CalTrigger

Introduction:
 
 

At Level 1 the calorimeter triggers are divided into 2 types: object triggers ( electrons, photons and jets) and global triggers ( E_T total and missing energy).

The object triggers are formed by applying thresholds to individual towers while thresholds for the global triggers are applied after summing energies from all towers.

The digitized calorimeter data are summed into trigger tower energies and weighted by sin(theta) on the front-end cards to produce the transverse E_T. Data are processed in the 6 L1CAL crates with final global sums and trigger summaries made in the Global L1 crate. The data are then forwarded to the L2 cluster finder after a L1 accept.

At Level 2 the clustering algorithms loops over all of the trigger towers, first comparing the transverse energies with different thresholds for the seeds. If a seed is found the information on eta, phi, and energy is stored in an array of information for a cluster. The seed is masked off, and four surrounding towers are flagged for the shoulder threshold passes. This continues until there are no more towers to be included in a given cluster. For each tower a word is created, containing the result of the clustering algorithm. (each tower can be seed or cluster member for a given pass). A schematic representation of the data flow in the Calorimeter code is shown below: 

Code

The following classes have been designed and coded (package CalTrigger):

• CalTriggerDataMaker : public AppModule is the driver module ( the implementation code is in several files:
• CalTriggerDataMaker.cc contains beginRun/endRun/beginJob etc general methods;
• CalTriggerEvent.cc contains the event method
• CalTriggerDB.cc contains the dabase access routine
• CalTriggerPatterns.cc contains database access routine to TriggerDB for DIRAC memory patter generation
• CalTriggerPatterns2.cc contains the code for the memory patter generation ( same as online java code)
• CaltriggerMonitors.cc contains the TC1D monitor mehtod
• TC2D_monitor.cc
.
• CalTriggerData: public StorableObject is the main object of the calorimeter simulation.


 It contains the array of trigger towers, arrays of L1 boards ( 96 DIRAC and 12 CRATESUM cards) and quantities resulting from the L1 simulation and L2 clustering and isolation algorithms.
Through methods of CalTriggerData the emulation of the digitization stage is performed ( methods: FillTower, CorrectTransEnergy), L1 is emulated through the methods: dirac_sim, cratesum_sim, L2 clustering through the methods: DoClustering() and DoIsolationSums(). Modifiers and accessors are of course provided plus a number of utilities like the method TC1D_output_conversion to convert the output of dirac_sim and cratesum_sim into the right TC1D bank format.
 Some information on DIRAC
CalTriggerData does not implement streamers.

• Class TriggerTower contains generic trigger tower information: eta, phi of the tower, hadronic energy, EM energy and transverse quantities.
• Class TriggerTowerTypeX : public TriggerTower contains specific trigger tower information ( PMT sums).
• Class TriggerCellKey maps the transformation of the physical towers into trigger towers.

 

 
 
 
 
 
 
 

• void FillTower() where
• TriggerTower::fillTriggerTower() is called and the actual implementation of this method is found in the TriggerTowerType class.

fillTriggerTower takes the individual photomultipliers information for each detector and sums them based on the type of the tower.
• void CorrectTransEnergy() calls the correctTransEnergy method of class TriggerTower, which uses an array of sin$\theta$ values for each trigger tower index. The energy in each trigger tower is multiplied by this sin(theta), and then integerized. The resulting energies are stored in the data members of TriggerTower.
• TriggerTower* towerMake() creates the elements of the 24x24 trigger tower array, after selecting the correct TriggerTowerType.
• Several L1 simulation methods }.
• void dirac_sim (int card_id);
• void cratesum_sim( int card id);
• PfSumet_sim( ..): ( PreFred SumEt calculation);
• PfCaltrg_sim( ...): PreFred CalTrigger word calculation;
• bool DoClustering() loops over all of the trigger towers, first comparing the transverse energies with thresholds for seeds, based on the physics looked at. The thresholds will be read from a database. If a seed is found the information on eta, phi, and the energy is stored. The seed is masked off, and four surrounding towers are flagged for shoulder threshold passes. This continues until there are no more towers to be included in a given cluster. For each tower a result word is created, to be stored in the TC2D_StorableBank.
• bool DoIsolationSums() performs isolation sums on the clusters.

 When running CalTriggerDataMaker, always remember to include the talk-to to CalibrationManager and CalorimetryModule . An example of talk-to to CalTriggerDataMaker, CalibrationManager and CalorimetryModule is in run_TRGSim++.tcl.

CalTrigger input and output

CalTrigger reads in input the calorimeter D-banks: CEMD, CHAD, PEMD, PHAD, WHAD.
It can also get its input from CalData, but in this case we need to control that the verb makeDbanks  is set t true in the CalorimetryModule talk-to.  To get CalData, one might need to puff it, via the talk to to PuffModule:
module talk PuffModule
    puffonly add CalData
exit

CalTrigger  gets also the track input from XTRD_StorableBank.

Starting witn 4.5.0int3 the verbs standalone and simulation have disappeared !

It is now possible to use the following verbs to run on  real or simulated data:

    use_software_CAL_banks : this verb allows the user to select real or simulated CALORIMETR data ( CEMD, CHAD,PEMD,PHAD,WHAD or CalData)
    use_xtrp  : this verb allows to use the XTRP track input, if it is set to f then no XTRP input is present ( --> no ELCTRONS!)
    use_hardware_xtrd:  this verb allows the user to select real or simulated XTRP data ( namely real or simulated XTRD_StorableBank)
    use_hardware_L1: this verb allows the user to select real or simulated trigger tower energies in the clustering simulation ( namely real TC2D or simulated energies)
 
 

So, for example to run on simulated data  without XTRP ( old standalone option) the user will need to set the following talk-to:

use_software_CAL_banks set  t
use_xtrp  set f
use_hardware_xtrd set f
use_hardware_L1 set  f
 
 

To run on simulated data  with XTRP the user will need to set the following talk-to:

use_software_CAL_banks set  t
use_xtrp  set t
use_hardware_xtrd set f
use_hardware_L1 set  f
 
 

CalTrigger produces in output the following banks:

• TC2D with the list of trigger tower energies and clustering summary word for each tower ( description "Simulated Trigger Bank".
• TC1D with the results of the DIRAC triggers: description "Simulated Trigger Bank".
• TL1D with the PreFred bits for CalTrigger and SumEt: description "Simulated Trigger Bank".
• TL2D where the only block filled is the clustering block: description "Calorimeter TL2D Bank".

Last updated  Sept 04 2001; Simona Rolli