold time test
3
Ground Support TTCS (I) The System
Water IN/OUT
hot spots
The bottle is always open (pressure const. p 60
bar)
Here is placed a restriction, giving a Dp PT02
PT01 5 bar
Pump IN/OUT
4
Ground Support TTCS (II) Temperature Monitoring
hot spots
Tracker OFF
14 May 2008
dT/dt gt 1C/h
entrance/exit

• Typical temperature path
• 1 hour of ramping temperature many calibrations
  and short runs
• 10 hours of nominal conditions 1 calibration
  for each hour

5
Typical Calibration
14 May 2008
Y-side
Noisy Strips Fraction
X-side
6
Calibration Stability
318 calibrations (from 23 April to 10 Jun)

• average dead channels fraction (DSP) 0.04
  (0.26) for Y (X) side
• average noisy channels fraction (DSP) 2.11
  (3.36) for Y (X) side
• average channel noise (ltsgt) 2.71 (3.31) for Y
  (X) side

7
Calibrations Dependence from Temperature

• Pedestals drift depends from temperature
  (electronics)
• Noise depends from temperature (silicon)

Thermal Chamber Measurement (2007)
Cosmic Rays Test (2008)
From WJB
8
Cluster Reconstruction Efficiency

• Tracks Reconstruction
• Development of new Tracker SW
• Alignment (next talk by S. Haino)
• Excluding the layer under analysis
• Minimum 6 hits per track
• An hit on the layer on the opposite side of the
  same plane is required
• Cut on c2 to avoid multiple scattering
• Sample used
• 10 M events of 14 May 2008
• 170 k tracks reconstructed

under analysis
9
Geometric Inefficiencies

• Out of ladders hits
• Out of sensor hits

edges effect
1 mm
Layer 2
1 mm

• Sensitive area is reduced to a fiducial area due
  to edges effect (1 mm)

10
Efficiency

• For a window of 500 mm
• Y-side eff. 96
• X-side eff. 95

Layer 2

• Sources of inefficiencies
• Dead/Noisy strips inefficiency, around both for
  X and Y
• TDR cluster reconstruction (cut on the
  Signal-over-Noise), estimated around few from
  the Landau profile

11
Ladder efficiency

• For a window of 500 mm

All Layers
12
Signal Analysis (I)

• Sample used
• 07-15 May 2008 the largest sample in stable
  condition
• 48 M events
• Reconstructed 5.6 M Tracks
• Preselection
• Reconstructed track
• Clusters attached to the track
• No one strip cluster
• Ladders 410, -807 are excluded

13
Signal Analysis (II) Inclination Signal Scaling
Each slice is fitted with Landau-Gaussian
convolution
0
45
A correction up to 30 of the ladder signal
14
Signal Analysis (III) Function of Impact Point
and Angle
Particle Trajectory (fit of n-1 hits)
Y-side
Impact Point
p-strips
Impact Angle
YZ Projection
X-side

• Inclination signal scaling
• each point is the MPV of a Landau-Gaussian fit
• A reasonable pattern is found

Impact Angle and Impact Point dependence is of
the order of 1.5 (1.5) for p (n) side
15
Signal Analysis (IV) VA Gain Spread

• Inclination scaling IA, IP corrections
• The 640 (384) channels signals are amplified 10
  (6) VA, each one amplifies 64 channels
• The signal of each VA was fitted using
  Landau-Gaussian convolution

Y-side
The 10 Y-side VAs
Hybrid
The 10 (6) VAs gain spread is 2.9 (4.5) for p
(n) side
16
Signal Analysis (V) Ladders Gains

• Inclination scaling, IA-IP and VA corrections
• Refitting of the ladders Landau distributions
• a Ladder gain global spread of 7 (4) is been
  achieved

Ladders have a uniform behaviour (without
corrections) at the 7
17
Signal Analysis (VI) Hit Correlation

• Applying all the correction a unique gain is been
  achieved

• Further developments
• Hit correlation applied in reconstruction (? See
  A. Oliva PhD Thesis)
• Probability density functions (pdf) for muons can
  be defined
• Track charge likelihood construction

18
Other Test (I) Ladders Shaping Time Test

• Scan between 1900-5100 ns
• 11 points with statistics approx. 2 M events per
  point

Test Beam 2003 (6 ladders mean)
CR 2008 (190 ladders mean)
Presence of an Extra-Delay?
19
Other Test (II) AMS Rotation
90
TRD
ECAL RICH
-30

• Second week of June

Tracks reconstructed rate
RUN taken during rotation
? See next presentation by S. Haino
20
2008 Tracker Cosmic Muons TestA Preliminary
Signal Analysis
21
Eta Distribution

• Presence of a track
• Clusters attached to the track
• Excluding 1 strip clusters
• Two higher strips
• p-side is indep. from qxz
• n-side is indep. from qyz

saturation
z
qxz
q
y
f
x
22
CofG3 Distribution
p-side

• Presence of a track
• Clusters attached to the track
• Excluding 1 strip clusters
• nstrip gt 3
• Three central strips

n-side
is not used
23
CofG4 Distribution
p-side

• Presence of a track
• Clusters attached to the track
• Excluding 1 strip clusters
• nstrip gt 4
• 4 higher strips

n-side
the best fourth strip is added
24
Number of Strips VS Impact Angle
Seed S/N gt 3.5 and Neig. S/N gt 2.
The 1 MIP signal is well described by 3 strips
clusters
25
IP VS Eta

• Presence of a track
• Clusters attached to the track
• Excluding 1 strip clusters
• Impact point is predicted by a nhits-1 fit
  extrapolation
• Only straight tracks projections are taken into
  account
• A correlation between IP and eta is seen

26
About the Cosmic Rays Test

• What full tracker does it mean?
• Ladder 410 bias not connected
• Ladder -807 K cable disconnected (photo)
• 2 ladders with a low K-gain (requires definition
  ? )
• TDR 5 crate 5 has problems
• Other problems
• Ladder -310, 303 very low K gain
• Ladder 614, 815 very low statistics, maybe a SW
  problem

27
Tracker SW DevelopmentToward an Affordable
Reconstruction