Time: 10:35 - 11:50, Dec. 6, 2023
Participants: Jiyuan Chen, Hongbin Diao, Dejing Du, Fangyi Guo, Yong Liu, Baohua Qi, Siyuan Song, Jiaxuan Wang, Xin Xia, Haijun Yang, Yunlong Zhang
Minutes by Yong Liu
- AHCAL simulation from SJTU and IHEP: summary by Jiyuan
- Geometry: minor differences in
- The first AHCAL layer
- PCB and ESR material definition
- Air gap
- Digitiser: two versions
- SJTU: an older version from Baohua in late 2022
- All SiPMs assumed to be HPK
- Simple model for SiPM non-linearity
- No saturation effects in ASIC (SPIROC2E)
- IHEP: up-to-date modelling for SiPM+SP2E
- Layer 1-38: HPK-SiPM, 39-40: NDL-SiPM
- SiPM non-linearity modelling with including pixel recovery effects (still with some artificial factor for the "effective pixel number")
- High Gain saturation from data: Zhen Wang's HG-LG inter-calibration from data; Low Gain saturation with assuming ideal 12-bit ADC (max. 4095 ADC tics) -> need LG saturation data (ongoing studies by Hongbin)
- Conclusions
- AHCAL energy observable (E_sum for hits over threshold) between MC and data
- Prominent discrepancy in electron data: results from Jiyuan, Dejing and Siyuan -> need further crosscheck
- Mostly consistent in muon data: minor difference could be due to the noise hits -> to be checked with more stringent muon track selection criteria
- Some-extent discrepancy in pion data: results from Dejing and Siyuan, not yet checked by Jiyuan
- Discrepancy in electron and pion data: beam-energy dependent
- Could not be explained by the uncertainty from material/geometry description
- Extra: Jiaxuan's previous investigation also suggested marginal effects in EM shower profiles/performance when changing CuW absorber density by ~12%
- It appears that non-linearity in SiPM readout may be a major dominating factor
- SiPM non-linearity -> correction needs the calibration of single photon spectrum per channel
- ASIC non-linearity in Low-Gain mode -> needs LED data with high light intensity
- Other discussions
- Material/geometry discrepancy: Jiaxuan/Yunlong plan to investigate further (ScECAL)
- Yunlong: secondary particles in showers hit SiPM directly?
- Yong: Acceptance ratio=(1.3x1.3mm)/(40x40mm)=0.1% seems a very small effect
- Yong: SiPM noise model -> need some time and person power to turn ideas into coding (with validation)
- ~10% noise-only events in PID studies; some-level noises could also be mixed in showers
- Later implement it into ScECAL performance studies: SiPM noise, poor S/N, relatively higher trigger threshold (trigger cutting into MIP spectra) -> how much performance degrading
- Zhiyu: LED calibration
- SPS in HG can be done first, then SiPM non-linearity correction
- Action items
- Jiyuan and Dejing: make sure use the same data sets with PID information (either FD or ANN) to compare with MC (with and without digitisation) -> data from Xin (FD) and Siyuan (ANN)
- Dejing and Siyuan: to plot scaling factor vs beam energy (like Jiyuan's plot)
- Scaling factor = ratio of Esum mean values of data and MC
- Quantify the discrepancy
- Zhiyu: LED calibration for SPS, SiPM non-linearity correction
- Dejing:
- (simple) modelling of LG saturation in simulation (~3000 ADC tics)
- Muon tracks selection with more stringent criteria (to avoid noises, multiple particles)
- Strategy/roadmap
- All: to define event selection criteria to select well-controlled data samples
- Also apply to MC samples to check selection efficiency
- Focus on the simulation+digitisation validation
- Energy observables: aim to be generally consistent with data