SCIPP Large-Scale Application of Silicon Detectors in...

Firenze 2001: Silicon Detectors Hartmut F.-W. Sadrozinski , SCIPP, UC Santa Cruz

SCIPPSCIPP

Large-Scale Application of Silicon Detectorsin Space

Hartmut F.-W. SadrozinskiSanta Cruz Institute for Particle Physics (SCIPP)

Development of Silicon DetectorsGLAST :•Gamma-Ray •Large Area •Space Telescope


SCIPPSCIPP Principle of Silicon Strip Detectors

25-200 µm

300-400 µm

Alat ~ 100V

n+ implant

Al SiO2

p+ implantat ground

Depletion region. Charged particletraversing region produces ~80electron/hole pairs per micron.

Readout electronics(S/N typically > 20)

holes

Reverse Bias of junction: thermal current generationScale : Band gap 1.12eV vs. kT = 1/40eVCooling needed only in ultra-low noise applications.Wafer thickness 300um = 24k e-h pairs = 0.3%RLDepletion Voltage ~ thickness2 <100V Collection Time of e-h pairs: ~30nsArea is given by wafer size: 4” & 6” => Ladders


SCIPPSCIPP Evolution of Silicon Detectors

Large Area Double-sided

Hybrid Pixels

Monolythic:CCD, MAP

Si Drift

3-D

n n n

nn

p p

n


SCIPPSCIPP The Rise of the Silicon Detectors

Development of Area of SSD and # of Electronics Channels follow Moore’s Law

Larger - CMS 10M Channels, 230m2

Faster - ATLAS 22nsCheaper - CMS ~$5/cm2

0.01

0.1

1

10

100

1000

1985 1990 1995 2000 2005 201

Silicon Area [m2]

Year

CDF

ATLASGLAST

CMS

AMS-02

AMS-01

D0

BaBar

NOMAD

LEP

LPSCDFMark2

Pamela

Agile MEGA

1

10

100

1000

104

1985 1990 1995 2000 2005 2010

# of Electronics Channels [in k]

Year

CDF

ATLAS

GLAST

CMS

AMS-02

AMS-01

WIZARD

D0

BaBar

NOMAD

LEP

LPSCDFMark2 Pamela

Agile MEGA



Edge joint and wire bonds before encapsulation

0.01

0.1

1

10

100

10 100 1000 104

Silicon Area vs. # of Electronics Channels

# of Channels [k]

CDF

ATLASGLAST

CMS

AMS-02

AMS-01 D0

BaBar

NOMAD

LEP

LPSCDF

Mark2 Pamela

Agile MEGA

Are

a[m

2 ]

Limited Resources (Power) in Space

Long Ladders possible with:Bonding and Encapsulation



Trends in the Cost of Silicon Detectors

Cost of processing wafers reduced ~ 4x

Increased Area 4” -> 6”Better utilisation of area

Improved Qualitye.g. GLAST detectors:

<2nA/ cm2

<2*10-4 bad channels 1

10

100

1985 1990 1995 2000

Cost /Area of Single-sided Silicon Strip Detectors(double-sided factor 2.5 higher)

4"6"

Cos

t /A

rea

[ $/c

m2 ]

Year

Mark 2DC coupl.

ZEUSDC coupl.

CDFNomad

(untested)

GLAST"4"

ATLAS

GLAST6"

Wafer Size

Blank Wafer Price4 "

6 "

CMS

(Guestimates by HFWS)


SCIPPSCIPP Tracking Milestones: Fixed Target

Silicon Detectors~ 5cm x5cm

Fanout-Cables

Amplifiers

That’s how it all began

Fixed Target experiments withhigh rates:

Na11 (ACCMOR)Na14E706E691

Detect heavy decaying particles through their finite decay distance


SCIPPSCIPP Tracking Milestones: Vertex Detectors

The big step forward in Mark2:ASIC’s (Terry Walker et al)

Vertex Detector ParadigmASIC’s,Few thin layers,Close in.

ALEPH

Every LEP Experiment has aVertex Detectors:

Double-SidedAC-coupled


SCIPPSCIPP Tracking Milestones: Speed and Rad.Hardness

56 planes, 50k channelsElliptical shapes!

2 chip set: Bipolar+CMOS

LPS at HERA (UC Santa Cruz & INFN)“Fixed Target” at ColliderImportance of Electronics:

rad hardfastlow noiselow power


SCIPPSCIPP Tracking Milestones: Highest Luminosity LHC

Temperature Range : -17oC (cooling pipe) to +16oC (ASICs)

Vertex Detector Inner DetectorChange in Paradigm:

coverage of large areaelectronics inside tracker volume

ATLAS: Silicon TrackerSimple Detectors,Optimized ElectronicsThermal management


SCIPPSCIPP Tracking Milestones: Highest Luminosity LHC

Continued Paradigm Change:Outside radius : ~1.1m~1R.L. in tracking volume

Silicon has arrived:all Silicon Inner DetectorSi Area 223m2, - 6” Wafers -


SCIPPSCIPP Silicon Tracker in Space: AMS

Very long ladders (65cm)Thin mechanical structuresElectronics outside trackingPrecision alignment

Tracking in Magnetic field:Minimize material

Conservative Adaptation of HEP Technology for ISS


SCIPPSCIPP Photon Detection: Direction crucial for Astronomy

Optical-X-raysNeed Focus:LensesMirrorsCollimatorsCoded MasksProximity

Photon energy (MeV)

1

0.01

0.1

10

100

1000

Mas

s at

tenu

atio

n co

effic

ient

(cm

2 g–1

)

photo-electric pair production

total

Si l icon

0.10.010.001 1 10 100 1,000 10,000

Rayleigh(coherent)

Compton

λ varies by 105!

Absorption of PhtotonsN(x) = Noe- λ x

ComptonPartial Direction

γ

calorimeter (energy measurement)

anticoincidenceshield

e+ e–

particle tracking detectors

conversion foil

Pair-ProductionDirection

Absorption coefficientλ = (7/9)/Xo


SCIPPSCIPP Application: Compton Telescope AstroPhysics

Reduced Compton circles of events with electron track

MEGAUse of electron directionto limit the Compton cone.

Stack of Silicon detectors

MPE - NRL

Multiple Compton +Energy Measurement

Classical Compton Event Circles


SCIPPSCIPPEGRET / GLAST / AGILE: Pair Conversion Telescope

γ

e+ e- calorimeter (energy measurement)

particle tracking detectors

conversion foils

charged particle anticoincidence shield

1

2

Converter Thickness tConversion Probability ~ tPointing RMS ~ √t

Gamma-rays convert into e+e- pairs,are tracked and their energy measuredGamma is reconstructed from e+e- tracks

MaximizeNumber of Converters

New Paradigm:Add material into tracking volume:


SCIPPSCIPP GLAST Gamma-Ray Large Area Space TelescopeAn Astro-Particle Physics Partnership Exploring the High-Energy Universe

• Precision Si-strip Tracker (TKR)• Hodoscopic CsI Calorimeter (CAL)• Segmented Anticoincidence Detector (ACD)• Advantages of modular design• NASA, DoE, DoD, INFN/ASI, Japan, CEA, IN2P3, Sweden

Challenges of Science in Space

• Launch

• Limited Resources• Space Environment

4 x 4 Arrayof Towers

AnticoincidenceShield

CalorimeterModule

Grid

TrackerModule

GammaRay

Resolving the γ-ray sky

Design Optimized for Key Science Objectives

• Understand particle acceleration in AGN, Pulsars, & SNRs• Resolve the γ-ray sky: unidentified sources & diffuse emission• Determine the high-energy behavior of GRBs & Transients

Proven technologies and 7 years of design, development and demonstration efforts


SCIPPSCIPP GLAST LAT: International Collaboration• expertise in each science topic (theory + obs.)• experience in high-energy and space instrumentation• access to X-ray, MeV, and TeV observatories by

collaboration for multi-wavelength observations

• expertise in each science topic (theory + obs.)• experience in high-energy and space instrumentation• access to X-ray, MeV, and TeV observatories by

collaboration for multi-wavelength observations

~ 100 collaboratorsfrom 28 institutions

~ 100 collaboratorsfrom 28 institutions

Organizations with LAT Hardware Involvement

Stanford University & Stanford Linear Accelerator CenterNASA Goddard Space Flight CenterNaval Research LaboratoryUniversity of California at Santa CruzUniversity of Washington

Commissariat a l’Energie Atomique, Departement d’Astrophysique (CEA)Institut National de Physique Nuclearie et de Physique des Particules (IN2P3): Ecole Polytechnique, College de France, CENBG (Bordeaux)

Hiroshima UniversityInstitute of Space and Astronautical Science, TokyoRIKENTokyo Institute of Technology

Istituto Nazionale di Fisica Nucleare (INFN): Pisa, Trieste, Bari, Udine, Perugia, Roma

Royal Institute of Technology (KTH), Stockholm

Organizations with LAT Hardware Involvement

Stanford University & Stanford Linear Accelerator CenterNASA Goddard Space Flight CenterNaval Research LaboratoryUniversity of California at Santa CruzUniversity of Washington

Commissariat a l’Energie Atomique, Departement d’Astrophysique (CEA)Institut National de Physique Nuclearie et de Physique des Particules (IN2P3): Ecole Polytechnique, College de France, CENBG (Bordeaux)

Hiroshima UniversityInstitute of Space and Astronautical Science, TokyoRIKENTokyo Institute of Technology


Royal Institute of Technology (KTH), Stockholm

TKRCALACD

CAL

TKR

TKR

CAL


SCIPPSCIPP GLAST LAT: International Collaboration

EGRET ~10

GLAST ~100

ATLAS/CMS ~1000


SCIPPSCIPP GLAST LAT TKR Collaboration

Organization of the LAT TKR Subsystem Activities

University of California at Santa CruzManagementFront-end Electronics

Stanford University & Stanford Linear Accelerator CenterFront-end ElectronicI & T

Hiroshima UniversitySSD


MechanicsAssemblyTower I&T

Organization of the LAT TKR Subsystem Activities

University of California at Santa CruzManagementFront-end Electronics

Stanford University & Stanford Linear Accelerator CenterFront-end ElectronicI & T

Hiroshima UniversitySSD


MechanicsAssemblyTower I&T


SCIPPSCIPP Criteria for large-scale Application in Space

FlexibilityAdapt to Space Environment (low power, light weight, rigid, radiation, inaccessibility)Use Conservative (“proven”) Approach

ModularityClean InterfacesLowered Risk in Performance and Schedule

SimplicitySimple, controlled AssemblyRobust Detectors and Electronics

RedundancyNo single-point Failures

Q/AParts SelectionProcedures – Work with IndustryEarly R&DTesting


SCIPPSCIPP The Large Area Telescope (LAT)

DAQ Electronics

Grid

Tracker

Calorimeter

ACD Thermal Blanket

•Array of 16 identical “Tower” Modules, each with a tracker (Si strips) and a calorimeter (CsI with PIN diode readout) and DAQ module.

•Surrounded by finely segmented ACD(plastic scintillator with PMT readout).

•Aluminum strong-back “Grid,” with heat pipes for transport of heat to the instrument sides.


SCIPPSCIPP Overview of TKR Tower Design

• 16 towers, each with 37 cm × 37 cm of Si (78m2 in all)

• 18 x,y planes per tower– 19 “tray” structures ~3cm high

• Si planes on top and bottom• 12 with 3% W converter on bottom• 4 with 18% W converter on bottom• 2 with no converter

– Every other tray rotated by 90°, so each W foil is followed immediately by an x,y plane

• 2mm gap between x and y• Electronics on the sides of trays

– Minimize gap between towers– 9 readout modules on each of 4 sides

• Trays stack and align at their corners• The bottom tray has a flange to mount on

the grid• Carbon-fiber walls provide stiffness and the

thermal pathway to the grid (∆T~11oC)

Carbon thermal

panel

Readout Cable

Electronics Module

2 mm gapCarbon-Fiber Wall

19 Carbon-Fiber Tray Panels


SCIPPSCIPP 2 Clean Interfaces per TKR Tower

4.1.4.3.1Silicon-Strip

Detectors

4.1.4.3.2Tray Mechanical

C-fiber panelW converters

4.1.4.3.3Tray Electronics (MCM)

F.E. ASIC; Controller ASIC; PC Board;Connector sockets; Pitch Adapter; Passive parts

4.1.4.4.1Tower Structure

C-fiber sidewallsFasteners

Spacers/pinsEMI shield

4.1.4.4.2Tower Cable Plant

Flexible multi-layer cables; Connector Plugs

Wire BondsScrews;

Adhesive tape

BiasCircuit;

Adhesive

Det

ecto

r Bia

s

Wire

Bon

ds

Nano- Connectors Machined

Cable RunsFasteners

TowerElectronics

Module

GridFlexure Mount

ThermalGasket

GLAST TrackerBlock Diagram

andInterfaces

Mechanical

ElectricalNo Inter-Tower Interfaces


SCIPPSCIPP GLAST LAT TKR

• Numbers: GLAST is modular:16 flight (+ 2 calibration towers) with 18 x-y SSD planes each4 x 4 SSD per plane ( 4 ladders with 4 SSD each)Number of SSD needed:

10368 (+ 5% spares + 5% wastage) => 11,500 • Total SSD Area: 83m2, ~1M channels, ~ 5M wire bonds• Simple mechanical assembly method:

Butt-join and wire-bond 4 SSD to “ladders”Glue 4 Ladders onto both sides of 3cm thick panels (“trays”) Attach MCM on the side of the panel via 90o interconnect Stack trays into towers

• QA:Tight specification increase reliability of SSDCharge manufacturer with all detailed testingTest important parameters before further integration step


SCIPPSCIPP TKR Flight-Tower Design & Assembly

Cable PlantUCSC I

Tower Structure (walls, fasteners)Engineering: SLAC, HytecProcurement: SLAC I

Tower Assembly and TestSLAC (2) Italy (16)

Tray Assembly and TestItaly I

Composite Panel & ConvertersEngineering: SLAC, Hytec, and ItalyProcurement: Italy I

2592

342

648

34218

SSD Ladder AssemblyItaly I

SSD Procurement, TestingJapan, Italy, SLAC I

10,368

Electronics Design, Fabrication & TestUCSC, SLAC I

Most Production and Assembly Steps done in Industry = I


SCIPPSCIPP TKR Interconnects: Industry Job

~ 1,000,000 TKR Channels~ 6,000,000 encapsulated Wire Bonds


SCIPPSCIPP GLAST Front-End Electronics ASIC

Binary Readout: •Low-power (~200uW/channel) •Peaking time ˜ 1.3 ms•Low noise (Noise occupancy <10-5)•Threshold set in every ASIC•Separate Masks for Trigger and Readout in every Channel•Trigger = OR of one Si plane

(1536 channels)

Pulse Height:Time –over-Threshold on the OR of every Si plane

Distinguish single tracks from two tracks in one strip

Electron Events

Photon Events


SCIPPSCIPP Redundant TKR Electronics

• Serial, LVDS readout and control lines.• Two readout and control paths for every 64-channel front-end chip.• Any single chip can fail without preventing the readout of any other.• Either of the two communications cables can fail without affecting the

other. 24 64-channel amplifier-discriminator chips for each detector layer

2 readoutcontroller chipsfor each layer

Con

trol s

igna

l flo

w Control signal flow

Data flow to FPGAon DAQ TEM board.

Data flow to FPGAon DAQ TEM board.

Control signal flow

Data flow

Nine detector layers are read out on each side of each tower.

GTRC

GTFEGTFE

GTRC

GTRC

GTRC

GTRC

GTRC

9-998509A22

• Trigger output = OR of all channels in a layer.

• Upon trigger (6-fold coincidence) data are latched into a 4-event-deep buffer in each front-end chip.

• Read command moves data into the GTRC.

• Token moves data from GTRCs to TEM.


SCIPPSCIPPTKR Tray Minimize Material Outside of Converters

• Lightweight 4 piece machined closeout frame, bonded to face sheets and core, form a sandwich structure (R.L. : SSD = .8%, Tray = 0.7%)

• Tray payload is bonded to the sandwich structure using epoxy, with the exception of silicone used to bond SSD’s

Converter Foils

SSD’s

Bias-Circuit

Structural Tray

TMCM

Bias-Circuit

SSD’s


SCIPPSCIPP Instrument Performance

FOV: 2.4 srSRD: 2.0 sr

Single Source F.o.M ~ Aeff / [σ(68%)]2 “sweet spot” 1 – 10 GeV


SCIPPSCIPP GLAST LAT SSD


SCIPPSCIPP Design Criteria for the GLAST SSDPredictability of Performance

Poly-silicon biasingOxide-Nitride combination for coupling cap dielectricLow leakage currentLow depletion voltageSilicon dioxide passivationLarge Aluminum overhang on stripsBalance strip width effects: capacitance vs. fieldAccurate mask alignmentNo voltage across saw-cut

Ease of TestingAC couplingLarge, redundant bonding/probing padsN-sub contact on top

Ease of AssemblySeparate fiducials for alignment, bonding and metrologyAccurate control of saw-cut


SCIPPSCIPP Prototyping of the GLAST SSD

The SSD design has been finalized and procurement is underway

11,500 SSD inlude 10% Spares

Qualify Prototypes from HPK (experience with ~5% of GLAST needs)

0.1*specs

+340

Additional Prototypes: Micron (UK), STM (Italy), CSEM (Switzerland)


SCIPPSCIPPSSD Prototype Quality

• The principal SSD vendor, Hamamatsu Photonics, is already qualified, with prototypes of the final design delivered and exceeding specifications.

• The average leakage current is sufficiently low that detectors with one or two bad strips can be identified by measuring the bulk bias current.

• The number of bad strips, 0.03%, is well below our tight spec (0.2%).

0

50

100

150

200

250

0 10 20 30 40 50 60 70 80

HPK @25deg CSLAC @19deg CHIroshima @19deg CINFN Pisa

Wafer #

Average SSD Current

Specification: 200nA

HPK: 143nA

TempCorr.

Leakage current in nA measured on prototype flight SSD (HPK vs. Pisa, Hiroshima, SLAC)

All Production Testing done by Manufacturer:

I-V (No single strip I-V)

C-V

Coupling caps (CC) on every strip

Resistors, Dimensions per batch

GLAST: Acceptance I-V

I-V and CC after bonding


SCIPPSCIPP SSD Production Quality

0

50

100

150

200

250

50561 50563 50565 50567 50569 50571 50573

HPK 1rst + 2nd LotAverage Run Leakage current

5056250563505645056550566

5056750568505705057150572

Run #

Run #

< i > = 186 nA

0

20

40

60

80

100

120

50561 50563 50565 50567 50569 50

HPK 1rst & 2nd LotAverage Depletion Voltage

5056250563505645056550566

Run #

Run #

The first 320 Production SSDLeakage Current and Depletion Voltage

GLAST LAT Specifications;Ave current < 200nA Depletion Voltage <120VMax Current < 500nA

571 50573

5056750568505705057150572


SCIPPSCIPP Radiation Hardness

Test Total Dose Dependence of Leakage currents, CInterstrip, CBody, RBias, RInterstrip, Ccoupling

Up to 20kRad (1kRad predicted)

Heavy Ion Test of SEE Immunity of Coupling Cap: ok

Live Baby Skinny010203040506070

0 5 10 15

perimeter/area ratio [/cm2]Le

akag

e cu

rren

t [nA

/ cm

2 ]

0kRad 10kRad

Max spec10.2nA/cm2

Max spec afterirradiation120nA/cm2


SCIPPSCIPP Dimensional Tests on HPK SSD

Thickness

Saw Cut(Crucial for Assembly):

Spec at 20um.Average tiltnow 2um!

Planarity

0.5

0.4

0.3

0.2

0.1

0.0

Thickness (mm)

38

37

36

35

34

33

32

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

15

14

13

12

11

109876

Detector ID

HPK preseries

Measured capacitanceWill yield thickness for every SSD


SCIPPSCIPP TRK of the Beam Test Engineering Module

End of one readout hybrid.

BTEM Tracker Module with side panels removed.

Single BTEM Tray

The BTEM Tracker, (~1/16 of the flight instrument) for the SLAC test beam (11/99 – 1/00)

- 2.7m2 silicon, ~500 detectors, 42k channels- all detectors are in 32 cm long ladders.

Si Detectors

HPK 296 (4”), 251 (6”)

Micron 5 (6” ) Leakage I: 300 nA/detector (HPK)

Bad strips: about 1 in 5000


SCIPPSCIPP Assembly of BTEM Tracker at SCIPP

4 trays, 10 eyes & 10 hands

17 trays!

2 delicate hands

2 trays and 2 observers

All done and all smiles.


SCIPPSCIPP Installation of the BTEM at SLACBeam Test in SLAC’s Endstation A ( Dec 1999/Jan 2000)

Silicon Tracker

CsI Calorimeter

ACD

•Test Fabrication Methods•Verify Performance

ResolutionsTriggerMC Programs


SCIPPSCIPP Beam Test at SLAC 1999/2000: e+ and γ in BTEMHigh efficiency (99.9%), low noise occupancy (<10-6)

ConversionPoint


SCIPPSCIPPEffective Area = Geometric Area * Conversion efficiency

PSFAeff ~ 6 % R.L

~ 25 % R.L

Back Section:3 layers of thickEfficient Conversionbefore calorimeter

Front Section:10 layers of thin converters:Precision Tracking

Effective Area:Vertex Distribution reflects the Converter Distributionshows expected Attenuation of the γ Beam


SCIPPSCIPPBalloon Flight:

1/16 GLAST – 45km up - ½ of Texas – 50g

August 4, 2001


SCIPPSCIPP GLAST LAT Project Schedule

2000 2001 2002 2003 2004 20052011

Formulation Implementation

SRR NAR M-PDR M-CDRI-PDR I-CDR Inst. Delivery Launch

Build & TestEngineering Models

Build & TestFlight Units

Inst.I&T

ScheduleReserve

Inst.-S/CI&T

Ops.

Calendar Years

SSD Procurement SSD Procurement

Ladder Production Ladder Production

Tray AssemblyTray Assembly


SCIPPSCIPP GLAST Development Process and StatusDate Activity Program

93-98 Conceptual study NASA SR&DDetector R&D DoE R&D Beam Test 98

98 DoE Review SAGENAP Endorsement

98-00 Technology NASA ATDDevelopment (BTEMIrradiations Full Size Modules

Manufact. ProcessBeam Test 99/00 ASIC’s, DAQ)

Fall 99 Instrument Proposal NASA AO GLAST LAT (Si TKR, CsI CAL, ACD)Endorsements, MoA

Feb 25, 00 Decision on AO GLAST-LAT selectedAug 01 Balloon Flight

IrradiationsJan. 11, 02 PDR Baseline Review

March 2006 Launch on Delta 2

What Kind of Surprisesare awaiting us?


SCIPPSCIPP Follow Moore’s Law into Space

0.01

0.1

1

10

100

1000

1985 1990 1995 2000 2005 2010

Silicon Area [m2]

Year

CDF

ATLASGLAST

CMS

AMS-02

AMS-01

D0

BaBar

NOMAD

LEP

LPSCDFMark2

Pamela

Agile MEGA

1

10

100

1000

104

1985 1990 1995 2000 2005 2010

# of Electronics Channels [in k]

Year

CDF

ATLAS

GLAST

CMS

AMS-02

AMS-01

WIZARD

D0

BaBar

NOMAD

LEP

LPSCDFMark2 Pamela

Agile MEGA

Better Hurry: by Year 2010, Moore predicts 1,000 m2 of SSD10,000,000 electronics channels


SCIPPSCIPP Follow Moore’s Law into Space

But

Moore predicts by Year 2010 a cost of ~ 150 $ /SSD

1

10

100

1985 1990 1995 2000

Cost /Area of Single-sided Silicon Strip Detectors(double-sided factor 2.5 higher)

4"6"

Cos

t /A

rea

[ $/c

m2 ]

Year

Mark 2DC coupl.

ZEUSDC coupl.

CDFNomad

(untested)

GLAST"4"

ATLAS

GLAST6"

Wafer Size

Blank Wafer Price4 "

6 "

CMS


SCIPPSCIPP Challenge #1 : Launch and CTE Mismatch

BTEM TKR tray undergoing random vibration testing at GSFC.

Vibration Testing of a live tray up to 14g.Leakage current before and after shaking identical

Space Qualification:Assembly Methods

MaterialsTests

Aluminum and carbon-fiber mechanical model

of 10 stacked tracker trays, used by Hytec,

Inc. to validate the design in vibration tests.

FEM analysis of (a) TKR tray deflections and (b) of a complete TKR module. Fundamental frequencies are above 550 Hz for the tray and 300 Hz for the module, clamped only at its base.


SCIPPSCIPP Challenge #2: On Board Cosmic Ray Rejection

Diffuse High Latitude gamma-ray flux

C.R. Rejection needed 105 : 1 segmented ACD segmented CAL segmented TRK

Radiation Levels: 1krad in a 5year missionIssue: SEE from Heavy Ions (SEU & Latch-up)See below

LVL1 : 5kHzDownlink: 30Hz


SCIPPSCIPP Challenge # 3: 1M channels, 250W Power

• Noise occupancy <5x10-5 with efficiency of >99% for MIP (within fiducial region).

• Low power consumption (<240 mW/ch).• Self triggering.• Sustain 10 kHz trigger rate with <10% dead time.• Radiation hardness >10 kRad.• Single event latchup resistance to >20 MeV cm2/g LET.• Single event upset: configuration registers resistant to >3 pC charge deposition.• Redundant read-out scheme to minimize the possibility of catastrophic single-point failure• Compact, to minimize inter-module dead space.


SCIPPSCIPP Redundant, ultra-low Power, low-noise FEE

Boss for mechanical and thermal attachment to the wall.

28 Amplifier chips

TermResis4 layers of 1/2 oz coppTEM

needs shielding around cable.

Digital Analog

Cross-over into the side arms

Bias + Analog 3.3VAnalog GroundAnalog 1.5V

Digital 3.3VDigital GroundLVDS Signals

Kapton Cable down the Tower Walls

Hybrid:

Electrical & mechanical Challenge

Digital readout controller chip at each end

25-pin Nanonics connector


SCIPPSCIPP

0 200 400 600 800 1000 1200 1400

10-5

Strip Number

Layer 6x

Occ

upan

cyChallenge #4: Tracker Noise and Efficiency

• Noise occupancy determines the noise rate of the LVL1 trigger, a coincidence of 6 OR’dlayers.

• Noise RMS σ = 130 + 21*C/pF [e-] , τ =1.3µs• Hit efficiency was measured using single

electron tracks and cosmic muons.• The requirements were met: 99% efficiency

with <<10−4 noise occupancy.0 200 400 600 800 1000 1200 1400

Threshold (mV)

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

Effi

cien

cy

Layer 10 xLayer 10 y

1 2 3 4 5Detector Ladder

95

96

97

98

99

100

101

Hit

Effi

cien

cy

Layer 6xCosmic RaysElectron Beam

Noise occupancy and hit efficiency for Layer 6x, using in both cases a threshold of 170 mV. Nochannels were masked.

Hit efficiency versus threshold for 5 GeV positrons.

100,000 triggers


SCIPPSCIPP Challenge #5: Space Environment: RadiationGLAST is in a Low-Earth Orbit (550km):

Shielding of Atmosphere and magnetic Field Avoid (most!) of the radiation belts

Orbit co-determined by Re-entry > 10 Years, < 30years.

USA on ARGOS

Radiation Belts: - High LatitudeSouth Atlantic Anomaly (SAA)-Trapped electrons and protons

responsible for Total Dosecause huge trigger rate(Detectors will be switched off)

Outside radiation Belts:Charged Cosmic Ray Background (p, e, heavy ions) Responsible for Single Event Effects (SEE)


SCIPPSCIPP

Long-term Radiation Damage:

Entirely given by electron and proton flux trapped in the SAA

Extremely soft spectrum: Self shielding of Instrument:Blanket, ACD, walls: 2.50g/cm2

Cut-off at 80MeV protons

Radiation: Total Dose & Displacement

0

5

10

15

20

25

30

0 5 10 15 20 25 30

GLAST Silicon Tracker End-of-Mission Signal-to-Noise

S/N E-o-M 1x

S/N E-o-M 5x

Temperature [deg C]

Total Dose 1kRad (5 yrs) -NASA safety factor: 5x-Leakage current increase 50% surface, 50% bulk(same temperature dependence).

Increase in shot noise due to radiation constrains operating temperature to below 25oC.

1

10

100

1000

10000

100000

0.01 0.1 1 10

ElectronsBremsstrahlungProtonsTotal

Full dose - Spherical shield550 km 28° circular orbit

5-year mission - Solar Minimum

Depth (g/cm2 Al)


SCIPPSCIPP Heavy Ion Radiation: Temporal Effects (SEE)Linear Energy Transfer LET governs Single Event Effects: SEU, SEL, Punch Through LET is dE/dx: LET (Min ion) ≈ 1.3*10-3 MeV/(mg/cm2), LET ~ Z2 : LET (Fe) ≈ 1-2 MeV/(mg/cm2).

Fe

GLAST IRD

Update from AMS


SCIPPSCIPP Challenge # 6: System Reliability

• The LAT Tracker is being designed to operate at its full capability for >5 years.

• There are no consumables in the system, and the solid-state technology is inherently robust and stable.

• Comprehensive radiation testing (SEE and total dose) has been carried out on SSDs and ASICs.

• Long-term environmental testing is in progress for the SSDs.• A preliminary reliability analysis has been made of the electronics

readout system: LAT-TD-00178.– The mean time to failure is conservatively estimated to be 2.5

years.– The vast majority of failures considered here would have little or

no effect on the Tracker performance.– Modularity and redundancy in the system limit the impact of more

serious failures.– Modularity and redundancy in the system allow finite number of

failures without loss of performance


SCIPPSCIPP Testing/QA for the GLAST SSD

Acceptance QC: flight sensors• Order high quality SSD testing of detectors is done by vendor:

i-V, C-V, bad channels, R’s, . NO single channel current testing!• Measurement of parameters crucial for assembly by GLAST assembly institution:

i-V and dimensions

Process Control: test structures• Incorporate Test structures, test one out of every lot (48) at Hiroshima U. & INFN:

Measure all detector parameters specified in specifications, radiation effects• Test wire bonding and glueing on test structures

Assembly QA: flight sensors / ladders• Testing after bonding and pre-post encapsulation by GLAST assembly institution:

“Vital” parameters (i-V of ladders and coupling caps on every strip)• Test before tray assembly:

i-V on ladders

This program is based on our experience with the >500 HPK SSD in the Beam Test Engineering Module (BTEM, SLAC-8471).


SCIPPSCIPP 6” wafer of the GLAST SSD

Each wafer has a GLAST2000 SSD and a GLAST cut-off.We have established the correlation between SSD and test structure performance.

GLAST Test Structures“Baby”, 32 strips, 3.5cmMos Structures Bonding Test StructurePhoto Diodes“Skinny”,8 strips, full length,

GLAST2000 SSD8.95cm x 8.95cm


SCIPPSCIPP Test Structures on the GLAST SSD Wafer

“Skinny”, 8 strips, full length

Cint, Rint, RAl,

Bonding Test Structure

MOS Structures

“Baby”, 32 strips, 3.5cm long

I-V (10kRad), Vdep,

Fiducials:AlignmentBonding


SCIPPSCIPP Ladder assembly (INFN Pisa)

GlueingAlignment

Service box for bondingTransfer bridge


SCIPPSCIPPSuperGLAST tray


SCIPPSCIPP Placement of ladders on trays – 1st exercise


SCIPPSCIPP Assembly Challenge: QA, Yield, AlignmentLeakage current on 32cm long ladders after assembly

Assembly Yield: 98%

Bad Channels Fraction: 0.06%

Start with good SSD’sQC during Assembly

End up with good ladders

Simple Mechanical Alignment:RMS: 8um


SCIPPSCIPP TKR Tower Thermal PerformanceTKR Tower Thermal Performance

• Tower Conduction results in an 11.8 °C ∆T down the length (0.35W heat input/TMCM)– Standard tray closeout ∆T is 0.1°C – The temperature drop between

sidewalls and thermal boss is ~0.5°C (2X)

– Sidewall ∆T is 10.3°C– Bottom tray ∆T is 0.13°C – 0.22°C ∆T across V-Therm gasket

along the bottom of each Tower

Tower Temperature Contour Plot

Tower-to-Grid Thermal Interface


SCIPPSCIPP Space Tests

Thermal vacuum

• Facility: thermal & vacuum chambers (10-5 Torr)

• Test: -30°C ⇒ +50°C

• Duration: 160 hr /Tray

Acceptanceα1=α2=1

Qualificationα1=1.25α2=1.4 (Shock)

Dynamic

• Facility: Shaker

• Test:

1. Sine Sweep

2. Random vibration

3. Mechanical Shock

• Duration: ≥ 10 min

Static Load

• Facility: Centrifuge

• Test: ±12 g × α1 on secondary structure applied to each axis independently

• Duration: ???

Warning: the duration of the test is only approximate and will be fixed in accordance

with the facility characteristics


SCIPPSCIPP Space tests: Trays

dynamic (sine sweep/random vibration)Qualification (USA)

thermal

dynamic (random vibration):• fixture (4.6Mlt. + VAT)• test:1) lab use for 1.5 days (shaker) (2.8Mlt. + VAT)2) data acquisition and reporting (1.474Mlt. + VAT)3) leakage current measurement:2 hrs/tray (400klt. + VAT), 380 trays

Acceptance (Italy)

thermal


SCIPPSCIPP Space tests: Towers

dynamic (sine sweep/random vibration)Qualification (USA)

towers A/B thermal

thermal•288 hrs/tower

cosmic rays•1-2 weeks/tower

dynamic (random vibration):• fixture (4.6Mlt. + VAT)• test: lab use for 1day/tower (shaker)

(1.6Mlt. + VAT)

Acceptance (Italy)flight towers


SCIPPSCIPP GLAST LAT Balloon Flight


SCIPPSCIPP GLAST LAT Balloon Flight August 2001


SCIPPSCIPP


SCIPPSCIPPThe LAT Hardware


SCIPPSCIPPGLAST Science Capability

Key instrument features that enhanceGLAST’s science reach:

• Peak effective area: 12,900 cm2

• Precision point-spread function (<0.10° for E=10 GeV)

• Excellent background rejection: better than 2.5×105:1

• Good energy resolution for all photons (<10%)

• Wide field of view, for lengthy viewing time of all sources and excellent transient response

• Discovery reach extending to ~TeV


SCIPPSCIPP Science Capabilities - Sensitivity

200 γ bursts per year prompt emission sampled to > 20 µs

AGN flares > 2 mntime profile +∆E/E ⇒ physics of jets and acceleration

γ bursts delayed emission

all 3EG sources + 80 new in 2 days

⇒ periodicity searches (pulsars & X-ray binaries)⇒ pulsar beam & emission vs. luminosity, age, B

104 sources in 1-yr survey⇒ AGN: logN-logS, duty cycle,

emission vs. type, redshift, aspect angle⇒ extragalactic background light (γ + IR-opt)⇒ new γ sources (µQSO, external galaxies, clusters)

1 yr

100 s

1 orbit

1 day

3EG limit

0.01

0.001

LAT 1 yr2.3 10-9

cm-2s-1

large field-of-view


SCIPPSCIPP Science: High-Energy Behavior of GRBs

Important GLAST properties for achieving science objectives:

• Large area• Low instrument deadtime (20 µs)• Energy range to >300 GeV• Large FOV

Expected Numbers of GRBs and Delayed Emission in GLAST

GLAST will probe the time structure of GRB’s to the µs time scaleSpectral and temporal information might allow observation of quantum gravity effects.

Time between detection of photons


SCIPPSCIPPMulti-wave Length Campaign:AGN, Pulsars, SNRsMulti-wavelength Observations are crucial for the understanding of Pulsars and AGN’s. Flares are largest at high energy.

Overlap of GLAST with ACT’s provides Needed energy calibration.

Crab

Mk 501

Synchrotron Radiation Inverse Compton

Flares


SCIPPSCIPPConclusions

During the next decade, silicon strip technology will expand our knowledge on the enigmatic gamma-ray sources in the heavens:

• active galactic nucleai• gamma-ray bursts• supernova remnants• super massive black holes• pulsars• ...?


SCIPPSCIPP Back-up


SCIPPSCIPP TKR Bottom Tray to Grid Interface

• Radial blade flexure configuration• Titanium blade flexures used to isolate

tracker tower from thermal expansion of aluminum grid

• Thermal gasket to transfer heat to the grid/heat pipes

Corner Blade Flexure

Double Blade Flexure


SCIPPSCIPP GLAST SSD Prototyping with HPKGLAST

1996GLAST

1997GLAST

1998GLAST

1999 GLAST

2000GLAST Specs

Wafer Size 4” 4” 6” 6” 6” 6”

Sensor Size [cm x cm] 6 x 6 6.4 × 6.4 6.4 × 10.7 9.5 × 9.5 8.95x8.95 8.95x8.95

Pitch 236um 194um 194um 208um 228um 228

Implant Width [um] 57 50 50 52 56 56

Thickness 500um 400um 400um 400um 400um 400um

Biasing Punch Through

Poly-Si Poly-Si Poly-Si Poly-Si Poly-Si

Depletion Voltage [V] 140 100 100 120 60, 100 <120

Bias Resistors [MOhm] 30 60 60 38+- 2

20 – 80+-10

Current [nA/cm2] ~2.5 ~2.5 ~1.7 1.8<2.5

2.5<6

% bad strips averagemax /SSD

0.02 0.04 0.04 0.03 0.25 <0.9

# ordered 20 300 240 35 35

# delivered 20 296 256 35 63+310

Date ordered Jun 1997 Sept 98 July 1999 Sept. 2000

Use BT’97 BTEM BTEM <100> Wafer

EMFlight


SCIPPSCIPP Justification of Key SSD Specifications

Leakage Current: (av)<200nA , max<500nA• The low detector leakage current is an indication of a mature

manufacturing process. • A low detector leakage current specification allows us to eliminate the

time consuming leakage current measurement on every strip and measure instead the entire current on the detector only.

• The leakage current has to be kept low to reduce shot noise ( 35cm long strips!)

• Single detector strip with ~20nA has increased noise level.• One of the major limitation for the GLAST LAT is the available power.

The power assigned to the detector biasing is 4W at end of mission, mainly due to radiation damage.

• At 150V, this is 2.6uA/SSD, and the initial detector current should be a small fraction of this number.

• Because we observe a factor 2 increase of the leakage current from production testing to finished ladders, our specs mean actually an initial current of about 500nA/SSD, about 20% of the end of mission limit.


SCIPPSCIPP Justification of Key SSD SpecificationsDetector Thickness: 400um• The TKR has sufficient S/N as specified. The signal is proportional to the

path length. For normal incidence, it’s the thickness (400um), for large angles it’s the pitch (228um). So if S/N is a problem, the pitch has to be increased as well, which hurts the science.

• The depletion voltage has to be kept low to reduce power and noise. We set an upper limit of 150V, but would like to operate at 100V ( like in the BTEM).

• The depletion voltage depends on the square of the detector thicknessand the inverse of the resistivity. Thus changing the thickness from 400um to 500um requires an increase in resistivity of 56%, from 4kΩ-cm to more than 6kΩ-cm. We are told that the wafer manufacturers can’t guarantee a stable supply of 6” wafers with resistivity higher than 4kΩ-cm on our time scale.

• We see now depletion voltages at 120V, which would increase to 188V with 500um thickness at the same resistivity. This is too high.

• Our principle supplier prefers 400um detectors and fabricates (exclusively?) in that thickness. Mixing different thickness sounds like an assembly nightmare.


SCIPPSCIPP Leakage current measurements (HPK SSD)


SCIPPSCIPP Leakage current measurements (HPK SSD)

Accepted sensor Not accepted sensor


SCIPPSCIPP TKR Tray Sandwich Structure

• Lightweight 4 piece machined closeout frame, bonded to face sheets and core to form a sandwich structure

Gr/CE Face Sheet

MCM Closeout Wall

Thermal Boss

Aluminum Honeycomb Core

Structural Closeout Wall


SCIPPSCIPP Application: Optical Imaging with CCD’s

36k2

16k2

8k2

4k2

2000Year

20051990 1995 2010

2k×2k

MACHO

Suprime

UH4K

NOAO 4KCFHT MOCAM

NAOJ

BTC

RGO

UH8K EROS2 QUEST NOAO 8k

UW

DMT

WFHRI_1

SDSSCFH12K ESO OmegaCAM

NOAO 8k (thinned)3

3

3

107

108

109

1011

Nu

mber

of

pix

els

DEIMOSUH8K (thinned)MAGNUMCTIO 8kESO 8kMSSSO/WFI

CFHT MegacamSAO/MMT Megacam

1976

SNAPsat

SNAP: 200 CCD @ 10$each

BTC: 4 CCD @ 120k$ each

Progress:Astronomy: CCD replaced film completely:

ideal match for

Astronomical observations.

long integration time,

long readout time,

low noise

Increase efficiency

Shining from the backside,

high resistivitycollect from bulk

X-rays: Chandra-Newton


SCIPPSCIPP Optimization of Converter Thickness t

0

500

1000

1500

2000

2500

0 5 10 15

Effective Area vs. Conversion Plane

Graded Converter (2.5%, 25%)Uniform Converter (3.5%)

x-y Plane

0.01

0.1

1

10

0.01 0.1 1 10 100

Gamma Angular Resolution PSF(68)

68% Front68% Back

Gamma Energy [GeV]

Aeff ~ t

PSF(68) ~ √t

For Background limited Sources:(Significance) = Aeff / PSF(68) 2

is independent of Converter Thickness

For High Latitude Sources:Number of detected gamma’s count.

0.9038%26%4Back

0.3938%3.8%12Front

PSF(68)@1GeV

[o]

γConversion

X0 per Layer

# of Layers


SCIPPSCIPP FEA simulation of thermal cycles

• temperature range: -30°C ÷ +50°C;• number of cycles: 4;• 12 hr @ -30 °C• (dT/dt) ≤ ± 5 °C/hr;• 12 hr @ +50 °C• estimated duration : 56 hr / cycle / tray

GEVS & LAT req.

Strain gauges:• Resistance: 120 W, Gauge Factor: 2,07;• Configuration:1/4 Wheatstonebridge;

t

-30

50

56 h

12 h 12 h16 h 16 h

T (°C)


SCIPPSCIPP Strain during thermal cycles


SCIPPSCIPP

Controlaccelerometer

Random vibration (independently on x,y,z):• Range: 20 ÷ 2000 Hz• Notching at normal modes frequencies• Duration: few minutes

• Normal modes search (>500Hz);• Range: 10 ÷ 2000 Hz;• Amplitude: 0.25 g ÷ 0.5g;• Frequency scan velocity: 2 ÷ 4 oct / min; • Test duration : ~ 1 min.

Sine sweep (independently on x,y,z):

FEA simulation of vibrational cycles

max (Uz)rms 160µm(at tray centre)

max (σeq)rms 4MPa(at tray centre)


SCIPPSCIPP Application: Compton Telescope

Compton effect: 2-body reaction correlates energy transfer with scattering angle

−−=

12

2 111cosEE

cmeϕPrecise Energy measurement of Compton electrons constrains the scattering angle.

T. Kamae: Multiple Compton scattersallow good determination of the incident angle, and energy

Thick Silicon detectors increase the conversion probability and energy resolution

SCIPP Large-Scale Application of Silicon Detectors in...

Documents

Transcript of SCIPP Large-Scale Application of Silicon Detectors in...