ATLAS Silicon Pixel and Strip Detector: Commissioning ... - MPG HLL

ATLAS Silicon Pixel and
Strip Detector:
Commissioning
with Cosmic Data
Maria Fiascaris (University of Oxford)
on behalf of the ATLAS Collaboration
11 th European Symposium on Semiconductor Detectors
June 7 th 2009

The ATLAS Detector at the LHC
LHC:
14 TeV p­p collisions
25 ns Bunch Crossing Time
L = 10 34 cm ­2 s ­1
ATLAS (General Purpose Detector)
➢ Fast and radiation hard electronics and
sensor elements
➢ High Granularity
Inner Detector
(|| < 2.5, p T
> 0.5 GeV):
➢ Hermetic and robust pattern recognition
➢ Excellent momentum resolution
➢ Primary / secondary vertex measurement
07/06/2009 M. Fiascaris
2

The ATLAS Inner Detector
In 2T solenoidal field:
➢ Pixels
➢ Silicon Strip Detector (SCT)
➢ Transition Radiation Tracker
➢ Radiation damage:
Si sensors cooled to ­5 to ­10 C
(C 3
F 8
)
Pixel Detector
➢ Barrel Region : 3 layers
➢ End­caps: 3 disks per side
➢ 3 space points
➢ 80.4 M read­out channels
SCT
➢ Barrel Region : 4 layers
➢ End­caps: 9 disks per side
➢ 4 space points
➢ 6.3 M read­out channels
➢ 63 m 2 Silicon Area
07/06/2009 M. Fiascaris
3

Pixel Modules
Pixel Module:
➢ Active surface: 6.08 x 1.64 cm 2
➢ 16 FE chips bump bonded to sensor
➢ Flex hybrid with Module Control Chip
Pixel Sensors:
➢ n­in­n technology
(optimal radiation tolerance)
➢ Bulk depth 250 m
➢ Typical pixel size: 50m x 400m
➢ Accuracy: 10 m (R) , 115 m (z, R)
➢ Radiation tolerance:
F neq
~ 1.1 x 10 15 cm ­2 (10 years of operations)
➢ Inner vertexing layer to be replaced
after 3 years of operations
07/06/2009 M. Fiascaris
4

SCT Modules
SCT Module:
Barrel Module
➢ 2 planes of Si (2 sensors each)
➢ Stereo angle of 40 mrad
➢ Barrel: identical modules
➢ End­caps: 3 different designs
SCT Sensor:
➢ Single sided p­in­n technology
➢ AC­coupled readout strips
➢ Bulk depth 285 m
End­cap Modules
➢ Typical strip pitch 80m
➢ Accuracy (from 2 wafers):
17 m (R) , 580 m (z , R)
➢ Radiation tolerance:
F neq
~ 2 x 10 14 cm ­2 (10 years of operations)
07/06/2009 M. Fiascaris
5

Read-Out System
Preamplifier + Discriminator
Bump­bond
pad
Address
Time stamp
ToT
End of Column
Buffer
L1
Trigger
MCC
Pixels
➢ 16 FE chips per module
➢ 0.25 m CMOS technology
➢ Digital read­out of 2880 pixel cells
➢ End­of­column readout buffers
SCT
➢ 12 ABCD3T chips per module
➢ DMILL technology
➢ Binary read­out of 128 channels
➢ 132­length binary pipeline
➢ Data compression, read­out
Detector
Amplifier
Shaper
Comparator
Binary Pipeline
Output
Buffer
3 bins read­out
centered on L1 Trigger
Data
Compression
& Readout
Optical links
➢ Data from on­detector to off­detector electronics
➢ Trigger, timing, configuration data to the modules
07/06/2009 M. Fiascaris
6

Installation & Commissioning
Aug 2006
Apr 2007
Jun 2007
SCT Barrel inserted in cavern
SCT Barrel signed­off
SCT End­caps inserted in cavern
Pixels installation
Installation of Pixel Detector
Feb 2008
Sep 2008
Dec 2008
SCT End­caps signed­off
Pixels optical, elect. connections
First beam in ATLAS
Calibrations and cosmic runs
Installation of ID Barrel
May 2009
Autumn
2009
Operations resumed:
Calibrations, combined runs
Beam expected again
→ Full characterization of Pixels and SCT
prior to first beam in 2009
07/06/2009 M. Fiascaris
7

Services and Connectivity
Evaporative Cooling System:
➢ Major problems encountered in 2007­2008 → fully operational from August 2008
➢ Non­operational loops in autumn 2008:
➢ 3 pixel loops (recoverable)
➢ 2 SCT End­cap loops due to a damaged cable (recovered, 23 modules)
a coolant leak (non recoverable, 13 modules)
➢ Upgrade in first months of 2009
Optical Links
➢ Test of optical connections and tuning of parameters
➢ Elevated mortality rate of TX plug­ins on off­detector boards observed
➢ Evidence of electrostatic discharge during manufacture
➢ TX plug­ins replacement with new production ongoing
→ to be completed by August 2009
Power Supplies and Measurement of Environmental Variables
➢ Electrical connections and power supplies
➢ Detector Control System:
monitoring power supplies and environmental variables
➢ Fully commissioned in 2008
07/06/2009 M. Fiascaris
8

Commissioning: Noise Occupancy
Pixels Noise Occupancy
SCT Noise Occupancy
Specification:
5 x 10 ­4
➢ After masking of noisy pixels
➢ Noise occupancy ~ 10 ­10
➢ SCT binary read­out:
threshold calibration crucial !
➢ Nominal threshold: 1fC
➢ Noise ~ 10 ­5 , well below specifications:
noise occupancy < 5 x 10 ­4
efficiency > 99%
07/06/2009 M. Fiascaris
9

Cosmics Data Taking
➢ Number of tracks collected with
magnetic field on/off:
Field off Field On Total
All tracks 4.94 M 2.67 M 7.61 M
SCT tracks 1.15 M 0.88 M 2.03 M
Pixel tracks 0.23 M 0.19 M 0.42 M
Track through Pixels and SCT (field on)
➢ Fraction of operational modules:
Pixels*: 96%
*About 2% due to unavailable cooling loops,
recoverable
SCT Barrel: 99%
SCT EndCaps**: 97%
** Almost 2% due to unavailable cooling loops
Expect > 99% operational in 2009
➢ Cosmic runs used for:
● Timing in
● Performance studies
● Alignment
07/06/2009 M. Fiascaris
10

Intrinsic Hit Efficiency
Pixel Barrel
Intrinsics hit efficiency from cosmic run:
→ only active modules considered
→ aligned geometry
SCT Barrel
➢ Hit efficiency per layer ~ 99.8%
07/06/2009 M. Fiascaris
➢ Barrel hit efficiency per layer ≥ 99.6%
➢ 4 layers with 2 sides
➢ Incidence angle > 40 degrees
➢ End­caps also studied:
efficiency ~ 99%
11

Timing-In the ID
Pixels
➢ During cosmic data­taking:
read­out 8 bunch­crossings (BC)
➢ Plans: 5 BC (initial collisions)
1 BC (LHC designed luminosity)
➢ Cluster arrival time:
all modules timed­in within ± 1 BC
SCT
➢ Reads out 3 BC
➢ SCT only track segments in barrel
➢ Well timed­in (within 1 BC)
➢ For LHC collisions:
further improvements needed
07/06/2009 M. Fiascaris
12

Lorentz Angle
➢ Important to understand the detector:
alignment, spatial resolution
➢ Dependence on V bias
, temp. , external B field :
➢ Measured studying cluster size vs incidence angle
L
Lorentz angle (SCT barrel):
good agreement data ­ MC!
Data:
L
= 3.93±0.03 (stat.)±0.10 (syst.)
Monte Carlo:
L
= 3.69±0.19 (syst.)
Minimum ~ 0
(B field off)
→ Lorentz Angle measured also with Pixel Detector: similar results
07/06/2009 M. Fiascaris
13

Inner Detector Alignment
➢ Barrel region aligned down to module level
➢ End­caps: large structures aligned (too low statistics for module­to­module alignment)
Residual distribution for x­ coordinate in Pixels and SCT
→ Great improvement in residuals: Pixels 128m 24 m , SCT 123m 30 m
→ Resolution approaches the MC simulation of a fully aligned detector
07/06/2009 M. Fiascaris
14

Conclusions
Great progress in the commissioning of ATLAS Pixel and SCT Detectors in 2008
Work to improve the detector operational reliability (evaporative cooling system
and optical links):
upgrade on cooling system in 2008/2009 shut­down period
TX replacement on­going (until August)
Autumn 2008: long period of calibrations and cosmic runs
96% of Pixels and 99% (97%) of the SCT Barrel (End­caps) were operational
Performance (hit efficiency and noise occupancy) within specifications
Cosmic Data:
High statistics (2M tracks in SCT, 400k in Pixels)
Time­in with ATLAS, Lorentz angle measurement
Great improvement in alignment in barrel region
Expect > 98% of Pixels and > 99% of SCT to be operational in 2009
Recently (end of May) resumed operations:
calibrations and cosmic runs over Summer
Excellent shape, waiting for beam in autumn 2009!
07/06/2009 M. Fiascaris
15

BACK­UP SLIDES

First Beam in ATLAS
➢ Pixel Detector off (safety mode)
➢ SCT Barrel off
collimator
➢ SCT End­caps on, for safety:
● low voltage (20 V)
● raised threshold (1.2 fC)
07/06/2009 M. Fiascaris
17