CMS_Hardware by xiangpeng


									                 Rutgers Contributions to
                      CMS Hardware
                        Steve Schnetzer
                     for the Rutgers CMS Group

NSF Site Visit

                        1) Prior Accomplishments
                              Forward Pixels

                        2) Ongoing Work
                              Forward Pixels
                              Beams Conditions Monitor (BCM)

                        3) Future Plans
                              Commissioning of Forward Pixels and BCM
                              Pixel Luminosity Telescopes (PLT)
                                (dedicated CMS luminosity monitor)

Our program benefits enormously              Robert Stone  Senior Scientist
from the excellent staff, that are                             fully supported by Rutgers
fully or substantially supported by          Edward Bartz  Electrical Engineer
Rutgers University.                                            highly subsidized by Rutgers

                                             John Doroshenko  Computer Scientist
                                                                     fully supported by Rutgers

                 Others contributing: Alick Macpherson, Dmitry Hits (current)
NSF Site Visit                        Lalith Perera, Steven Worm (former)                         2
                            Prior Accomplishments

      • Rutgers joined CMS in 1997
      • Members of the US-CMS
        Forward Pixel Group
      • Lead US group for pixel readout
        electronics (1997-2005)
                                                             the CMS pixel system
      • As Fermilab came up to strength,
        turned, over responsibility for high
        density circuits (VHDI, HDI) and Port Card
      • Sole responsibility for two major projects:
                    Token Bit Manager (TBM) Chip
                    Front End Controller (FEC)
      • Only US group to contribute to both barrel and forward pixel electronics

      • Close collaboration and partnership with Roland Horisberger‟s group at PSI
      • Developed world-class chip design capability from scratch
      • Work held in high regard by both forward and barrel pixel management
NSF Site Visit
                                   Overview of Pixel Electronics

                                                                           Rutgers contributions in scarlet
                             Pixel FEC
                                                            Tracker FEC
                               Digital                     CCU System                                (6 Fibers)
                             Opto Hybrid
                                                                                         Analog              Analog
    RClk         RDa                     PLL                                           Opto Hybrid         Opto Hybrid
                                     CLK     L1A
                                                                                        Alt Chip              Alt Chip
                                  Delay 25                           Blade 1 Side A
                            SDa      CLK     L1A                      TBM
                                                                      TBM & ROCs
                                                                            Side B
                       GateKeeper                                     TBM
                                                                      TBM & ROCs

                           SDa     CLK     L1A                       Blade 2 Side A
     RClk RDa
                                                                      TBM & ROCs
                             3 Fan Out Chips                                Side B
                                                                      TBM & ROCs
                                 2 Fan In Chips
                                 1 Reset Chip                        Blade 3 Side A
                                                                          TBM & ROCs
                                    5 Differential Pairs                    Side B
                                                                      TBM & ROCs
                                  1 Single Ended Reset
NSF Site Visit                                Each                                                                       4
                       Token Bit Manager (TBM)

                                                      Analog              Analog
    • Custom rad-hard IC                              Readout             Power

    • Orchestrates readout of group of            Stacks
      readout chips (ROC‟s)
    • Contains front-end control network hub                              Sections

                                                                                     4.4 mm

    • Developed in IBM 0.25 m process
    • Five prototype submissions

    • Production chips delivered in 2005                     Regulators

    • Meets all design specifications                       3.2 mm

    • TBM‟s now being assembled into detector modules

    • Project successfully completed

NSF Site Visit
                                      TBM Functions

 •     Distribute to Pixel Readout Chip’s (ROC)
         – L1A triggers (2 Outputs/TBM)                                      L1A
         – Clock (2 Outputs/TBM)                              Clock
                                                                                                      To Flash
         – Switchable to single TBM mode

 •     Control readout through token pass
                                                                            TBM TBM                       +
 •     Stack triggers awaiting token pass

         – 32 event deep
         – No readout after 15 deep               Token Out           CLK      L1           Token In

 •     Write header and trailer (2 bit analog
       encoded digital)
         – 8 bit Event Number (Header)
         – 8 bit Error status word (Trailer)
                                                    ROC               ROC             ROC
 •     Respond to signals on the L1A line
         – Resets
         – Calibration Signals for the ROC                            Analog Output

         – Sync Triggers
NSF Site Visit
                                            TBM Chip

    • Two Token Bit Manager’s
            – Each TBM Controls one ROC Chain         Control Analog 40 MHz    Analog
                                                              Output Clock L1A Output
            – 8 to 24 ROCs per TBM                   Commands

    • Control Network Hub
            – Port addressing for control commands      HUB
            – 6 to 14 Dual TBMs per Control Link

                                                         Command           L1
            – 120 Links in the Pixel Detector              Ports

                 (Barrel & Forward)

    • Operating Conditions                                         TBM          TBM
            – Clock/Data Rate = 40 MHz
            – Temperature = ~ -10 degree C                                        Output
            – Supply = 2.5V, Low Power Design
            – Fluence = 4 x 1014 cm-2 per LHC year
                                                     Chain of Read Chain of Read
                 Full luminosity @ 4 cm radius
                                                       Out Chips     Out Chips

    ~1100 TBMs & 20,000 ROCs (Barrel & Forward)

NSF Site Visit
                            Token Bit Manager

TBM sends output analog data
packet for each L1 trigger
                                     TBM with 1 ROC
          • Header
          • ROC hit info                                                            TBM
          • Trailer                Event
                                   Number                 Status Bits

                                                                        Forward Pixel HDI
                                                                        with Ten Readout

Located in high radiation
environment on detector
modules                           Header ID      ROC   Trailer ID


                                                                          used in barrel
                                                                          and forward
                               Barrel HDI with
NSF Site Visit                  16 Readout
                                   Chips                                                    8
                                Control Network Hub

                                                                     Control Link
    TBM responsible for routing of
    of front-end control commands
                                                                               Dual TBM
•     40 MHz Serial Control Link.
•     Up to 32 hubs addressable.                                             Bi Directional
                                                                            Serial Data Port
•     5 Ports per hub.
        – 4 Write Only external  ROCs.
        – 1 Bidirectional internal  TBMs.                             TBM           TBM
•     Functions of addressed hub.
        – Selects addressed port.                                         4 SDa Ports, Output Only
        – Strips off byte containing hub/port address.
          Passes remainder.
        – Reflects output data back to
          Front End Controller.                          ROC     ROC                ROC        ROC
        – Return Hub/Port Address Received.

NSF Site Visit
                      Front End Controller (FEC)

     Custom VME module that sends
         • clocks
         • triggers
         • control commands
     to front-end chips

                  Rutgers proposed modifying existing
                  CMS Strip Tracker FEC
                      • Alternative to development of custom
                        FEC from scratch
                      • Our proposal saved much time, resources
                        and money
                      • One small hardware change to tracker
                        FEC implemented
                      • Rutgers successfully developed custom
                        firmware for mezzanine gate array chip
                      • Project successfully completed (minor
                        debugging changes being implemented)
                      • Rutgers designed FEC under use at Fermilab,
                        Cornell and PSI
NSF Site Visit
                                         The Tracker FEC

                                mezzanine cards
                                custom firmware
                                on gate arrays

•   VME64x Compatible 9U Board.

•   Supports 1 – 8 mFEC Modules.

•   Control information/Pixel Data pass
    through the VME Bus

•   Fast Signals pass through the TTC link.
      (Trigger FPGA will be modified for the Pixel System).
      – L1A
      – Resets
      – Trigger Data Download

    for pixel FEC one extra
    trace needed on motherboard

NSF Site Visit
                             Principle FEC Design Features

         •       Rapid upload of ~ 66 million pixel threshold trim bits
                  – Single event upset refresh
                  – Startup or after general reset
                  – Measured rate

                         Time to load at startup
                             • Barrel (48M pixels)   14 s
                             • Forward (18M pixels)  5 s

         •       Multiple data transfer modes
                  – Normal mode  individual commands
                  – Column mode  transfer of thresholds (full pixel column)
                  – Compressed column mode  all thresholds set to same value

         •       Local Level 1 Trigger and Calibration generation (stand alone operation).
         •       Synchronize of data transfer to the TTC global trigger and timing system
         •       Automatic error checking of transferred addresses and data

NSF Site Visit
                                           Other Projects
                                                                                       testbeam telescope
• System test facilities (Bartz)
        designed and built forward pixel test systems                          silicon strips   pixel   silicon strips
        primary test facilities
                                                         120 Gev Proton
            • used at Rutgers, Fermilab, Purdue,
              UC Davis, Cornell, Nebraska, Kansas, PSI
            • testing and calibration of readout chips
              and sensors
            • testing of production modules
        evolving into final pixel readout system
• Graphical user interface (Doroshenko)
        designed and built graphical use interface (Cosmo)
        being modified to be part of final FEC DAQ software                       example of testbeam result
        critical tool during commissioning
                                                                                     inefficiency map
• Testbeam infrastructure (Perera)
        built test beam telescope

                                                               150um (Column)
            • eight microstrip planes
            • scintillating fiber array trigger
        provided all associated electronics
        provided DAQ, alignment and analysis software
        achieved ~ 1 micron track extrapolation                                                            4 pixel
        test beam runs crucial for selecting                                                               array
         final sensor design
NSF Site Visit
                 •   Fermilab test beam (2004 - 2005)                                                                 13
                                                                                        100um (Row)
                                   On Going Work

                 • Major hardware projects (TBM, FEC) successfully completed

                 • Our group will continue our commitment to these projects
                  and will play an active role in their commissioning through
                  the successful operation of the completed detector.

                 • Our current activities include:
                               Development of pixel DAQ software
                               CMS Beam Radiation Monitor (BRM)

                 • With the large augmentation of the Forward Pixel Group
                  in 2005, production and testing of modules and assembly
                  of the detector is being successfully addressed.

                 • We will turn our focus to the CMS physics program
                   and to work on a proposed CMS luminosity monitor.

NSF Site Visit
                                       Pixel DAQ Software

      • Our group is part of the overall Pixel DAQ effort
                 – about 15 people: Buffalo, Cornell, FNAL, PSI, Rutgers, Vanderbilt
                      • led by (K. Eckland, Buffalo)
                    three from Rutgers: Stone, Doroshenko, Bartz

      • We are responsible for the FEC supervisor software
                 – issue initialization and control commands to TBM’s and ROC’s
                 – each of ~60M pixels needs to have 4-bit trim threshold set
                 – fast download needed for run initialization
                 – four operation modes:
                      •   Debugging / commissioning
                      •   Calibration of all pixels: gain, pedestal, saturation, threshold
                      •   SEU updating of all ROC settings during private orbits (if needed)
                      •   Power up / run initialization

NSF Site Visit
                              Beam Radiation Monitors

                 • Beam Radiation Monitoring (BRM) high CMS priority
                 • Needed for protection of inner detectors (tracker, pixels)
                 • Multiple component system
                      BCM  diamond sensors (monitoring beam induced current)
                      BSC  scintillation counters (beam halo, abort gap)
                      RADMON  passive monitoring in cavern

                 • Primary measurement
                      Beam induced current in array of
                       diamond sensors near beam line

                 • Rutgers responsibilities
                      BRM Project led by Alick Macpherson
                          (joint Rutgers/CERN Research Associate)
                      Depositing electrodes on BCM diamonds
                      Characterizing BCM diamond sensors

                 • Project on course for operational system in „07

NSF Site Visit
                                      BCM Functionality

• Use synthetic diamond sensors as simple Flux/rate monitors
        – 2003-2006: Prototypes tested in CDF, CERN East Hall, KAZ (Karlsruhe), PSI

• Generate warnings/alarms/beam aborts based on sustained above-
  threshold readings or very rapid rate of increase towards threshold
        –    Deadtime free monitoring
        –    CMS configurable thresholds
        –    Rolling history of conditions available online,
        –    Initiate CMS beam abort if conditions above abort threshold

        – CMS:
                 • Direct warning/alarm, monitoring, diagnostic and postmortem information
                 • User selectable time scales over which above-threshold conditions assessed

        – LHC:
                 • Provide full monitoring + beam dump post-mortem reporting to CCC
                 • Close interaction with LHC operations and LHC machine protection group

NSF Site Visit
                            BCM Locations
                                                         CMS BCM Units
                                              1   BCM1L: Leakage current monitor
3                                                 Location: z=±1.9m, r=4.5cm
                                                  4 stations in 
                                                  Sensor: 1cm2 PCVD Diamond
                                                  Readout: 100kHz
                                                   No front end electronics

                                              2   BCM1F: Fast BCM unit
                                                  Location: z=±1.9m, r=4.3cm
                                                  4 stations in 
                                                  Sensor: Single Crystal Diamond
                                                  Electronics: Analog+ optical
                                                  Readout: bunch by bunch (Asynch)

                                            3 BCM2: Leakage current monitor
                                              Location: z=± 14.4m, r=29cm, 5cm
                                              8 stations in 
                                       1      Sensor: 1cm2 PCVD Diamond
                                               Readout: ~20kHz
                                              Sensors shielded from IP
2 Sensor Locations, 3 Monitoring   Timescales Off detectors electronics
 NSF Site Visit
                                  Future Projects

             • Active role in installation and commissioning
               of the Forward Pixel detector.

             • Continue to lead development of pixel control DAQ software

             • Develop dedicated CMS luminosity monitor
                   Presented an original concept for a dedicated CMS luminosity
                   monitor at an October 2004 CMS luminosity workshop.
                   Evolved into the Pixel Luminosity Telescopes (PLT) proposal
                   Concept: Array of small angle telescopes
                             based on diamond pixel detectors.
                   Collaboration with Princeton, UC Davis, CERN
                   Seeking CMS authorization in Spring „07
                   Installation for 2008 LHC run

NSF Site Visit
                                 PLT Overview

 Measure relative luminosity bunch-by-bunch
      •   Small angle (~1o) pointing telescopes

      • Three planes of single-drystal CVD diamond
        sensors (4 mm x 4 mm) active area
      • Diamond pixels bump bonded to CMS pixel ROC
      • Form 3-fold coincidence from ROC fast out signal

      • Eight telescopes per side
      • Located at    r ~ 5 cm, z ~ 1.7 m
      • Total length 9 cm

      Count 3-fold coincidences
      for each bunch crossing

NSF Site Visit
                           Location of Telescopes

    • End of Be section of beam pipe
      (~ 1.7 m from IP)

    • Just outside of beam pipe
      (~ 5 cm from beam line)

NSF Site Visit                            IP        21
11/29/06           Pixel Service Tube
                 Rutgers Work on CVD Diamond Sensors

  We have a very long history in
  CVD diamond sensor development

                 Our group pioneered the development
                 rad-hard CVD diamond sensors in 1988.

                 In 1993, we became charter members of
                 the CERN-based RD42 Collaboration to
                 develop diamond sensors.
                 From 1998-2003, our work on diamond
                 sensors was supported by a Major Research
                 Instrumentation grant from NSF.
                 Our work on diamond sensor R&D has led
                 to the high quality diamonds now available
                 from industry that make the PLT possible.
                 The diamond pixel sensors in the PLT
                 represent the fruition of our nearly 20
NSF Site Visit   years of work on CVD diamond sensors.        22
                      Single Crystal Synthetic Diamond

  Essential component of PLT                          Distribution cleanly separated from zero

         Result of diamond sensor
         development pioneered
         by our group

• Radiation hard (few x 1015 cm-2)
• No need for cooling
• Full charge collection at 0.2 V / m               pedestal    90 Sr Pulse Height (electrons)
           ― 18,000 e― for 480 m diamond
          ― Landau 60% narrower than Si

                                                                                          150 m
• 5 mm x 5 mm single crystal diamonds
          ― detector grade
          ― commercially available
          ― ordered 12 (delivery Jan.)                                          100 m

                                                                    4 mm x 4 mm pixel
NSF Site Visit                                                      pattern for testing      23
                                        Dual Readout of PLT

     Unique dual readout capability
                    1) fast (bunch-by bunch) hit information
                    2) full pixel tracking information

       Luminosity mode:
       Fast output level (each bunch crossing)                     CMS pixel chip has “fast”
                                                                   multiplicity counting built in
                 • 0, 1, 2, 3, … double column hits
                 • individual pixel thresholds adjustable           CMS pixel chip active area
                 • individual pixels can be masked

        Tracking mode:                                                    80 x 52 pixels
                                                                                                 8 mm
                                                                         100 m x 154 m
        Full pixel readout (~ 1kHz)
                 • pixel address and pulse height of each hit
                 • diagnostic of fast out signal      IP
                 • determination of track origin      Scattering
                                                      Beam halo                8 mm
                 • determination of IP location

NSF Site Visit
                       PLT Performance Capabilities

                                                                     Lalith Perera
      numbers for
                                                                     Dmitry Hits

                 • 1.6 tracks per bunch crossing
                      Pythia simulation

                 • Relative bunch-by-bunch luminosity
                      1% measurement each second

                 • Interaction point centroid relative precision
                      100 m radially each second
                      2 mm longitudinally each second      scales as (time)-1/2

                 • Beam in abort gap identified
                      crucial for CMS protection

                 • Beam halo measured
                      horizontal tracks

NSF Site Visit
                              PLT Status and Plans

                                                                      Begin construction
    Goal to have PLT installed for ‘08 run                            by March ‘07

Current status:       •   mechanical support structure designed
                      •   cabling ordered
                      •   necessary chips allocated
                      •   power supply system identified (same as pixel)
                      •   conceptual design of DAQ (based on modified pixel VME module)
                      •   conceptual plan for data logging and publishing (modeled on HF)
                      •   first 12 of 48 diamond sensors ordered

Milestones:       Jan ‘07:      Conditional approval by CMS management
                                    Begin assembly of first 12 detectors

                  March ‘07: PLT Readiness Review
                                    Complete test beam studies of diamond pixels
                                    Begin assembly of first 12 telescope planes
                                    Order additional 24 diamond sensors (12 week lead time)

                  June ’07:     Final endorsement of project
                                    Result from first 12 telescope planes
                                    Order remaining diamond sensors
                                    Begin production of remaining telescope planes
NSF Site Visit    April ‘08:    Delivery of PLT to CERN                                        26
                                     BCM1 Carriage structure

 Designed to accommodate PLT

                                               PLT telescope
                     BCM length= 1076mm           volume

                       Hinge joint


BCM Carriage            BCM Arm

                 •    Only lower arm attached to BCM carriage
                       – Used as push/pull rod to install BCM
                       – Provides cable tray to route out cables
                 •    Hinged arm due to
                       – Dedicated BCM installation rail track: Curved
NSF Site Visit
                       – Proved “longer wheel-base” stability to BCM structure
                          PLT Responsibilities

                 • Rutgers
                        overall leadership
                        diamond sensor characterization
                        deposition of pixel electrodes
                        test fixtures
                        system testing
                        electronics
                        DAQ
                 • Princeton
                        hybrid board
                        production and testing of telescope planes
                        interfacing of data to CMS and LHC
                        luminosity database
                 • UC Davis
                      bump bonding
                 • CERN
                      mechanic structure
                      cables
                      installation
NSF Site Visit
                                        PLT Summary

    Luminosity monitor based on
    diamond pixel telescopes

                 • Unique hybrid (fast coincidence, pixel) readout
                         most versatile luminosity monitor ever for hadron collider
                 • Precision bunch-to-bunch relative luminosity
                 • Precision measurement of IP location
                 • Stable reference for monitoring of detectors and trigger

                 • Measurement of beam hot spots and beam halo
                 • Based largely on existing components
                 • Modest cost ~ $300K
                 • Ready for first physics run
                 • Major milestone in development of radiation-hard detectors
                  for SLHC and other high intensity colliders

                               Long term test of diamond pixel
                               detectors in high radiation environment
                               under actual experimental conditions
NSF Site Visit
                               Funding Needed for the PLT
           •     50% support for Alick (CY „08)       45K   Funding distributed over
           •     Diamond sensors                      60K   two years (50/50). Needs
           •     Metallization                         5K   to start spring „07.
           •     Support for Ed (18 months at 70%)    40K
           •     HDI and Port Card circuits           20K
                                           Total:    170K
                                                            Source: CMS project
  UC Davis                                                  or possibly(?) NSF/DOE
           • Clean room time                          10K
                                                            supplemental requests.
           • Part-time tech                           40K
                                           Total:     50K
           • Hybrid board                             10K
           • Part-time tech                           40K
                                           Total:    50K

 CERN            (already spent > 50K)
          • Crates, electronic modules, PC            30K
          • Optohybrids                                5K
                                            Total:    35K
NSF Site Visit
                     Personnel on CMS Hardware

                 • Steve Schnetzer (faculty)
                 • Robert Stone (Senior Scientist)
                     fully supported by Rutgers

                 • Alick Macpherson (Research Associate)
                     supported 50%

                 • Ed Bartz (Electrical Engineer)
                     highly subsidized by Rutgers

                 • John Doroshenko (Computer Scientist)
                     fully supported by Rutgers

                 • Dmitry Hits (Graduate Student)

                 • Jordan Ledvina (Undergraduate)

                 • Scott Robinson (Undergraduate)
NSF Site Visit
          Publications Related to Our CMS Hardware Work
                                                                                               last five years
                 1) Radiation hard sensors for future tracking applications.
                    W. Adam et al. (RD42 Collaboration), NIM A565:278-283,2006.

                 2) The control and readout system of the CMS barrel pixel detector.
                    D. Kotlinski et al., NIM A565:73-78,2006.

                 3) Development of a CVD diamond beam condition monitor fro CMS.
                    L.Fernandez Hernando et al., NIM A552:183-188,2005.

                 4) The 0.25 m token bit manager chip for the CMS pixel readout.
                    E. Bartz, Proc. 11th Workshop on Electronics for LHC and Future Experiments, Heidelberg, 2005.

                 5) New Developments in CVD diamond for detector applications.
                    W. Adam et al. (RD42 Collaboration), Eur.Phys.J. C33:S1014-S1016,2004.

                 6) The development of diamond tracking detectors for the LHC.
                    W. Adam et al. (RD42 Collaboration), NIM A514:79-86,2003.

                 7) The token bit manager chip for the CMS pixel readout.
                    E. Bartz, Proc. 9th Workshop on Electronics for LHC and Future Experiments, Amsterdam, 2003.

                 8) The CMS pixel detector.
                    S. Schnetzer (for the CMS Pixel Collaboration), NIMA501:153-159,2003.

                 9) CVD diamond pixel development.
                    R. Stone, et al., IEEE Trans.Nucl.Sci. 49:1059-1062,2002.

                 10) Beam test results of the US-0CMS forward pixel detector.
                     M. Atac et al., NIM A488:271-281,2002.

                 11) Pixel readout token bit manager.
                     E. Bartz, Proc. Pixel 2002 International Workshop on Semiconductor Pixel Detectors, Carmel, 2002.

                 12) Diamond pixel detector development.
                     R. Stone, NIMA473:136-139,2001.

                 13) Diamond pixel detectors.
                     S.Schnetzer (for the RD42 Collaboration), Proc. of ICHEP, Osaka, Vol.2:1229-1230,2001.
NSF Site Visit
                     Summary of Effort on CMS Hardware
                 • Leadership of forward pixel electronics
                 • Completed two major projects (TBM, FEC) key
                   elements of both barrel and forward pixel systems
                 • Primary developer of instrumentation for
                   forward pixel test fixtures and system tests

                 • Will continue to play important role in development
                   of pixel DAQ software
                 • Providing leadership of the essential Beam Conditions Monitor
                 • Will play important role in commissioning of the
                   pixel detectors
                 • Originated and leading a proposal for a unique and powerful
                   dedicated luminosity monitor based on diamond sensors

                 • On to exciting CMS physics
                          (see talk by Amit Lath)
NSF Site Visit
                      Eyes on the ILC: Accelerator R&D

 RU Grad Student Tim Koeth lead the
 Fermilab Capture Cavity II effort, the
 1st ILC like cavity at Fermilab, into
 operation at 31.3MV/m (ILC spec is

     FNAL Gained Experience in:
        Input Coupler Conditioning
        Cavity processing
        1.8K Cryogenics
        1.3 GHz High Power RF
        1.3 GHz Low Level RF control
        Clean Room Techniques

 This cavity will upgrade the Fermilab
 A0 Photo Injector to 40MeV and will
 ultimately be used as the injector for
 testing ILC cryomodules.

NSF Site Visit
                           Emittance Exchange at FNAL
15 MeV e- beam from
                            ex < ez           D1        3.9 GHz TM110 Cavity
                           Initial e- bunch

                                                                                               ex > ez
                                                                                        D4     final e- bunch


                                                                                     Beam Dumps

Tim Koeth‟s Thesis Expt
Place a 3.9 GHz TM110 cavity between two
dog-leg magnetic bends to reduce the
momentum spread, thereby reducing the
longitudinal emittance. This comes at the cost
of introducing a transverse betatron amplitude,
thus an increase in transverse emittance. A
complete longitudinal to transverse emittance
exchange should be observed

NSF Site Visit
                 Backup Slides

NSF Site Visit
                 IP Distribution

NSF Site Visit
                       Acceptance Profiles
 IP offset in radial
 direction of one

  1% flat in r to 12 mm @ z=0

  1% flat in z to 320 mm @ r=0

NSF Site Visit
                       Acceptance Profiles

 IP offset in radial
 direction between
 two telescopes

  1% flat in r to 8 mm @ z=0

  1% flat in z to 320 mm @ r=0

NSF Site Visit
                     Acceptance with Four Telescopes
      Four telescopes per side
                 no flat region in r
  Need 8 telescopes per side
 IP offset in radial
 direction between
 two telescopes

NSF Site Visit

To top