Project

General

Profile

Actions

FF130613

This will be in Fermilab all day on 13th June. If we run out of time we can overrun into the software workshop on 15/16 June.

Rooms

9 am - 1 pm: Hornet's Nest (WH 8th floor crossover)
1 pm - 5pm: Theory room (WH 3rd floor NW)

Attendees

  • Chris Rogers
  • Durga Rajaram
  • Ian Taylor

Agenda

  • Maybe some overview (Rogers)
  • Calibration code
    • Physics
    • Code structure
    • Code
  • Monte Carlo
    • Geometry and survey
    • MC Digitisation
  • Ckov
  • Reconstruction
    • Data structure
    • Digits, slab hits, space points
    • Integration tests, documentation
    • Online plots

Notes

Calibration - General

1. Three parameters trying to access
  • Trigger time delay which is pixel by pixel
    • delay between trigger station and the signal coming off a particular pixel
  • Time walk correction which is PMT by PMT
  • Speed of light in material
ACTION:
  • calibration should record version of the code that was used;
  • any inputs (e.g. datacards);
  • this should be stored somewhere permanent, best in the calibration itself
    ACTION:
  • TOF calibration - should be a documented procedure for what data to take
  • TOF calibration - should be runnable by non-expert
    • Add a reducer - MapTOFPlot.py convert into a reducer
    • Output root histograms get put into end_of_run directory
    • Secondary script which:
      • runs the TOF calibration on data in end_of_run directory, taking a selected numbre of files
      • copies root histograms to PPD server micewww
      • uploads calibrations to config db
      • SOC trained to appropriate level to check if there are problems
        ACTION:
  • Calibration should tie into geometry/survey; needs to be able to use
    ACTION:
  • Calibration data should be sorted by data that the first run in the series was
    done
    ACTION:
  • We will need a RAID array in PPD cluster or equivalent for calibrations
    • ~ 1 Gb required for TOF calibration histograms

Calibration - Yordan Code

calib.cpp
  1. time walk
  2. t0 calib
  3. trigger calibration
    • In order, one depends on previous
    • All hard wired
      • e.g. central pixel is hard coded
      • e.g. distance between stations is hard coded
    • Does fit to existing root file
    • No comments
Time walk calib
  • Uses minuit fit A + B/(x+C) + D/(x+C^2)
  • Score function calculates chi2
  • Fit has no limit
  • Fit uses MIGRAD
  • Fits on binned data
T0 calibration
  • Fits to 3 peaks and require a minimum gap
  • Full of lurking evil - needs lots of checks
Trigger calibration
  • Looks okay
Common comment
  • all calibrations should check chi2 and raise a warning if chi2
    is out of range
  • check number of entries is reasonable
  • Path length affect on time of flight? ~ 30 ps systematic
    • Cross check by looking at effect of quadrupole currents on calibration constants
  • Particle speed affect on time of flight? - 20 ps systematic smearing of the ADC pulse electron velocity vs muon velocity

Calibration - Rayner

ACTION: rajaram to push to launchpad
  • DONE

MCDigitizer

BUG: TOF hits outside of a slab
ACTION: header file needs comments
ACTION: Noise study (predicated on some trigger MC)
  • Look at high rate data, decide if there is an issue
  • Add trigger window
  • Add ADC model
  • Add noise

ACTION: Use MiceModules from GlobalReconstruction
ACTION: check_sanity should check for recon_events
ACTION: all_tof_digits.push_back(Json::Value()) is useless, line 81
ACTION: strig.str() needs to be moved down line 96
ACTION: line 82, line 110, line 111 are redundant (tof_evt)

make_digits()
ACTION: line 128 calling clear on a new std::vector is redundant
ACTION: make debug a datacard
ACTION: some of the type checking on hit members needs to be done
ACTION: ROGERS to add daq_event_type to MapPyReconSetup()
ACTION: encode station, plane, slab into a lookup table
ACTION: ask Maurizio what the time resolution of the PMT is
ACTION: break the function down

check_sanity_mc()
Nowt

get_npe()
ACTION: line 350 should be TOFpmtQuantumEfficiency
ACTION: random seed should be set off the primary

find_trigger_pixel()
ACTION: ROGERS - how do we set up trigger station
  • When running pure MC
  • When running MC of a run
    ACTION: should find earliest trigger pixel, rather than first trigger pixel

fill_tof_evt()
Nowt

Digits

ACTION: SetConfiguration does nothing
ACTION: YORDAN write up DAQ

SlabHits and SpacePoints

ACTION: TOF reconstruction should make slab hits for every combination of PMT hits
ACTION: Review Issue 568 and repeat that sort of analysis
ACTION: Write down which calibration and cabling was used in the data structure

Updated by Rajaram, Durga over 7 years ago ยท 10 revisions