Sep 19, 2019 Calorimeter

From GlueXWiki
Jump to: navigation, search

Meeting Time: 11:00 a.m.

  1. To join via a Web Browser, go to the page [1]
  2. To join via Polycom room system go to the IP Address: ( and enter the meeting ID: 907185247.
  3. To join via phone, use one of the following numbers and the Conference ID: 907185247.
    • US or Canada: +1 408 740 7256 or
    • US or Canada: +1 888 240 2560
  4. Upon connection all microphones are automatically muted. To unmute your mike on a Polycom or equivalent unit, enter *4. Unmuting on a computer is trivial as there is a microphone button than can be clicked.
  5. More information on connecting to bluejeans is available.

Participant Direct Lines

  • JLab Phone in CC L207 is 757-269-7084 (usual room)
  • JLab Phone: in CC F326 is 757-269-6460
  • Phone in the Regina Video-conference Suite is 306-585-4204
  • Athens Phone: in Christina's office is 011-30-210-727-6947


  1. FCAL HDFCAL log book
  2. BCAL HDBCAL log book

Goals for Calorimetry Group

  1. Determine preliminary photon reconstruction efficiencies as a function of E, phi and theta in data and simulation with a point-to-point precision of at least 5%.
  2. Measure systematics of pi0/eta mass calibration as a function of detector position to a precision of at east 5 MeV.
  3. Demonstrate agreement of photon reconstruction efficiency and resolution between data and simulation as a function of E, phi and theta to within 5%.

Action Items

Short term items:

  1. Investigate uniqueness tracking for efficiency calculations in calorimeters (Jon)
  2. Complete Calorimeter work packages

Long term items:

  1. From Mike's work: -cos(phi) type dependence in the data and a phi dependence at the downstream end of the bcal. See Sep 21, 2017 Calorimeter Meeting.
  2. FCAL: It has been noticed that the pi0 mass shifts when the LED pulser runs at high rate. This is still not understood. See Log Entry 3595928 and references. Several runs have been taken in raw mode. These runs exhibit the shift in pi0 mass with LED running at 1 kHz. There is a need to analyze these to try to determine the cause:
    1. 50634 - FCAL Green LED: 1kHz, 29 V, Livetime is 9%, 1.7 M events
    2. 50635 - FCAL Violet LED: 1kHz, 22 V, Livetime is 9%, 1.4 M events
    3. 50636 - FCAL Blue LED: 1kHz, 15 V, Livetime is 9%, 1.5 M events

Tentative Agenda

  1. Announcements
  2. Action Items
  3. Collaboration meeting
  4. Run Updates
    1. FCAL
    2. BCAL
  5. Calibrations
  6. Monitoring
    • Varun : BCAL LED monitoring update[2]
  7. Efficiencies
  8. Simulations
    1. Tolga: I will present some updates and by default, questions regarding some findings for my Neutron ID stuff
  9. BCAL Deep Learning
  10. NIM articles
  11. Any other business


Attending: Elton, Mark, Jon, Colin, Matt (JLab); Curtis, Will (CMU); Christina, George (Athens); Sean (FSU): Tolta, Churamani (FIU); Foda, Varun, Zisis, Karthik (UofR).

  1. Announcements
  2. Action Items
    • Karthik - no update on pi0 mass z dependence
    • Mark - Fiducial cuts have been implemented into the analysis library by Sean. They should be included into future launches.
  3. Collaboration meeting
  4. Run Updates
    1. FCAL
      • Sasha has implemented new firmware into one FCAL crate and some checks have been completed by Sean. Still need to take data for the entire detector to check whether new firmware would be implemented before the next run. The firmware should take care of 4 ns offsets, but requires taking data and calibrating the firmware. This feature has been implemented and used in Hall B for some time.
    2. BCAL
      • Elton will go down to the hall with Jon and go over the BCAL electronics this afternoon
      • Mark: Will repeat the baseline measurement to compare pedestal widths to data taken a couple of years ago. Degradation of the pedestal widths might indicate degradation due to radiation. This will be done when he returns from his conference.
  5. Calibrations
    • Karthik: Checked gain calibrations fro 2019. Finds no change aside from 3 points in layer that are likely due to fitting problems. He will put the information onto DocdB for reference.
  6. Monitoring
    • Varun : BCAL LED monitoring update[3]
    • Most channels look very stable. Five different modules (upstream and downstream) show huge variations.
    • Mark: suggested to printout data run by run to investigate and possibly look at runs that have full waveforms.
    • Elton: appears to be a software problem.
  7. Efficiencies
    • Matt: Graduate student from Primex has determined run-by-run efficiencies for Fcal channels from LED runs in 2019. Should use the same program to determined efficiencies for 2018.
    • Efficiencies currently in database come from a limited study for 2017 data by IU student that has now left.
  8. Simulations
    1. Tolga: I will present some updates and by default, questions regarding some findings for my Neutron ID stuff
      • Update on timing for photons and neutrons.
      • Elton: suggests understanding data at a fixed photon energy and angle before moving to data with a range of energies and angles. Data at a fixed energy/angle should be able to be corrected for geometry (i.e. path length) and verify that the timing difference between reconstruction and generated times has a small width and centered on zero. Moving to another kinematics with also fixed values, repeat and check that corrections also place the time difference at zero with a narrow width. Then generating a range of energies/angles at the same time should give a distribution centered at zero with a width comparable to the time resolution of the system (~ 0.15 ns).
  9. BCAL Machine Learning
    • Building on work done by Tegan in 2014 with code re-checked earlier this summer to verify it still runs correctly.
    • Use a sample of Bggen events and take "good showers" that give good agreement between reconstructed and generated energy.
    • Use ML to distinguish (single) good showers from bad (i.e. attempt to differentiate photon showers from showers originating from anything else).
    • Mark: During studies of pi0 distribution, note that the background includes not only "bad showers" but also combinatorial background that cannot be eliminated with any analysis.
    • What next? Work on ML algorithms that use information from two showers that reconstruct to a pi0 in order to identify good pi0's
    • Mark: One needs to be careful that one does not let the ML algorithm pick photons that yield the pi0 mass because that would bias the distribution (as it does in kinematic fitting and one looses the ability to distinguish signal from background). One needs to consider how the algorithm will be used in a practical analysis.
  10. NIM articles
    • GlueX NIM paper documentation
    • Mark: Working on analyzing eta mass and width vs momentum to determine resolution. This will increase the statistics to analysis in BCAL paper and extend the analysis to FCAL. He will submit MC so it can be ready when he returns from his conference.
    • Matt: Has some text/paragraphs originally intended for an FCAL NIM paper which could be useful. The beam test in Hall B might have some useful material. Colin has started drafting some paragraphs also.
  11. Any other business