NSS 2004

1
The Silicon Strip Tracker of the GLAST Large Area
Telescope L. Baldini INFN - Pisa Nuclear
Science Symposium 2004 Rome October 18, 2004
2
Outline
GLAST Burst Monitor (GBM) 10 kev 25 MeV
Large Area Telescope (LAT) 20 Mev 300 GeV
Launch Vehicle Delta II 2920-10H Launch
Location Kennedy Space Center Orbit Altitude 575
Km Orbit Inclination 28.5 degrees Orbit Period 95
Minutes Launch Date February 2007

• Talk overview
• The science case for GLAST.
• Instrument design.
• Silicon Tracker construction.
• Conclusions.

3
A brief history of the g-ray astronomy...

• SAS-2 (1972 - 1973)
• Energy range 30 MeV 1GeV
• Energy resolution 100
• Peak effective area 100 cm2
• Field of view 0.25 sr

• EGRET (1991 - 1996)
• Energy range 20 MeV 30GeV
• Energy resolution 15
• Peak effective area 1500 cm2
• Field of view 0.5 sr

1970
1980
1990
2000
Time

• COS-B (1975 - 1982)
• Energy range 30 MeV 5GeV
• Energy resolution 40
• Effective area 70 cm2
• Field of view 0.25 sr

Balloon flights, Small satellites ( 621 photons
above 50 MeV detected by OSO-3!)
4
The need for a high-energy g-ray detector

• Broad spectral coverage is crucial for
  understanding most astrophysical sources.
• Multiwavelenght campaigns space based and
  ground based experiments cover complimentary
  energy ranges.
• The improved sensitivity of GLAST will match the
  sensitivity of the next generation of ground
  based detectors filling the energy gap in between
  the two approaches.
• Overlap for the brighter sources cross
  calibration, alerts.

• Predicted sensitivity to point sources
• EGRET, GLAST and MILAGRO 1 year survey.
• Cherenkov telescopes 50 hours observation.
• (from Weekes, et al. 1996 GLAST added)

5
The sky above 100 MeV the EGRET survey

• The heritage of EGRET
• Diffuse extra-galactic background ( 1.5 x 10-5
  cm-2s-1sr-1 integral flux).
• Much larger ( 100 times) background on the
  galactic plane.
• Few hundreds of point sources (both galactic and
  high latitude).
• Essential characteristic variability in time.

6
Point source sensitivity and sky-map

• 3rd EGRET catalog
• 271 point sources found.
• Based on 5 years of data.
• GLAST 1 year sky survey
• Will discover thousands of new sources (based on
  the extrapolation of the number of sources vs.
  integral flux).
• Key points large FOV and effective area.

Integral Flux (Egt100 MeV) cm-2s-1
7
Unidentified sources

• 170 point sources of the EGRET catalog still
  unidentified (no know counterpart at other
  wavelengths).
• GLAST will provide much smaller error bars on
  sources location (at arc-minute level).
• GLAST will be able to detect typical signatures
  (spectral features, flares, pulsation) allowing
  an easier identification with know sources.
• Most of the EGRET diffuse background will be
  resolved into point sources.
• Large effective area and good angular resolution
  are crucial!

Cygnus region 15o x 15o, E gt 1 GeV
Counting stats not included.
8
Active Galactic Nuclei

• AGNs phenomenology
• Vast amount of energy from a very compact
  central volume.
• Large fluctuations in the luminosity (with
  hour timescale).
• Energetic, highly collimated, relativistic
  particle jets
• Prevailing idea accretion onto super-massive
  black holes (106 1010 solar masses).

• AGN physics to-do-list
• Catalogue AGN classes with a large data samples
  (at least 3000 new AGNs)
• Detailed study of the high energy spectral
  behavior.
• Track flares (t minutes).
• Large effective area and excellent spectral
  capabilities needed!

9
Gamma Ray Bursts

• GRBs phenomenology
• Dramatic variations in the light curve on a very
  short time scale.
• Isotropic distribution in the sky (basically
  from BATSE, on board CGRO, but little data _at_
  energies gt 50 MeV).
• Non repeating (as far as we can tell).
• Spectacular energies ( 1051 1052 erg).

• GRBs physics
• GLAST should detect 200 GRBs per year above
  100 MeV (a good fraction of them localized to
  better than 10 in real time).
• The LAT will study the GeV energy range.
• A separate instrument on the spacecraft (the
  GBM) will cover the 10 keV 25 MeV energy range.
• Short dead time crucial!

• Simulated 1 year GLAST operation
• (Assuming a various spectral index/flux.)

10
Requirements on the instrument

• High sensitivity, pointing accuracy
• Large effective area.
• Large field of view.
• Good angular resolution (Point Spread Function).
• Good spectral capabilities
• Good energy resolution.
• Wide energy range.
• Fast alert
• Short instrumental dead time.
• Long observation time
• No consumables.
• Requirements connected with the operation in
  space
• Modularity, robustness, redundancy.
• Severe power and mass budgets.

11
Experimental technique

• The instrument must measure the direction, energy
  and arrival time of high energy photons (from
  approximately 20 MeV up to 300 GeV).
• Pair production is the dominating interaction
  process for photons in the GLAST energy range.
• e e- pair provides the information about the g
  direction/energy.
• e e- pair provides a clear signature for
  background rejection.

• Pair conversion telescope
• Tracker/converter (detection planes high Z
  foils) photon conversion and reconstruction of
  electron/positron tracks.
• Calorimeter energy measurements.
• Anti-coincidence shield background rejection
  (cosmic rays flux typically 104 higher than g
  flux).

12
Design drivers

• Science drivers
• Background rejection
• Drives the ACD design.
• Also impact on TKR/CAL design.
• Effective area and PSF
• Drive the converter thicknesses and layout.
• PSF also drives sensor performance, layers
  spacing and overall tracker design
• Energy range/resolution
• Drive the thickness/design of the calorimeter.
• Field of view
• Basically sets the aspect ratio (width/height).

• Mission drivers
• Allocated space on the launcher
• Forces the maximum possible lateral dimension
  (and geometric area).
• 1.8 m for GLAST.
• Power budget
• Restricts the number of readout channels in the
  tracker (i.e. strip pitch, number of layers).
• 650 W for GLAST.
• Mass budget
• Basically bounds the total depth of the
  calorimeter.
• 3000 kg for GLAST

TKR
TKR
CAL
CAL
13
Overview of the Large Area Telescope

• Overall modular design
• 4x4 array of identical towers - each one
  including a Tracker, a Calorimeter and an
  Electronics Module.
• Surrounded by an Anti-Coincidence shield.

• Anti-Coincidence (ACD)
• Segmented (89 tiles).
• Self-veto _at_ high energy limited.
• 0.9997 detection efficiency (overall).

• Tracker/Converter (TKR)
• Silicon strip detectors.
• W conversion foils.
• 80 m2 of silicon (total).
• 106 electronics chans.
• High precision tracking, small dead time.

• Calorimeter (CAL)
• 1536 CsI crystals.
• 8.5 radiation lengths.
• Hodoscopic.
• Shower profile reconstruction (leakage
  correction)

Tower DAQ (TEM)
14
Triggering and On-board Data Flow

• Level 1 trigger
• Hardware trigger, single-tower level.
• Three_in_a_row three consecutive tracker x-y
  planes in a row fired. Workhorse g trigger.
• CAL_LO single log with E gt 100 MeV
  (adjustable). Independent check on TKR trigger.
• CAL_HI single log with E gt 1 GeV (adjustable).
  Disengage the use of the ACD.
• Cosmic rays in the L1T! 13 kHz peak rate.
• Upon a L1T the LAT is read out within 20 ms.

• On-board processing
• Identify g candidates and reduce the data
  volume.
• Full instrument information available to the
  on-board processor.
• Use simple and robust quantities.
• Hierarchical process (first make the simple
  selections requiring little CPU and data
  unpacking).

x
x
x

• Level 3 trigger
• Final L3T rate 30 Hz on average.
• Expected average g rate few Hz
• (g rate cosmic rays rate 1 few).
• On-board science analysis (flares, bursts).
• Data transfer to the spacecraft.

15
GLAST vs. EGRET

• 1After background rejection.
• 2Single photon, 68 containment, on axis.
• 31s, on axis.
• 41s radius, high latitude source with 10-7
  cm-2s-1 integral flux above 100 MeV.
• 51 year sky survey, high latitude, above 100 MeV.

16
A view on the Tracker

• Tracker/converter design determines the PSF
• Low energy PSF completely dominated by multiple
  scattering effects ( 1/E).
• High energy PSF set by hit resolution/lever arm.
• Converter foils layout/detectors design/layers
  spacing determine the rollover energy and the
  asymptotic value of the PSF _at_ high energy.

• Final design
• 19 trays structures providing the basic
  mechanical framework.
• 18 x-y detection planes immediately following W
  converter foils.
• Front section 12 trays with 3 X0 converter.
  Excellent PSF.
• Back section 4 trays with 18 X0 converter.
  Increase effective area.
• Bottom 3 trays with no converter at all.

• Compact and aggressive mechanics
• Less than 3 mm spacing between x and y layers.
• Front end electronics on the four sides of the
  trays (90º pitch adapters).
• Few mm gaps between adjacent towers.

17
The Tracker electronics system

• Key features
• Low power consumption
• ( 200 mW/channel).
• Low noise occupancy
• (lt 10-4 at the single channel level, mainly for
  the trigger).
• Redundant architecture.
• Complete digital zero suppression onboard.

• Overall design
• 24 custom Front End chips 2 Readout Controller
  chips handle one silicon layers.
• Data can shift left/right to one of the
  Controllers (single dead chips can be bypassed
  without loosing data).
• Zero suppression takes place in the controllers
  (data stream includes IDs of the strips fired
  TOT of the layer OR, which also provides the
  trigger).
• Two flat cables per side complete the redundancy.

18
Tracker construction work flow

• SSD procurement and testing

• Ladders assembly

• Towers assembly

18
10,368
342
2592

• Trays assembly and test

648

• Panels fabrication

• Readout electronics fabrication, test and burn-in

19
Silicon Strip Detectors procurement/testing

• Standard design
• 6 wafers size.
• 8.95 x 8.95 cm detectors.
• 400 mm thickness.
• 384 strips, 228 mm pitch.
• Reliable, small rate of defects.

• Aggressive specifications
• Depletion voltage lt 120 V.
• Breakdown voltage gt 175 V.
• Leakage current _at_ 150 V lt 500 nA
• (lt 200 nA averaged over 100 SSDs).
• Rate of bad strips lt 0.2.

• Procurement and testing
• Hamamatsu Photonics qualified producer.
• 11500 SSDs delivered (10368 for 18 towers
  spares wastage).
• All SSDs already electrically and geometrically
  tested.

• Electrical test
• Performed at wafer level.
• Leakage current 110 nA on average ( 1 nA/cm2).
• Depletion voltage 70 V on average.
• Electrical test
• Error on wafer cut 2.5 mm.
• 0.5 final rejection rate!

20
Ladders assembly/testing

• Ladders assembly
• 4 wafers glued head to head.
• Wafers are wire-bonded together as to form an
  unique 9 x 35 cm detector.
• Wire bonds are encapsulated.

• Ladders production/testing
• 1330 flight ladders produced (more than 50 of
  the total production).
• 500 ladders under construction.
• 1200 ladders geometrically/electrically tested.
• 2 rejection rate (including start-up
  problems).
• 0.016 bad channels caused by wire bonding or
  probing.

21
Panels production/testing

• Bare panels construction
• Aluminum internal honeycomb, carbon-carbon
  closeouts.
• Carbon-fiber face-sheets.
• W tiles (converters).
• Bias circuits.

• Bare panels production/testing
• Final design frozen, flight production started.
• ESPI acoustic excitation interferometry of
  laser beams. Resonance frequencies measurement
  and defects search.
• Reliable, non-destructive. Allows early
  identification of problems before the final
  tooling.

22
Tray assembly

• Trays assembly
• Glue spots deposition with automatic dispenser.
• Micro-bonding with automatic wedge bonder.
• Good experience with pre-production panels.
• 1 assembly chain ready, 5 under construction.
• Maximum rate 15 trays/week.
• Foreseen assembly rate 10 trays/week.

23
Tray testing

• Tray testing
• Trays stored in aluminum shipping containers
  before leaving the production site.
• All testing (electrical/functional, thermal
  cycles) is done without opening the box.
• No contamination or accidental mishandling.
• Boxes opened right before the tower assembly.

• Stacked trays
• Functional tests/CR burn-in for a whole tower in
  parallel.
• External trigger capability.

24
Assembly of tower 0

• The history of tower 0
• Some issues in the production of flight trays
  recently identified. Few design changes needed.
• First full tower (originally Tower A, then
  renamed as Tower 0) assembled anyway.
• Not all the trays instrumented with silicon
  detectors.
• All the hardware handled as flight hardware.
• E2E test and validation of the assembly
  procedure.
• First data collected in Tower configuration.

25
Assembly of tower 0

• Tray is positioned and the handlers are removed

• Trays are slit into the assembly jig by means of
  service handlers

• The stack grows upside-down and is completed
  with the bottom tray

26
Assembly of tower 0

• Each side is completed with two readout flex
  cables (to be connected with the TEM) and the
  tower is covered with a dark box for a CPT.

27
Tower 0 testing

• Noise measurement
• Single strip noise occupancy (averaged over each
  layer) measured as a function of the threshold on
  discriminators.
• Plateau _at_ high threshold consistent with the
  expected rate of accidental coincidences with
  cosmic rays.
• Noise occupancy measured for each strip at the
  nominal threshold.

Design specification
Nominal DAC setting (1/4 MIP)

• Detection efficiency
• Measured by means of external scintillators on
  the stack.
• Detection efficiency (within the active area)
  found to be typically gt 99.5 _at_ the nominal
  threshold setting.

28
Assembly of tower 0 (II)

• Completion of the assembly procedure
• Sidewalls are put in place (gt 700 screws per
  side!).
• Tower craned to the CMM and fixed on the marble
  for the alignment procedure and verification.
• 200 mm maximum shift in one direction, 60 mm
  leaning (within specs).
• Ready for electrical tests/data taking.

29
First events in tower configuration...
XZ view
YZ view
30
Online distributions

• Online monitoring
• All relevant quantities (hit-map, hit
  multiplicity, TOT distribution) monitored online
  for all the layers.

31
Preliminary offline analysis
32
Conclusions

• GLAST will survey the sky in the 20MeV1TeV
  g-ray band, where the most energetic and
  mysterious phenomena in nature reveal their
  signature.
• GLAST is equipped with state-of-the-art particle
  detectors, resulting in an order of magnitude
  improvement in sensitivity and resolution with
  respect to previous missions
• The GLAST LAT Tracker is the largest Si tracker
  ever built for space applications (gt 10K SSDs,
  gt80 m2, 106 channels).
• A highly modular design was chosen for a
  simpler, cost-effective and more reliable
  construction.
• Construction is organized in a well-defined
  sequence of increasingly complex operations
  wafers testing completed, ladders assembly well
  on the way (50 produced and tested).
• Final production phase just begun, delivery of
  the last tower to IT next year.

33
(No Transcript)
34
Extra - Showers
35
Extra - Showers
36
Extra detection efficiency