Tracker DAQ by dffhrtcv3

VIEWS: 4 PAGES: 13

									Tracker DAQ

Makoto Yoshida
  (Osaka Univ.)

MICE meeting @ Frascati
      2005/6/27
                   KEK beam test
   Test for the SciFi tracker with 1T solenoid
    magnetic field
   p/p/m test beam line in KEK-PS
   Slow extraction with 2-sec flat top in 4-sec cycle
       Typically, 1000 particles in a beam spill
   Newly developed Cryostat with Cryocooler
       2 VLPC cassettes embedded
   Newly developed buffer module, VLSB
       4 VLSB modules for 4 AFEII boards
    DAQ Software for KEK beam test
   Will use Unix-based DAQ programs
       “Unidaq” developed by KEK
          http://www-online.kek.jp/~online/Unidaq/
          A lot of experience at the KEK beam tests

          Work on Linux PC

       Network distributed system
          Transport data via NOVA buffer manager
          Processes can be distributed among PCs on the
           network.
                An Event Builder
                Collectors for TOF etc. and VLPC
Linux PC
                               System Overview
                                                                           AFE II (L)   8x64 ch
            PCI-VME
                                Slow Control                                   VLPC Cassette #1




                                                         VLPC backplane
            1553                                                                        8x64 ch   1024 ch
                                                                           AFE II (R)
            SASeq #1
                                    AFE II Control
                                                                                        8x64 ch
            SASeq #2                                                       AFE II (L)
VME BUS




     6                                                                         VLPC Cassette #2
            VME ADC/TDC
    U




                                                                                                  1024 ch
                                       4                                   AFE II (R)   8x64 ch
            LVDS-VME #1
                                           4                                                VLPC Cryostat
            LVDS-VME #2
                                               4                          Serialized ADC DATA
            LVDS-VME #3
                                                     4

            LVDS-VME #4                                                        PMTs
          4x8bit = 32 bit / board
                           Status
   Setup LinuxPC with Bit3 VME interface at
    Fermilab in Dec. 2005
       Start porting the Windows VB routines to C codes on
        Linux PC
   Successful readout pedestal data on a AFEII
    board in the end of May.
       Initialization AFE boards via 1553
       Control SASeq
       Readout buffer module, VLSB
       GUI is prepared for the test
   VME ADC and TDC modules have been tested in
    May KEK beam test to readout TOF, ACC and
    other beam counters in UniDAQ framework
             Tracker DAQ at Fermilab
       Backplane
                   AFEII
Cryocooler




                           Bit3
                                                SASeq
                                  1553

                                         VLSB
VME ADC/TDC at KEK beam test




                              multi-event ADC (CAEN V792)
        Multi-hit multi-event TDC (CAEN V1290)
                       UK colleagues provides VME ADC/TDC
         To do for KEK beam test
   Extend slow-controller to handle 4 boards of
    AFEII
   Implement AFEII readout routines on UniDAQ
    framework
       Merge AFEII data with the other ADC/TDC data
   Integration test at Fermilab
       Time stamp of integration window on Trip chip
       Tracker prototype will be shipped from UK in July
       LED and Cosmic-ray data taking
   Setup the system in KEK in Aug.
      KEK beam test triggering scheme
      Take coincidence of TOF counters
      Check the timing with charge integration window on Trip chip
               Trip has 48 x 132ns time slots
      Distribute trigger to CAMAC and VME modules
Trip integration window



VLPC analog



TOF coincidence



Event trigger




                                   Dt : relative timing of trigger to integration window
                     Trigger (cnt’d)        21 usec

         132 ns x 53 = 6996 ns




                                   396 ns x 12 = 4752 ns




                   Live time = 132 ns x 36 / 21 usec = 23%




                                                 SASeq
trig                                               trig
       DG2020
                                                             Trig + CLK
TOF Trig                                                                  AFE
                                                    sync
                To do for MICE

   Event-buffering in VLSB
       Need to trigger 1MHz (1k muons in 1msec
        from ISIS)
       Implement zero suppression scheme
       Bench mark test of readout speed
   Establish event numbering scheme /
    synchronization to the other counters
               Data Rate in KEK test
   Prototype has 2048ch readout channels
   Beam structure @ p2 beam line: continuous 1000 muons / 2 sec

   AFEII  VLSB (VME LVDS SERDES Buffer)
       2 Bytes / channel
       20 bits @ 53MHz (multiplex 2 channel = 10bits + 10bits)
       64 x 1/53MHz ~ 1 msec for a trigger
   VLSB  VME bus  PC
       Measured Transfer rate by KEK-online group with Bit3
            23MB/s in DMA mode
            2MB/s in PIO mode
       4MB / 4-sec-spill
       Need to transfer 4kBytes in 1 msec w/o buffering

   In case of zero suppression by DISCR, data rate could be reduced to
    ~1/10
                     Summary
   We are preparing stand-alone DAQ system
    for tracker
       VLPC readout
       VME ADC/TDC
       Not yet on slow-control
   To fit to MICE case
       Need event buffering (~1k muons / 1msec)
       Need trigger / event numbering scheme

								
To top