020924.ppt - HEP Group Research Pages by nyut545e2


									           Trigger issues for the CALICE
           beam test
                    Paul Dauncey
                    Imperial College London, UK

24 September 2002                Paul Dauncey     1
     Trigger requirements
     •       Need clean data to compare carefully with simulation
         •     No double particle events, beam halo, cosmics, etc.
     •       Must veto any events with more than one particle
         •     Obvious: no trigger if previous
               particle within ~1ms
         •     Less obvious: no trigger if
               following particle with
               ~150ns, i.e. before VFE
               sample-and-hold time

             CR-RC shaping in units of
             shaping time tS ~ 180ns

24 September 2002                         Paul Dauncey               2
     Trigger issues
     •       Want to discuss the issues for triggering here
         •     No commitment for providing trigger yet (certainly not from the UK!)
         •     I hope this starts a discussion of what will be needed
     •       UK interest here as we are building readout electronics and
             DAQ which uses the output of the trigger logic
         •     Soon need to decide what logic goes where
     •       I will describe a possible system
         •     “Straw man” proposal to get people thinking
         •     How much we actually use is still to be decided

24 September 2002                        Paul Dauncey                                 3
     “Activity” and “Trigger”
    •        Assume for now a simple scintillator trigger
        •      Concept works fine for scintillating fibre hodoscope also
    •        Define two levels of scintillator logic condition
         •     “Activity”; OR of hits
               in the scintillators, i.e.
               any activity
         •     “Trigger”; AND of
               hits in the scintillators,
               i.e. good central beam

24 September 2002                           Paul Dauncey                   4
     “Activity” and “Trigger” (cont)
     •       Use Activity to veto Trigger if too close in time
         •     If Trigger within ~1ms of last Activity, then prevent Trigger from being
         •     If Activity within ~150ns of Trigger, then abort Trigger
     •       Cannot hold Trigger during the ~150ns as it must reach the
             VFE chip in time to produce sample-and-hold (~180ns)
         •     Electronics must allow Trigger without subsequent readout
         •     Only other possibility is to read out state of Activity and not use events
               with it set when reading offline
     •       Need to be aware of “race condition”!
         •     Trigger clearly also produces Activity so must be sure it does not kill
         •     Requires careful signal shaping/timing

24 September 2002                          Paul Dauncey                                     5
     Cosmic pollution?
     •       Do we need to include “cosmics” top-bottom scintillators in the
             Activity logic OR so as to veto events with a cosmic as well?

         •     Will lower scintillator fit under support table for calorimeters?
         •     Risk of bias by vetoing events with large showers leaking out of sides?
         •     What is the cosmics probability to be in time with our few ms sensitive
               time? If low, then can be neglected

24 September 2002                        Paul Dauncey                                    6
     BaBar ECAL beam test trigger logic
                         Example of trigger including some of
                             the elements suggested here

24 September 2002     Paul Dauncey                              7
     Different Triggers
     •       Do we want other trigger conditions besides beam?
         •     Cosmics? (Now selecting them!)
         •     Anti-beam (i.e. noise events)?
         •     Clock trigger?
         •     Any others?

     •       Should these be parasitic between beam bunch trains?
         •     Exactly equivalent conditions as data taking
         •     Makes control more complex
     •       Need VME control of trigger type
         •     From where? UK electronics???

24 September 2002                        Paul Dauncey               8
     Logic implementation
     •       Where will the vetos and aborts be done?
         •       Activity preventing Trigger needs
             •        Logic to generate veto
         •       Trigger aborted by Activity needs
             •        Logic to generate abort
             •        VME interface to read out Activity status
     •       What are the signal levels to be used?
         •       NIM? LVDS? TTL?
         •       We need to be able to plug the cables into the UK crate
     •       How much of this logic should be in the UK electronics?

24 September 2002                               Paul Dauncey               9
     UK trigger board
     •       Single board which handles trigger input to UK electronics
         •     Distribution to readout boards within crate (and elsewhere?)
         •     Holds off further triggers until readout complete
         •     Hold off is reset via VME at end of event read
     •       Minimum functionality needed
         •     Inputs: Trigger, Abort; from trigger logic. All other timing, veto, abort
               etc. done in NIM
         •     Output: Reset; to trigger NIM latch to allow further triggers
     •       Maximum extra functionality we might add
         •     Input: Trigger, Activity; all veto, abort, etc. done in logic on trigger board
     •       Also need to decide how to select different types of trigger
         •     More VME-controlled output levels? How many?
     •       Hard to build a general trigger board to cover both extremes…
         •     …and even harder to cover all possibilities between them

24 September 2002                          Paul Dauncey                                         10
     Main questions
     What are the requirements for the trigger?
     •   Different types?
     •   Different veto conditions?
     Who will supply the hardware?
     •   Scintillators, PMT’s, HV power supplies, NIM (or similar)
         trigger logic units, etc?
     •   How much will be common to the beam position system?
     Do we want post-trigger abort in hardware or software?

     Timescale to decide must be soon!
     •   UK should be building trigger board (one off) in ~7 months
     •   Needed to test prototype readout boards
     •   Must know functionality of trigger board with ~2 months
24 September 2002                Paul Dauncey                         11

To top