The LHCb VELO Upgrade

Size: px

Start display at page:

Download "The LHCb VELO Upgrade"

Robyn Skinner
6 years ago
Views:

1 The LHCb VELO Upgrade Kristof De Bruyn On behalf of the LHCb VELO group 13th Trento Workshop on Advanced Silicon Radiation Detectors Munich February 19th, 2018 Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

2 The LHCb Detector JNST 3 (2008) S08005 Physics Programme CP Violation Rare Decays Forward EW Physics Spectroscopy ons... Forward arm spectrometer to study b- and c-hadron decays Pseudo-rapidity coverage: 2 < η < 5 High Precision Physics Rely on identifying displaced vertices VELO plays crucial role Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

ntro VeloPix Cooling The Current Vertex Locator (VELO) Upgrade 2 arxiv:1405.

CO2 cooling Centres around P based on self-measured vertices Closest active sensor strip: 8.

3 ntro VeloPix Cooling The Current Vertex Locator (VELO) Upgrade 2 arxiv: JNST 9 (2014) P modules, each with R and φ sensor Moves in and out at every fill 300µm thick n-in-n type silicon strips Evaporative CO2 cooling Centres around P based on self-measured vertices Closest active sensor strip: 8.2mm away from P Thin RF foil to separate vacua and protect against beam charge Tough radiation environment for Silicon detector ( ) Except 2, who are n-in-p sensors Kristof De Bruyn (CERN) The LHCb VELO Upgrade See talk M. Maciejewski Trento Workshop 19/02/ / 22

Upgrade Plans for the LHCb Detector Upgrade 1: Move to a full detector readout at 40 MHz (currently 1MHz after hardware trigger) ncrease the number of visible interactions from 1.1 to 5.

4 Upgrade Plans for the LHCb Detector Upgrade 1: Move to a full detector readout at 40 MHz (currently 1MHz after hardware trigger) ncrease the number of visible interactions from 1.1 to 5.2 Goal: collect total luminosity of 50 fb 1 in Run 3+4 Need to be (radiation) Harder, Better, Faster, Stronger Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

ntro VeloPix Cooling Main Challenges for the VELO Upgrade 1 Upgrade 2 VELO Upgrade TDR [CDS 1624070] Side-by-Side Comparison: Readout Channels/Module (Tot) Max Radiation Dose Visible nteractions (µ)

5 ntro VeloPix Cooling Main Challenges for the VELO Upgrade 1 Upgrade 2 VELO Upgrade TDR [CDS ] Side-by-Side Comparison: Readout Channels/Module (Tot) Max Radiation Dose Visible nteractions (µ) Operating Temperature Max High Voltage Kristof De Bruyn (CERN) VELO VELO Upgrade 1 MHz 4096 (172k) neq /cm2 (9.6 fb 1 ) C 500V 40 MHz 790k (40.9M) neq /cm2 (50 fb 1 ) C 1000V The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

The New VELO Upgrade Module Grey: Sensor Tiles Cyan: Silicon microchannel substrate Brown: Hybrids + Readout cables Grey: CO 2 supply pipes Black: Module support A New Layout: Move from Silicon

6 The New VELO Upgrade Module Grey: Sensor Tiles Cyan: Silicon microchannel substrate Brown: Hybrids + Readout cables Grey: CO 2 supply pipes Black: Module support A New Layout: Move from Silicon Strips to Pixel technology Replace 42 C -shaped modules with 52 L -shaped modules Each module: 4 sensor Tiles (2 front + 2 back) 3 VeloPix ASCs per sensor Closest active element: 5.1mm from P Evaporative CO 2 cooling through microchannels etched into silicon substrate Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

7 VeloPix Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

pixel matrix with 55 55 µm 2 pixels General Triggerless, binary readout (time-stamped hits) Peak rate: 800 Mhits/s/ASC or 20.

8 VeloPix Specifications The Sensor Manufactured by Hamamatsu (Japan) n-in-p sensor with 200µm thickness Elongated pixels in between ASCs 450µm wide guard ring DRE-etched, rounded corners (foil clearance) The ASC Derived from TimePix3 130nm CMOS Manufactured by TSMC (Taiwan) pixel matrix with µm 2 pixels General Triggerless, binary readout (time-stamped hits) Peak rate: 800 Mhits/s/ASC or Gbps Highly non-uniform irradiation ( r 2.1 ) Average data rate per ASC per event Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

The Sensor: Charge Collection Extensive testing before/after irradiation Using TimePix3 telescope at the SPS testbeam Charge Collection Efficiency Results after full fluence irradiation Can reach

9 The Sensor: Charge Collection Extensive testing before/after irradiation Using TimePix3 telescope at the SPS testbeam Charge Collection Efficiency Results after full fluence irradiation Can reach 6000 e target with bias voltage < 1000V Cluster Finding Efficiency Efficiency - rradiated [µm] y [µm] y HPK, 39 µm implant S17, 500 V x [µm] x [µm] Results after full fluence irradiation Can reach efficiencies 0.92 > 99% 30 Micron (n-on-p), 36 µm 30 Micron (n-on-n), 36 µm implant 0.9 implant At 1000V, S23, 500 V also in the 20 corners S27, 500 V Production choice: µm implants x [µm] x [µm] 15/12/2016 Emma Buchanan Sensor PRR [µm] y [µm] y HPK, 35 µm implant S22, 500 V Un JS Sma seen throu pixe! Ther the d wher effici othe Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

10 ntro 20 VeloPix Cooling Upgrade 2 Micron n-on-n Spatial 18 Resolution - Binary - Data The Sensor: 16 Spacial Resolution Binary resolution 14 for pre-irradiated sensors using Timepix3 data HPK n-on-p Data 12 Simulation Reminder of the simulation Micron n-on-p S6 Micron n-on-n 18 8 S S11 4 S23 14 S25 2 S30 12 Reminder of the simulation S6 Angle [degrees] 8 S8 6 - S11 The simulation and data compare well - 4 S23 The best resolution is seen at ~8.5 µm - S25 Poorer resolution can LHCb be seen Preliminary for the 2 S30 thinner n-on-n sensor 0 [µm] Binary σ x [µm] Binary σ x Angle [degrees] 15/12/2016 Emma Buchanan Sensor PRR 18 -Results The simulation pre-irradiation: and data Best compare resolution: well 8.5µm - The best resolution is seen at ~8.5 µm The - Poorer measured resolution resolution can be matches seen for well the with expectations from simulation Results thinner post-irradiation: n-on-n sensor Best resolution: 9.5µm Follow expectations, but no direct comparison with simulation available (S30 is a thinner sensor) Micron n-on-p 15/12/2016 Emma Buchanan Sensor PRR 18 Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

searching for sub-dominant effects dentified ssues: 1 Single Event Latch-ups (SEL) 2 Spurious resets due to SEU 3 Continuous data from

11 The ASC: rradiation Campaign for VeloPix1 X-ray irradiation up to 500 Mrad Testing for Single Event Upsets (SEU) at Louvain-la-Neuve heavy ion facility 4 campaigns to identify all problems Effects have different probability to occur Masked regions to continue searching for sub-dominant effects dentified ssues: 1 Single Event Latch-ups (SEL) 2 Spurious resets due to SEU 3 Continuous data from pixel column FFO after SEU Solved in 2nd generation: VeloPix2 Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

12 Single Event Latch-Up What? Clear current increase with every SEL Recovery only possible by power cycling the chip Temperature dependent effect: Not seen < 10 C New with TSMC, not observed with previous manufacturer (BM) Why? Local short-circuit on the power lines after charge deposit Caused by high resistance of n-well and p-substrate Depends on distance between the contacts and the n + source/p + drain Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

absorption of near R photons shot at back of ASC Observed SEL (Red points) 3µm Scan CAD prediction of SEL susceptibility Distance between

13 ntro VeloPix Cooling Upgrade 2 Single Event Latch-Up Montpellier Laser Facility: Confirmation of the effect and localise the cause 30mA current rise per SEL Allow precise scan locations: steps of 3µm (but step size too small to achieve full coverage) Charge injection through absorption of near R photons shot at back of ASC Observed SEL (Red points) 3µm Scan CAD prediction of SEL susceptibility Distance between contacts SEL occurrences match regions with large distances between contacts Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

14 Resubmission Modifications for VeloPix2 Counter Radiation Effects Reduced distance between p + and n + contacts No current increase observed Triple redundancy of receiver No more spurious resets observed Redesign of pixel column read-out logic No more FFO issues observed Counter Transmitter Jitter mprovements to PLL design Reduced internal PLL jitter and better eye diagram Larger time spread (skew) for pixel matrix readout Reduce power spikes nternal clock signal is less sensitive to matrix activity Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

15 ntro VeloPix Cooling Upgrade 2 Developing the Quality Control Test Procedures Comparison no ffi ci al All S LH Cb U Abs very som sli 29 V curves V curve testing rig LH Cb U no ffi ci al Sensor Tile & Readout Boards Sensor exposure to Sr Pixel [0, 1] : 5.53 Pixel [11, 1] : Pixel [12, 1] : 6.09 al Pixel [14, 1] : 7.57 no ffi ci 40 U 30 Cb 20 Kristof De Bruyn (CERN) V curve testing rig The LHCb VELO Upgrade LH 10 MiniDAQ readout system Pixel noise scans Trento Workshop 19/02/ / 22

16 Evaporative CO 2 Cooling Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

Evaporative CO 2 Cooling through Silicon Micro Channels Advantages: CO 2 has high latent heat capacity Radiation hard Reduced material budget (incl.

17 Evaporative CO 2 Cooling through Silicon Micro Channels Advantages: CO 2 has high latent heat capacity Radiation hard Reduced material budget (incl. overhanging sensor) Same thermal expansion coefficient Challenge: Avoid Thermal Runaway Keep sensor tips at 20 C to minimise radiation damage Liquid CO 2 arrives at 30 C and 20 bar Power consumption: 30W/module Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

Racetrack Design 19 channels Size: 120 200µm

inlet and outlet Channel Output Channel nput

channels done by connector Kristof De Bruyn

18 Racetrack Design 19 channels Size: µm 2 Courtesy CEA-LET Etched in 500µm silicon substrate (Afterwards thinned) Final thickness: 500µm Each channel has its own inlet and outlet Channel Output Channel nput Channel Restriction Distribution among channels done by connector Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

Connector Soldering Challenges: Status: Must withstand

Must be leak tight (sensor in vacuum) Must remain planar

process Went through many design iterations First

19 Connector Soldering Challenges: Status: Must withstand pressures up to 186 bar (nominal operation = 20 bar) Must be leak tight (sensor in vacuum) Must remain planar Avoid use of Flux ( corrosion) Difficult and complex process Went through many design iterations First successful soldering on September 1st, 2017 Passed all requirements but further testing ongoing Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

20 Upgrade 2 Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

21 Start Thinking About Upgrade 2 LHCb upgrade aims to collect 50 fb 1 by end 2028 Should maximally exploit the High-Luminosity LHC Considering a Phase- upgrade during 2029 shutdown Goal: Collect 300 fb 1 with 50 visible interaction per bunch crossing Wish List for VELO Again need to be (radiation) Harder, Better, Faster, Stronger To be based on TimePix4 or later Reduced pixel size, thinner sensors and thinner ASCs (move to 65nm technology) Add pixel hit timing information to assign tracks to correct collision vertex Should reach 200ps resolution SuperVeloPix needs to be radiation hard up to n eq/cm 2... or easily replaceable (Like NA62 GigaTracker) Keep low material budget Consider removing the RF foil (if LHC allows) Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

22 Summary Preparing the Upgrade of LHCb s VELO detector Well on schedule: Moving into production phase To be installed by 2020 Key Features: 40 MHz triggerless, binary readout Pixel sensors with 55 55µm 2 pitch Sensors placed closer to the P and better segmented ASCs have Gbps maximum output bandwidth Evaporative CO 2 cooling through microchannels etched into silicon substrate Kristof De Bruyn (CERN) The LHCb VELO Upgrade Trento Workshop 19/02/ / 22

PoS(VERTEX2015)008. The LHCb VELO upgrade. Sophie Elizabeth Richards. University of Bristol

PoS(VERTEX2015)008. The LHCb VELO upgrade. Sophie Elizabeth Richards. University of Bristol University of Bristol E-mail: sophie.richards@bristol.ac.uk The upgrade of the LHCb experiment is planned for beginning of 2019 unitl the end of 2020. It will transform the experiment to a trigger-less