Teilchenphysik-Kolloquium Universität Heidelberg, May 8, 2012 Institut für Experimentelle Kernphysik, Karlsruhe Institute of Technology KIT University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association www.kit.edu
The Large Hadron Collider: Present and Future Pixel Detectors for Collider Experiments The CMS Pixel Detector The CMS Phase 1 Pixel Upgrade CMS Pixel Upgrade: Current Status and Plans 2
The Large Hadron Collider: Present and Future 3
LHC the Large Hadron Collider LHC Accelerator: proton-proton and lead-lead collisions CMS Experiment: multi-purpose experiment Lake Geneva Jura LHCb Experiment: CP violation and B physics CERN accelerator complex, about 100 m under ground LHC circumference: ~27 km ALICE Experiment: heavy ion physics ATLAS Experiment: multi-purpose experiment 4
LHC Luminosities LHC 2012: 8 TeV Luminosity ramped up in record time Already >1 fb 1 delivered Expected for full year: 15 20 fb 1 LHC 2010/2011: 7 TeV Record instantaneous lumi: 3.5 10 33 cm 2 s 1 About 5 fb 1 (= 800,000 top pairs) per experiment Exceeding expectations 5
LHC Physics 2010/11: Rediscovery of the SM Precision measurements of W and Z bosons, top quarks, boson pairs, etc. Jump in center of mass energy (2 TeV 7 TeV): many searches for physics beyond the standard model [pb] σ total 5 10 4 10-1 35 pb -1 35 pb ATLAS Preliminary L dt = 0.035-4.7 fb s = 7 TeV -1 3 10 Theory Data 2010 Data 2011 2 10-1 1.0 fb Example top pair 10 production: 6 7% uncertainty -1 0.7 fb -1 4.7 fb -1 1.0 fb -1 4.7 fb W Z tt t WW WZ ZZ [https://twiki.cern.ch/twiki/bin/view/atlaspublic/combinedsummaryplots] 6
LHC Physics 2011/12: The Hunt for the Higgs 2011: Large increase in integrated luminosity: 35 pb 1 5 fb 1 2012: Center of mass energy further increased: 7 TeV 8 TeV Even more searches for new physics no smoking gun yet 95% CL limit on σ/σ SM 10 CMS Preliminary Observed s = 7 TeV Expected (68%) -1 L = 4.6-4.8 fb Expected (95%) 1 Intriguing hints for the Higgs ) 2 (GeV/c m 1/2 700 600 500 No 400 evidence for SUSY 300 CMS Preliminary τ = LSP 2011 Limits q ~ (1250)GeV 2010 Limits tanβ = 10, A 0 q~ (1000)GeV q ~ (750)GeV 1 Lepton = 0, µ > 0 MT2 SS Dilepton s OS Dilepton = 7 TeV, Ldt Jets+MHT α T CDF 1 fb -1 g~, q ~, tanβ=5, µ<0 D0 g~, ~ q, tanβ=3, µ<0 ± LEP2 χ LEP2 1 ~ ± l g ~ (1250)GeV -1 ) g ~ (1000)GeV Razor (0.8 fb g ~ (750)GeV -1 10 100 200 300 400 500 Higgs boson mass (GeV) [CMS-HIG-12-008] 200 q~ (500)GeV 0 200 400 600 800 1000 2 m 0 (GeV/c ) q~ (250)GeV Multi-Lepton (2.1 fb -1 ) g ~ (500)GeV [https://twiki.cern.ch/twiki/bin/view/cmspublic/physicsresultssus] 7
LHC Challenges Ahead Z µµ with 25 Pileup Vertices High luminosity comes at a price: pileup LHC design luminosities 2808 proton bunches per beam, 25 ns spacing Instantaneous luminosities up to 10 34 cm 2 s 1 Up to 25 simultaneous proton-proton collisions per bunch crossing ( pileup ) Pileup 2012: 1380 bunches per beam with 50 ns bunch spacing Already now: 25 32 pileup vertices 8
LHC Long Term Plan Data Taking Long Shutdown Technical Stop 2012 2015 2018 2021 until 2030 Long Shutdown 1 LS 2 Long Shutdown 3 Tentative date for CMS Phase 1 pixel detetor upgrade CERN: long-term commitment to LHC Goal: deliver 3000 fb 1 of integrated luminosity by 2030 at least 5 increase in instantaneous luminosity Detectors must be upgraded: current detectors suffer from aging and radiation damage, keep similar performance, improve radiation hardness at high luminosity According to current planning: three long LHC shutdowns for upgrades 2013/14: LHC center of mass energy to 13 14 TeV 2018: several machine upgrades After 2022: start of high-luminosity phase of the LHC (HL-LHC) Phase 2 upgrades 9
LHC High Luminosity Upgrade: Physics Case From 2007 slightly outdated by now 7 TeV 13/14 TeV Key element for (almost) all physics channels: precise reconstruction of charged particle tracks Long Shutdown 1 Long Shutdown 2 Long Shutdown 3 now: LHC-HL 10 2010 ~2018* ~2022* *estimate
Pixel Detectors for Collider Experiments 11
Tracking, Vertexing, and B-Tagging Tracking & vertexing Charged particle tracking at small distances (~5 cm) from collision point: precise reconstruction of vertices Charged particle tracking at large distances (~1 m): precise momentum measurement Tracks Secondary Vertex Protons Decay Length x y Protons z Primary Vertex Space Point Vertex Detector Layer Charged Particle Trajectory B-tagging: Identify hadrons with b-quarks via their long lifetimes (picoseconds) Parts of the tracks from B hadron decays: large impact parameters and/or displaced secondary vertex Collision 12
LHC Choice for Tracking Detectors: Silicon Innermost part of the tracking detectors: silicon hybrid pixel detectors Detector = semiconductor diode with pn junction in reverse bias depletion zone Charged particles ionize detector material electron/hole pairs induce signal Electrical Signal Readout Chip: Amplification (Digitization) PbSn or Indium Balls ( bump bond ) 250 300 µm + + + + Implants: n + doped pixels (CMS: 150 100 µm 2 ) Bulk (n doped) Backplane (p + doped) Charged Particle Reverse Bias Voltage approx. 150 V 13
Tracking and Vertexing: Why Pixels? Resolution and material budget Low Occupancy track Small pixels high hit resolution high track and vertex resolution Material budget: 3D space point with a single detector layer Tracking advantages of highly segmented detectors Low hit occupancy (= fraction of bunch crossings in which a pixel is hit) low hit combinatorics Track seed from region with smallest probability for wrong assignment of hits to tracks High Occupancy track track Rule of thumb: tracking works for occupancies of 1% or less fake track 14
T. Rohe Silicon PixelSensor Detectors A Brief History Concepts for Pixel Detectors in HEP First ideas for a hybrid pixel detector Gaalema (1984): first pixelated X-ray detector (Hughes Aircraft Co.) followed by first serious work on hybrid pixel detectors for particle physics at CERN, SLAC, LBNL [PSI] ssues. Hybrid Pixel Detectors Silicon sensor and readout chip: separate devices chip: one circuit p containsreadout ~ 500k transistors per pixel k diodes Interconnects: solder bumps 2 15 Carmel, Sept. 9 12, 2002 Late 1980ies: pixel detectors = tracking option for SSC and LHC CERN: R&D Collaboration (RD19) First small scale applications: fixed target and LEP Today: ALICE, ATLAS, and CMS use hybrid pixel detectors For a historical review see: E. Heijne, NIM A465 (2001) 1
First Experience: Heavy Ion Pixel Telescope First use case at CERN: WA94/WA97/NA57 fixed target experiments strangeness production in heavy ion collisions (Pb) Occupancy similar to LHC, but much smaller rate Pixel telescope: 7 layers parallel, 500,000 channels total Fig. 1. Highly schematic drawing of proposal for array of Si diodes for particle tracking in a magnetic field B (1960, Bromley [1], after Mayer and Friedland, unpublished). 14 Omega2 E.H.M. Pixel Heijne / Nuclear Array Instruments and and Methods in Readout Physics Research A 465 Electronics (2001) 1 26 Event Display 16 Fig. 17. Photograph of an Omega2 array mounted on the WA97 support frame, and the staggered, matching array lying in front. A local card (right) and the remotely placed VME board read out one array with 36 288 pixels in 576 ms. particle hits with charge sharing have been production of seven planes one has tested 200 [Heijne] Fig. 2. A fixed target Pb Pb event reconstruction with 153 tracks, using a seven-plane pixel telescope in WA97 (setup 1995,
First Long-Term Test: DELPHI Upgrade of DELPHI micro-vertex detector for LEP2: two pixel layers in very forward region Long term test of pixel detector technology in collider experiment before LHC start many lessons to learn F. Hartmann / Nuclear Instruments and Methods in Physics Research A 666 (2012) 25 46 [NIM A412 (1998) 304] 17 Fig. 25. The DELPHI micro-vertex detector [19].
The CMS Pixel Detector 18
CMS Compact Muon Solenoid Muon Detector CMS fact sheet length: 21 m diameter: 15 m weight: 14 ktons channels: 80 M Calorimeters Tracking Detectors 19
CMS Photo Gallery Tracker Insertion Lowering YB0 (central wheel & magnet) Off-detector Electronics 20
ARI 17 September 2011 7:10 CMS Tracker Design a 1,200 1,100 1,000 900 800 700 600 500 400 300 200 100 b 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 Quarter Section of the CMS Tracker Tracker inner and outer barrel Pixel (barrel and end cap) General CMS tracker design Tracker end cap [Hartmann, Kaminski, Annu. Rev. Nucl. Part. Sci. 2011.61:197 ] 0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 2,200 2,400 2,600 2,800 Large all-silicon tracker, > 200 m 2 of active area High magnetic field (3.8 T) excellent momentum resolution Pixel detector design considerations in CMS Small radii (4 11 cm) excellent impact parameter resolution, low cost High magnetic field exploit charge sharing through Lorentz drift 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 Note time scales: pixel technology choice dates back to CMS Tracker Technical Design Report 1998 21 1.1 m
Courtesy of Paul Scherrer Institut BPIX: three layers FPIX: two times two disks Mechanics: Light-weight and modular Easy installation/removal 20 µm FPIX Half Disk d [CERN] 22 d BPIX c [CERN] FPIX BPIX Half Shell Figure 5 20 µm wo-dimensional position information. The electronics amplifier chip has the [R.provide Horisberger, (a) So-called hybrid active pixel sensors intrinsicpsi] two-dimen same size and channel pattern as the sensor and is bump bonded (fli n-in-n sensor with 100 150 µm2 pixels per cell. (b) An electron m structured (photolithography) indium bumps before and after the re of the end cap (reproduced courtesy of CERN). erated current (A cm 3) Separated into barrel and forward pixel detectors Annu. Rev. Nucl. Part. Sci. 2011.61:197-221. Downloaded from www.annualreviews.org by 149.172.137.50 on 10/28/11. For personal use only. The Current CMS Pixel Detector Data 0.020 a 0.015 0.010
Pixel Sensor: Precision Through Sharing CiS n + -in-n sensor collect electrons :47 WSPC/139-IJMPA S0217751X10049098 Erdmann Improve hit resolution by charge sharing: March 22, 2010 14:47 WSPC/139-IJMPA S02 Current sensor technology sufficient for Phase 1 upgrade Almost quadratic pixels: 100x150 µm 2 similar resolution in rφ and z Exploit strong electron Lorentz drift in 3.8 T magnetic field 1322 W. Erdmann Most accurate measurement of pulse height: analog readout B - Field ( 4 T ) Q1 Q2 n + - pixel implants Bias grid Bias dot electrons Aluminum p-implant depleted E>0 holes Silicon (p-type) Reduced p-spray Full p-spray dose undepleted E ~ 0 ionizing particle track p + - implant ( - 300 V) [W. Erdmann,Int. J.Mod. Phys. A25 (2010) 1315] lustration of charge sharing induced by Lorentz drift in the CMS pixel barrel detector. 9 ved 23 position resolution the A New charge Pixel Detector shouldfor be the collected CMS Experiment by at least two pixels. After, the detector stays functional beyond the point where it cannot be fully depleted, but Bump (in) Contact via Fig. 4. Photo of four pixel cells with moderate The pixel size is 100 µm 150 µm.
The PSI46 Readout Chip CMS Pixel Read Out Chip 24 Chip features 4160 pixel / chip 250 nm IBM process Area: 7.9 mm 9.8 mm, five metal layers Analog readout at 40 MHz pixel size 100 m x 150 m 251 transistors /pixel Basic layout Pixel array: 52 columns, 80 pixel unit cells in each column Periphery: transfer and store pixel hits, trigger timestamp Discriminator threshold: TS buffers 12 deep <2500 electrons 60 2 /FET 35 W/pixel, pixel ampl. 20nsec peaking on chip regulators 2.6-2.1V 1.9V analog coded readout of addr. & p height operating pixel threshold = 2500 e radiation hard design (~4 x10 15 p/cm 2 ) designed for pixel hit rates <100MHz/cm 2 Time Stamp & Data Buffers in DCOL DB buffers 32 deep Readout by double column and buffer sizes: Buffer serious depth in rate DCOL limitations are leading above order 10 limitation 34 cm 2 of s 1 ROC at 50 eff. ns at bunch high rate spacing LHC. (100 MHz/cm 2 ) upgrade required! Data throughput in Column Drain not our problem yet later yes 9.8 mm IBM_PSI46 CAD layout Column Drain Architecture 2 x80 pixel double column 8.0 mm timestamp & data buffers
CMS Barrel Pixel Module Token bit manager bump bonding: High density interconnect 16 62 mm 2 66,560 pixels Read-out chips SiN Module dimensions: 66.6 26.0 mm 2 [W. Erdmann] 25
CMS Pixel Performance Optimal charge sharing [https://twiki.cern.ch/twiki/bin/view/cmspublic/dpgresultstrk] Single hit efficiency: >99%, measured rφ resolution: 13 µm Data taking efficiency: 97% of pixel channels operational, 99% uptime 26
The CMS Phase 1 Pixel Upgrade 27
General Phase 1 Pixel Upgrade Strategy Goal: similar pixel performance in much harsher environment Modification Impact New digital readout chip Front-end electronics ready for high rates More layers: 3 4 barrel layers, 2 2 2 3 forward disks More 3D pixel space points, more tracking redundancy Smaller radius of innermost layer Improved impact parameter resolution (key to excellent B-tagging at high pileup) Improved mechanics, cooling, and powering Reduced material budget: less multiple scattering, fewer photon conversion 28
New CMS Pixel Readout Chip Goal: overcome rate limitations of current readout chip (100 MHz/cm 2 250 MHz/cm 2 ) Strategy: modest evolution of current chip (staying at 250 nm) First chip iteration: Digital readout: 8-bit ADC for pulse height 6th metal layer reduce cross-talk, lower threshold Larger buffers for data and time stamps First version received from foundry, some minor issues, in testing phase Second chip iteration: Improved column drain architecture Submission planned in late 2012 [B. Meier et al., PSI] 29
Upgraded Barrel Pixel Module upgrade: micro twisted pairs miniaturized Token bit manager High density interconnect 16 62 mm 2 66,560 pixels bump bonding: new digital readout chip SiN [G. Steinbrück after W. Erdmann] 30
Changes to Barrel Pixel Detector Mechanics Barrel Pixel Upgrade Comparison New Present y x [S. Streuli, PSI] sor layers 68 sensor modules) nd full modules) Present BPIX system weight mechanics ng C6F14 3 Sensor Layers (768 Modules) 4 Sensor Layers (1184 Modules) Lightweight Mechanics C6F14 Cooling Support Tube with Readout/Control Electronics (total 1184 sensor modules) (only full modules) New BPIX system Ultra-Lightweight 4 sensor layers Mechanics CO2 Cooling Ultra lightweight mechanics Lightweight Support Tube, Electronics moved to η >2.2 Mass 2...3 times less than present system CO2 cooling much less cooling fluid needed ly tube 31 with control A New and Pixel readout Detector for the electronics CMS Experiment Lightweight supply tube, material moved to η>2.2
New Central Beam Pipe Longer conical section smaller outer diameter [C. Schäfer, CERN] Central part: 0.8 mm beryllium Lighter material in outer part: AlBe instead of stainless steel Installation in Long Shutdown 1 (2013/2014) flexible scheduling of pixel installation 32
New Four-Layer Barrel Pixel Detector [S. Streuli, PSI] The New 4 Layer BPIX Two identical half-shells Two identical half shells Each Layer spitted positions: into four half rings 1. R = 29 mm, 12 faces 2. R = 68 mm, 28 faces 3. R = 109 mm, 44 faces 4. R = 160 mm, 64 faces Full sensor-modules only Layer #1: R 29mm; 12 faces Layer #2: R 68mm; 28 faces Layer #3: R 109mm; 44 faces Innermost layer: Layer #4: R 160mm; 64 faces Radius reduced from 44 mm to 29 mm 2.5 mm clearance to new beam pipe (outer diameter: 45 mm) Total 1184 sensor modules Clearance to beam pipe only 2.5mm (beam pipe D=45mm) New outermost layer: additional 3D space point at 160 mm If present beam pipe to be reused Layer #1: R 39mm; 16 faces FPIX / BPIX Mechanics & Integration with CMS 04-25-2012 33 3 Silvan Streuli PSI Villigen
Pixel Services Detector services = all cables, pipes, optical fibers Most pixel services are buried deep in CMS detector Must reuse existing power cables, optical fibers, cooling lines But: need to supply power, cooling and readout to 1.9 times more readout chips Cooling solution: CO2 Powering solution: high-voltage power lines & DC-DC conversion Readout solution: 40 MHz analog readout 320 MHz digital readout 34
CO2 Cooling Advantages of CO2 cooling: Large latent heat, low viscosity, low coolant mass High pressure small tubing Radiation hard, noncorrosive, non-toxic Implementation à la LHCb/ AMS: 2PACL (two-phase accumulator controlled loop) Two redundant systems for barrel and forward pixels [H. Postema, B. Verlaat, CERN] 35
DC-DC Conversion Power dissipation in the pixel detector Front-end: digital and analog voltage Leakage current through sensors due to radiation damage Ohmic losses in cables Local down-conversion of voltages AMISx ASIC (CERN): buck converter Radiation hard, 80% conversion efficiency Small power supply modification needed Idea of DC-DC conversion (sketch): PS [K. Klein, Aachen] Long cables (~ 50m) E.g. 10V, 0.5A Vin DC-DC Converter E.g. r = 4 10V! 2.5V Buck Converter Schematic t on T 1 T 2 DC-DC Converter PCB Short cables (cm) Detector module Detector module L E.g. 2.5V, 2A Rload V out 36
Ultra-Light Mechanics and Support Tube Mechanical structure Airex foam and carbon fiber CO2 cooling integrated Mass of layer #1: 59 g including coolant 30% reduction Overview of New Supply Tube DC/DC converter area New Pixel Support Tube [S. Streuli, PSI] +Z Opto hybrid area Sensor module connector board area (BPIX layer 3&4 outside, 1&2 inside) Cable trench area (BPIX layer 3&4 outside, 1&2 inside) Away from interaction point: increasing stiffness but also increasing mass 2235mm [S. Streuli, PSI] Towards interaction point: decreasing stiffness but also decreasing mass (material budged) Z=0 FPIX / BPIX Mechanics & Integration with CMS 04-25-2012 6 Silvan Streuli PSI Villigen 37
38 change. This part of the system includes: the fast 40 MHz control links, the trigger, and the fast I 2 C communications. As the number of optical readout fibers remains the same but reading out more modules, the services that come in to the detector Impact on Material Budget Radiation radlen Lengths [X0] Radiation radlen Lengths [X0] 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 Pixel Barrel -2-1 0 1 2 η Pixel Forward Old Geometry New Geometry Pseudorapidity -2-1 0 1 2 η Pseudorapidity Fig. 3. Radiation length as aa function New Pixel of ZDetector of the barrel for the (top) CMS andexperiment forward (bottom) pixel detectors. The dots represent the present detector while the histogram shows the proposed upgrade. efficiency [A. Bean, NIM A650 (2011) 41] fakerate 1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 Central part ( η <1.1): new fourlayer detector with less material than old three-layer detector Standard Geometry Phase 1 Geometry Forward part: shifting of services on support tube further outward 0.5-2.5-2 -1.5-1 -0.5 0 0.5 1 1.5 2 2.5 η 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0-2.5 Standard Geometry Phase 1 Geometry Significant reduction of material budget -2-1.5-1 -0.5 0 0.5 1 1.5 2 2.5 η Fig. 4. Comparison of the full tracking efficiency (top) and fake track rate (bottom) as functions of pseudorapidity after requiring a minimum of three hits forulrich the Husemann current (squares) and upgraded (circles) detectors. Institut The instantaneous für Experimentelle luminosity Kernphysik is (IEKP) assumed to be 10 34 cm 2 s 1 with 25 ns bunch spacing.
17 StdGeom Impact on Tracking: Transverse Impact Parameter Upgrade R30F12 Impact Parameter Uncertainty [cm] )[cm] xy ( d Ratio )[cm] xy ( d Ratio 0.012 0.0 < < 1.0 FullSim 0.01 0.008 0.006 0.004 0.002 standard geom - PU0 ph1 R30, 285!m,100"150!m 2 - PU50 dloss 0 0 2 2 1 10 10 2 p [GeV/c] 1 10 1.5 1.5 1 0.025 0.02 0.015 0.01 0.005 [A. Tricomi] Muons (E-gun) Transverse Impact Parameter 1 10 0.03 1.5 < < 2.0 FullSim standard geom - PU0 2 10 p [GeV] ph1 R30, 285!m,100"150!m 2 - PU50 dloss )[cm] xy ( d Ratio 0.02 1.0 < < 1.5 FullSim 0.015 0.01 0.005 1 1 10 0 0 2 2 1 10 10 2 p [GeV/c] 1 10 1.5 1.5 1 1 10 Preliminary Preliminary 2 10 p [GeV] )[cm] xy ( d Ratio 0.04 0.02 0.01 1 standard ph1 R30 2.0 < < 2.5 FullSim 1 10 standard ph1 R30 39 Please note that PH1 performance at 50P Muon Momentum [GeV] Old Geometry (no pileup) New Geometry (50 pileup vertices) Upgraded 0.03 pixel detector at 50 pileup vertices: As powerful as current detector at no pileup Better resolution for low momenta
R30F12 vs StdGeom b tagging 23 Impact on B-Tagging: Compare ROC Curves ttbar sample at <PU>=50, high purity tracks non b-jet efficiency CSV: mistag vs. b tag efficiency CMS Combined Secondary Vertex Tagger @ 50 Pileup Vertices 1-1 10 10-2 stdg_520steo02-dlosspu50 DUSG r30_520step02-dlosspu50 DUSG stdg_520steo02-dlosspu50 C r30_520step02-dlosspu50 C Preliminary better Significant gain i btagging performance at h PU Current Pixe Upgrade Pixe 40-3 10-4 10 [H. Cheung] 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Phase 1 CDR Meeting Old Geometry New Geometry b-jet efficiency Alessia Tricomi Upgraded detector: significantly better b- tagging performance at high pileup H. Cheung CERN 2
CMS Pixel Upgrade: Current Status and Plans 41
Pixel Project Status and Planning Immediate plans Technical Design Report (Summer 2012): technology choices and physics case Qualification of new components, especially new readout chip Late 2012: ramp-up of barrel pixel module production lines pre-series of all ingredients : sensor, readout chip, HDI, cables, Install new beam pipe during Long Shutdown 1 (LS 1) Space for Pilot Blade Pilot blade project (LS 1) Install upgraded pixel modules at location foreseen for the third forward disk in-situ test of the full chain [S. Kwan et al.] 42
Module Production: BPIX Layer 4 in Germany Plans for distributed barrel pixel module production more logistics (current detector: PSI only) Layers 1 and 2: Swiss consortium (PSI, ETH Zürich, U Zürich) Layer 3: Italy/CERN/Taiwan/Finland Layer 4: German consortium (DESY, U Hamburg, KIT, RWTH Aachen) German consortium Two production lines, 350 modules each: UHH/DESY and KIT/RWTH Sharing development effort for precision tools, testing equipment, Macro-assembly of Layer 4 at DESY Investigations into cost-effective in-house bump bonding (DESY, KIT) Various test-beam studies (DESY, KIT) 43
Aufbau High-Rate Beam Tests ensoren New pixel chip designed for high hit rates (200 MHz/cm 2 ) test high-rate performance a.s.a.p en das Teleskop und ein Sensor das High-rate beam test: 8-layer pixel telescope with single-chip modules CERN H4IRRAD high rate test beam KIT: trigger firmware, tracking and simulation software (based on EuDet telescope software) urch einfachen Stekkartenprinzip r Sensoren kann Lorentzwinkel xperiment simuliert werden GEANT Simulation of a Simplified 8-Layer Telescope r Ortsauflösung zum Modultest entwickelen trahlenhärte in einem Teststrahl [R.-S. Lu, NTU] Teleskop [K. Harder, (Skizze) RAL] Kristian Harder, RAL der 44 Auslese A New Pixel die Detector mit for the CMS Experiment
Module Calibration with Am Source Pixel module calibration Pulse height: linear with energy at all temperatures precise charge sharing Defined energy: characteristic X-rays Source on Cooling Jig charge [e] 12 11 3 10 Chip Calibration T=20 C, Vsf=150 T=10 C, Vsf=165 T=0 C, Vsf=170 T=-10 C, Vsf=165 10 PSI46 Test Board 9 8 7 6 5 / ndf 34.02 / 2 intercept -1570 ± 7.214 slope 64.94 ± 0.04653 80 100 120 140 160 180 200 220 Vcal [DAC] 2 2 / ndf 1461 / 2 intercept -733.6 ± 43.53 slope 61.52 ± 0.2889 2 / ndf 1575 / 2 intercept -131.3 ± 42.41 slope 61.01 ± 0.2974 2 / ndf 4294 / 2 intercept -338.2 ± 71.6 slope 64.53 ± 0.5195 [T. Barvich, S. Heindl, J. Hoß, Th. Weiler, KIT] 45
Module Calibration with X-Ray Tube X-Ray Tube Pixel module calibration Pulse height: linear with energy at all temperatures precise charge sharing Defined energy: characteristic X-rays Calibration Sensor 1 T=20 C Vsf=150 3 10 charge [e] 12 11 Chip Calibration X-Ray tube, 60kV/2mA Am Source 10 Target Single Chip Module 9 8 7 intercept -2691 ± 75.43 6 slope 66.53 ± 0.5626 intercept -2235 ± 420.8 5 slope 64 ± 2.516 120 140 160 180 200 220 Vcal [DAC] [T. Barvich, S. Heindl, J. Hoß, Th. Weiler, KIT] 46
Evaluation of Bump Bonding Options Bump bonding: most cost-intensive step in module production DESY and KIT (Institute for Data Processing and Electronics): in-house options for (part of) bump bonding process Currently investigating flip-chipping and gold-stud bumping Bonder Gold Stud Bumps on a CMS FPIX Module [Th. Blank, M. Caselle, S. Heitz, B. Leyrer, KIT] 47
Three main module types considered All based on local correlation of two sensors: f low-p T tracks The CMS Tracker Beyond 2020 Long Shutdown 3 (2021): preparation for High- Luminosity LHC A: Pixel + Strips, vertical interconnections Upper Sensor (VPS module) B: 2 Strip sensors, connections at the edges (2S module) ATLAS and CMS: complete replacement of tracker Challenge: trigger on interesting physics keep thresholds for key triggers low CMS: exploit tracking information in the early trigger stages 100 µm C: Pixel + Strips, connections 1 mm at the edges (PS module) Pass Fail Novel concept: pt modules Goal: suppression of low-pt tracks (< 1 2 GeV) for the trigger Idea (R. Horisberger): local coincidence of two sandwiched silicon detector layers (strips + strips or strips + pixels) Lower Sensor 48 A B
Conclusions CMS pixel detector: excellent performance, key to tracking and vertexing LHC long term plan: data-taking until 2030 CMS tracker Phase-1 upgrade: new silicon pixel detector (2016/2017) Preparation for Phase-2 upgrade (2022) 49