Document Sample
AFP-review_070411 Powered By Docstoc
					Tracking detector for AFP Upgrade Project

                      Marek Taševský
        (with help of Petr Šícho and Václav Vrba)
    Institute of Physics, Academy of Sciences, Prague
          FD Review Meeting, CERN - 07/04 2011
                 Open questions
                 First cost estimates
                          AFP = ATLAS Forward Protons
 Proton leaves the interaction untouched, travels through LHC optics and is detected at 220 m

4 stations (-224m, -216m, 216m, 224m): 6 layers per station -> 24 FE-I4 chips & sensors
With spare: ~50 FE-I4 chips                                                                2
                               Deflected protons

Diffractive beam-1 protons deflected at 220m:

                                                 protons deflect
                                                horizontally in a
  BEAM 1
                                                region ~2x2 cm2
                                                  outwards the

                                                1) Only horizontal detectors needed

                                                2) Region of interest is ~2x2 cm2.
                                                 This can be fully covered by exactly
                                                 one FE-I4 chip (16.8mm x 20.2mm)

    10-15 σbeam            LHC appertures
Layout of tracker including beam pipe

                                    Tracker design

FE-I3 design: 2 sensors per layer

                              Position detectors
             The same requirements for 220 and 420 m regions:
Close to the beam => edgeless detectors
High lumi operation => radiation hard                             Silicon detectors
Mass resolution of 2-3% => 10-15 μm precision
Suppress pile-up => add fast timing det.
ATLAS, 1.5 mm (220) and 5 mm (420) from beam

                                               Reconstruct the central mass from
                                               the two tagged protons (from their
                                               trajectories and incorporating experim.

                                               Beam en.smearing σE= 0.77 GeV
                                               Beam spot smearing σx,y = 10 μm
                                               Detector x-position resol. σx = 10μm
                                               Detector angular resolution = 1, 2 μrad
                                  Options for Si sensors
The IBL Upgrade project has similar requirements on sensors as AFP:
1) Radiation hardness (sustain the radiation dose of 5x1015n/cm2)
2) Maximum bias voltage 1000 V
3) Maximum inactive edge 450 μm

 AFP is open to all viable solutions and closely watching the IBL decision to be made in June.

Planar Si n-on-n: - used in current ATLAS pixel detector which functions very well
                  - proven technology, double-side processing, slim with inactive edge ~250 μm
                  [ON Smiconductors (Czech rep.) delivered ~50% of all pixel sensors
                  Prague Institute of Physics (PIP) tested ~30% of all pixel sensors]

Planar Si n-on-p: - single-side processing, slim with inactive edge ~450 μm
                 - used by most of LHC upgrade projects (Si strips)
                 PIP participates in those projects -> natural to
                 make use of this expertise and money for AFP project

3D Silicon:      - excellent in the small inactive edge and radiation tolerance
                 - not proven in a real experiment
        Insertable B-layer
                                    M. Havranek, Pixel det.

, modules, 2FE-4 chips per module

                                 IBL and AFP synergy
  - Using the same detection technique as the IBL project would substantially reduce the manpower
    and funds needed.
  - AFP aims to install first movable beam pipes with Si pixel detectors during the 2013-2014

  Possible areas of know-how and technology shares:
  1) Sensors
  2) Readout chip
  3) Bump bonding
  4) Module assembly and testing
  5) Power supplies
  6) External services
  7) Detector Control System
  8) Off-sensor readout electronics
  9) Cooling plant
  10) Integration of the whole system
         IBL and AFP can profit from the need to solve similar problems.
         Let’s communicate more closely.
AFP was presented on the last IBL General meeting at CERN, 16.02.2011
Very positive feedback!
Module of the current ATLAS pixel detector

Sensor of the current ATLAS pixel detector
                                 M. Havranek,Pixel det.


                                   Sensor design
New design in progress (PIP group):

n-implantation to p-type bulk silicon
n-in-p sensor – one-side lithography (being tested by all LHC experiments and RD50)

  Standard inactive edge ~ 450μm.
  This may be reduced by modifying guard rings area and using finer cutting methods

   4 stations: 6 layers per station -> 24 FE-I4 chips & sensors (~50 with spare)

   Total number of channels: 24x80x336 = 645120
Development motivated by low edge design


- Short collection distance (faster collection)
- Low depletion voltage
- Active edge (inactive edge ~ 5 μm)
- High radiation tolerance

Tracker design for AFP420 with 3D-Si

                                       Read-out chips

FE-I3:                                                                    FE-I4:

- ATLAS pixel det.                                                        - ATLAS pixel det. – IBL
- Process: 250 nm IBM CMOS                                                - Process: IBM 130 nm
- Threshold tuning                                                        - Largest R/O chip in HE
- Time over threshold 8 bit                                               - Reduced dead space:
- Leakage current compensation                                            FE-I3: 26% → FE-I4:11%

     Radiation dose close to the beam at L=1034cm-2s-1 is 1015p/cm2 per year (30 MRAD)

         The size of FE-I4 chip is similar to the region of interest for the AFP physics.
         This significantly simplifies the AFP tracker design!
                   Radiation at outer tunnel wall
                       Annual levels for nominal LHC   T. Wijnands, TS Dep.

Radiation levels
in perspective

Box diagram of all services

                                          half stave
 End     2 sensors         2 sensors    2 sensors         2 sensors    2 sensors         2 sensors
Stave   FE-I4 FE-I4    FE-I4 FE-I4     FE-I4 FE-I4     FE-I4 FE-I4    FE-I4 FE-I4    FE-I4 FE-I4
                     NTC                            NTC                            NTC

                                         1 half stave
             DCS                         • readout by 1 opto board
                                         • 6 detector modules
                                         • 3 bi-modules

                                         1 DCS bi-module:
                                          1 HV for depletion of 2 (4) sensors
                                          1 LV for 4 front ends
                                          1 NTC for temperature monitoring

IBL DCS Overview
                                S. Kersten, IBL

                   •   Routing and grouping of
                       services at patch panels
                       determine operation units

                   •   EoS/PP1 transparent for DCS

                   •   Cooling not under control of IBL

                       AFP following the same
                       layout ?

                                                                           S. Kersten, IBL

                                               for IBL
                   starting   destination

type2 detector     PP1        PP2              4 HV, 4 LV, 4 Tmod, 4 Env

HV                 PP2        Iseg (USA)       4 HV

LV                 PP2        LV - PP4 (USA)   4 LV

Tmod               PP2        BBIM (USA)       4 Tmod

Env                PP2        BBM (PP3)        4 Env

type2 opto power   Opto box   PP2              1 Vvdc

type4 opto power   PP2        SC-OL (USA)      1 Vvdc

Opto services      Opto box   SC-OL (USA)      Vpin, Topto, reset, Viset

Tpp2               PP2        BBIM (USA)       4 Temp, 2 cable/PP2 crate

PP2 aux. power     PP2        USA              1 Lvplus, 1 LVminus

CAN                PP2        USA              1 CAN cable/PP2 platform

CAN                PP3        USA              1 CAN cable

                                          DCS Services
AFP tracking system = 4 detector stations
2 detector stations = 12 Pixel-Single-Chip-Assemblies (12sensors+12FE-I4 chips) ~ ½ IBL stave
                                                                                   Ideas of P. Sicho,
LV cables = 3 pairs leading to PP2 (voltage regulator) , length =10-12 m           Pixel det.
           3 pairs leading from PP2 to USA15 (Wiener power supply), it should be robust
            to have the same voltage drop over 300 m cable as pixels have over 70 m cable

Opto-board power supply & control (4 pairs) lead to optical link power supply (SCOLink
Crate), length of cable = 10-12 m

HV cables = 3 cable pairs leading to USA15 ISEG module,
       length of cable = 300 m! (pixel detector uses 50m)

Optical link (3 fibres?), length = 300 m – should not be a problem

CAN bus cables (to control PP2 and SCOLink), lentgh = 300 m should not be a problem

NTC + ENV cables (temperature and humidity monitoring), 8 cable pairs (?)

Interlock cables (USA15 – PP2 & SCOLink)
                           DCS Services in summary
•   LHC tunnel:
•   24 Pixel-Single-Chip-Assembiles (24 sensors and 24 FE-I4 chips)
•   2 optoboards (for el-opt interface), price =                       Ideas of P. Sicho,
•   2 simplified SCOLink crate, price = 400-500 Eur                    Pixel det.
•   2 simplified PP2 voltage regulator crate, price = 1000 Eur
•   Standard cabling and optical cables

•   USA15:
•   1 HV modul ISEG in pixel crate/rack, price = 5000 Eur
•   1 LV Wiener crate in pixel rack, price = 5000 Eur
•   1 interlock electronics
•   1 electronics for temperature and humidity monitoring

•   As in IBL, MCCs services are going to be distributed on the chip

Optical links

                C. DaVia, FP420


                             P. Morettini, IBL

           New VME based ROD/BOC
      - Commands to the ROD via Ethernet only

                                          Diffractive Trigger
         - 224 m    - 216 m                                                                          Right
                                                  ATLAS detector                      Left
           xA              xB                                                      Primitives      Primitives
     T                                                jet

 Front end                                                                                                 T
PA                                                                                                         R
SH                 xA xB                                jet
                                                                                                               +730 ns

                                xA - xB =0                                   xD - xC =0
                                               Forward Trigger Logic
                    T                                                                         T                1,0 sec
                                 +850 ns              8 bits
  Pipeline                      (air cable)
 (6.4 sec)                                                            ATLAS Standard
                                                   L1 Central
                                                    Trigger            2 Jets with
                                                                       Pt > 40 Gev/c
                                                Processor (CTP)                                           2,0 sec
                       Data                                            Max 75 KHz
                   concentrator                       ROD                                                 2,5sec
                      Alcove                                ROL
                                                                                   Refined Jet Pt cut
                                                      ROS              HLT         Vertice within millimeter
                                                                                    time < 5 to 10 psec
P. LeDu, AFP
                            Timing and Data flow (Rev nov08)
          0 ns                   733 ns                                                                            P. LeDu, AFP
Proton @ RP        Flight path
                                                     ASIC and FPGA
                                                     Average Rate = 4,16 Gbit/sec
                                           1024 ns                                             –Bit 1 : One Track left
                                                     (11ns through cable to Alcove)            –Bit 2 :More than one track left
Pretrigger Data available           Processing                                                 –Bit 3 :One track Right
    @ 220 m(Alcove)                                                                            –Bit 4 :More than One track right
                                                      ALCOVE CTA crate                        –Bit 5 :One track left AND one track left
         Detector response 2 ns                       Matching 2 strips                        within a time window.
         RO response ??? ns                           Trigger primitives 8 bit per Xing        –Bit 6-7-8 : for eventual other information
         20 ms cable          80 ns
         Primitives Processing 50 ns
                                                                          1921 ns
                                                                                      8 bit/BX x 40 MHz = 0,32 Gb/s
     Triigger Data @ ATLAS CTP                              Cable
                                                                                      80 bit @ 10 GB/s - 880 transfert time

                                                 LVL1 ACCEPT (75 KHz)            Processing                          588 ns
                                                                                                2500 ns
                                                                    RPs data @ ROD
                     Data Production to ROD
                     4 events x(20 detectors x10 bit word stored in the pipeline)
                     4 events x 1 MCP-PMT detector x (6 bit address + 8 bit fine timing)
                     Total per LV1 Accet = 796 bit
                     Total x 75 KHz =60 Mb/s

              Implementation and read-out block diagram
               MCPs             PB                  PA                        //                    IP
                                                                                                                       RP Right
                                                                         Trigger Left bits                               bits
   Detector                      X X                X X X
    ASIC                       XX XX               X X                                           ATLAS
                                                                                 1Cable           LVL1           P. LeDu, AFP
                                           MCC                                  L1 ACCEPT
                                                  FPGA                              DATA
     Local                     FPGA
                 RO           Or cables          Or cables                                        RODs
     Logic                                                                         8 fiberss

                                                             < 1 Gb/s

     20 m                                                      75 KHz
     Cables                   CTA crate                        60 Mb/s
                                                                                               Reference clock
               Trigger primitives Local concentrator          160 MHz CLK (fiber)
Shielded Alcove     control logic & Monitoring
                                                                                                                     USA 15
                                                                                                LHC CLK
                                                                                                LHC CLK

     3D Pixels                           Data                                                                    H
                                     Concentrator                       ROD                    ROS
     detector                                                            ROD                                     L
                                     Local Trigger                                                               T
      Timing                          Processing                                                                 &
     detector                                                                                  ROS
                                         Data                                                                    D
                                     Concentrator                       ROD
      AFP220                         ALCOVE                             USA15                   ACR              Q        28
       Tracking detector – Tasks and PIP contribution
Areas where the PIP engineer group is ready to contribute (x = manpower, f = financial)

                                                 Break-down of manpower possibilities of the PIP group

      Covered by CVUT (V. Vacek)                         In summary: Year # people        FTE
                                                                     2011   8             1.3
                                                                     2012    7            1.2
                                                                     2013    5            1.0

            First cost estimates
IBL sensors (kCHF)                              AFP220 sensors (kCHF)

            Tracking detectors (from IBL TDR)


• Decision about the sensor option has to be made soon

• AFP is open to all viable solutions and is closely watching the IBL decision

• Prague group has expertise in pixel detectors, prefers and has money
  for n-on-p option

• First rough estimates of the cost of the whole tracking detector presented

• Further collaborators are welcomed!


L1 AFP Trigger time to CTP

   Acceptances depend heavily on the distance from the beam and dead space!
   (if protons hit the dead space in 220 station, they are lost for 420 measurement)
   Acceptance for 420+420, 420+220 and 220+220. Numbers mean total distances.
   420 at 6 mm everywhere, 220 varying from 2mm to 7mm


                                                          Dead space = 1.1mm

                                                       220 at 2mm obstructs the tracking at 420 !

                                                         Dead space of 1.1 mm is too cautious.
                                                         Peter will make this plot for dead space
                                                         of 0.5mm.
                                                         In the following analyses, dead space=0mm

15 σbeam ~ 1.5 mm
(thin window (400μm) + safety offset (300μm) + edge (5μm) + alignment) ~ 0.7 mm
Conservative guess of distance between beam center and first sensor : 2.2 mm                   34

Shared By: