Tuesday, October 1, 2024

Agenda:

• Dedicated discussion of GEP trigger logic, efforts in hardware/firmware/software/simulation. Andrew's slides

Key discussion topics/action items:

• Don and/or Jimmy will provide the software group with information about their convention for row/column/cell mapping and realistic geometry info, which may differ somewhat from idealized simulation geometry, which comes from ECAL design (somewhat older J/T file which may be missing some important components/materials relevant for simulation)
• Ben clarified some of the questions regarding firmware implementation:
  • Each cluster can have a uniquely defined “seed” threshold in addition to a ”cluster sum” threshold. In evaluating the trigger with simulation, we have only considered a cluster sum threshold thus far.
  • Each FADC channel has a programmable digital delay to align signals in time—important to account for different cable lengths used in ECAL.
    • It could be useful for easier bookkeeping, but not essential, to group similar cable lengths together in the same FADC boards and/or crates
  • Each FADC channel has a uniquely programmable gain to convert FADC amplitude into energy units (MeV is standard)—this can help account for e.g. underperforming PMTs for which we can’t safely increase HV to match gain with other channels.
  • Each cluster center has a uniquely programmable trigger threshold (cluster sum and (optional) cluster seed).
  • The cluster groupings, cluster bins, and coincidence lookup tables will be compiled into the firmware for efficiency/performance reasons.
  • We can define several coincidence lookup tables for the respective kinematic settings and also to define tighter/looser correlations at the trigger—then there will be one VTP configuration parameter to select which lookup table to use for trigger/DAQ:
  • We envision perhaps 3 lookup tables per kinematic:
    • Very loose “commissioning” trigger with no kinematic correlation: coincidence between global OR of all ECAL clusters and global OR of all HCAL clusters.
    • ”Loose” correlated trigger: conservative lookup table to make sure we efficiently capture the entire useful range of proton scattering angles in the CH2.
    • “Tight” correlated trigger: more aggressive enforcement of kinematic correlation, to be used if and only if we have verified that the tighter coincidence logic does not reject useful events.
• Software group (when we are done testing/debugging and satisfied that the simulated logic behaves as we intend) will provide example tables with clearly documented format to Hanjie/Ben/Simona et al. to help with firmware implementation (and organizing the additional crate/slot/channel mapping when the DAQ is fully cabled/finalized).

Back to GEP software meeting main page