Pion Meeting January 20 2023 3:30PM ET

Back to Main Page >> Pion Meetings

Zoom link: https://jlab-org.zoomgov.com/j/1608212513?pwd=M25LR1hzN3hLWm0xaFFETjZ0U1lQZz09

Agenda

1. (David) General updates
2. (Elham & Sudip): QSim vs remoll for pion detector
3. (Elham) More analysis of beam tests:
4. (Simona) More on PMT gain studies

Attendance

Neven, David, Elham, Simona, Kate, Sudip, Carl

Minutes

1. (David): We will stay at the same day & time for meetings this semester.
2. (David): Ryan Biraben is adding some additional material to "fins" that support beam pipe and Pb donut, to allow adjustability in case of imperfect cicularity. We should approximate the additional material in G4, and see its effect on the pi/e signal ratio.
3. (David): asks Elham to setup meeting with Jie Pan to discuss switchable PMT base for pion detectors: how to connect to 1" diameter PMT; perhaps we wish to have the integrating mode preamp separated from the base itself, similar to what will be done for ShowerMax.
4. (Sudip and Elham): Have tried pion detector inside QSim, and compared to ShowerMax Qsim results. Find that the reflectivity of the photocathode has almost effect on the number of PEs; moving from 12.5% reflectivity to 100% reflectivity has no effect. Possibly explained by the fact that the reflectivity in G4 is not directly related to the quantum efficiency, and the chance of a reflected optical photon bouncing around in the lucite and returning to the photocathode is probably small. Wouter explains that in G4 you can have your properties either determined by an efficiency (treat the surface as sensitive), or you can have it determined by the reflectivity (treat the bulk as sensitive). Fi\or us "G4OpBoundaryProcess.cc#L714 -> DielectricMetal()" is what geant4 does at the interface of the photocathode for incident optical photons. That function calls DoAbsorption in "G4OpBoundaryProcess.hh#L302", which turns the absorption with the specified efficiency into a Detection. The bigger puzzle is that changing the quantum efficiency (in QSim) has little effect on the number of PEs - moving from the initial QE(energy) function, which has its maximum QE at 0.3 to one where the QE is 1 for all energies (wavelengths) only increases the number of PEs in ShowerMax from 113 to 155 by 37% (113 to 155), which cannot be correct, it seems. We need to check that we understand the units of the coded QE, and how that we understand how G4 implements the QE.
5. (Elham) Changing the G4 macro from using /remoll/tracking/set 1 to /remoll/tracking/set 3, which causes secondaries to be tracked creates a high-energy tail to the simulated PE distribution, qualitatively similar to the Mainz beam test data - good! Note that previous remoll G4 simulations where we were studying pi/e ratios did track secondaries (of course). Detailed comparison on the size of the tail to the sigma of the Gaussian part will require us to have the simulation give reasonable PE estimates (as, at the moment, the simulation gives more than an order of magnitude more PEs than the data suggest). Can also generated a low-energy "bump" or tail in the simulation if material is added far upstream of the lucite, due to multiple scattering causing some tracks to not pass through all the lucite. However, the thin quartz in some of the beam test data was actually located rather close to the lucite, so
6. (David) Suggest doing a simulation of the beam-test data conditions (i.e. incident pencil beam) where we replace 855 MeV electrons with 4 GeV pions, and/or 4 GeV muons) just so we assure ourselves that the high-energy tail is a showering effect for electrons, and that we don't expect it with cosmics or beam-generated pions.
7. (Elham) Some discussion of cosmic data comparison to simulation. Suggest that the trigger conditions may not have been ideal for the cosmic data in "setup number 2" (the one mocking up the Moller experiment configuration) based on looking at the trigger scintillator ADC spectra.
8. (Simona) Continued bench test on the PMT from Manitoba. At 1540 V, gain is 3x10^6; see figure below. To go to lower voltages, would need to amplify more (another stage of amplification). This is exactly the value for the Hammamatsu R580 PMT as given on the gain vs HV curves on page 69 of the Hammamatsu catalog (see figure below), but it disagrees with the typical gain of 0.79x10^6 for the H3178-S1 PMT/base combination (which uses the R580 PMT) as listed on page 77 of the same catalog. Note that the gain inferred from a different tube (same model) used in the Mainz beam tests was about 0.38x10^6. Is this possibly just the spread of gains seen for different instances of this model of PMT?

1. (Wouter and Elham) Have just received back the detectors used in the Mainz beam tests. Plan is to removed the two PMTs and send them to Simona for bench testing their gains at JLab, and to re-instrument the lucite detectors with other (same model) PMTs for further cosmic studies at Manitoba.