scintiPulses

Realistic Simulation of Scintillation Detector Signals

📖 Overview

scintiPulses is a Python package designed to generate high-fidelity photodetector signals from scintillation detectors. It models the complete signal chain—from the initial energy deposition to the final digitized output—incorporating key photochemical processes, electronic effects, and digitization artifacts.

This tool is ideal for post-processing data from Monte Carlo simulation frameworks (e.g., Geant4, MCNP, FLUKA), transforming raw energy deposition timestamps into realistic pulse waveforms for algorithm testing and detector design.

✨ Key Features

• Physics-Based Modeling: Simulates time-dependent fluorescence including prompt and delayed components, with Fano factor support.
• Ionisation Quenching: Optional Birks-law quenching of the prompt yield, driven by an electron stopping-power model (Bethe + Joy-Luo).
• Delayed (TTA) Fluorescence: independently selectable yield model (dE/dx-dependent triplet-triplet annihilation, or a legacy fixed-fraction model) and kinetics model (non-exponential Voltz bimolecular, or plain exponential), so any combination can be paired. See docs/photophysics_model.md for the full derivation.
• Multi-Channel Support: Simulate signals across multiple independent detector channels.
• Photodetector Physics: Incorporates quantum shot noise, after-pulses, and thermionic (dark) noise.
• PMT Modeling: Configurable gain, gain fluctuation, signal inversion, and charge spreading.
• Electronic Simulation: Models Johnson-Nyquist thermal noise and signal shaping.
• Analog Filtering: Simulates RC preamplifiers and CR$^{n}$ fast shapers.
• Digitization: Includes anti-aliasing low-pass filtering, ADC quantization, and saturation effects.

Technical details and validation can be found in:

DOI: https://doi.org/10.1051/epjconf/202533810001

📦 Installation

You can install scintiPulses directly via pip:

pip install scintiPulses

🚀 Quick Start

The following example simulates a single 10 keV energy deposition with a fixed arrival time, using a fast scintillator (e.g. LSO/LYSO-like), and plots the transimpedance amplifier output of the PMT.

import scintiPulses as sp
import matplotlib.pyplot as plt

t, v0, v1, v2, v3, v4, v5, v6, v7, v8, y0, y1 = sp.scintiPulses(
    # Source parameters
    [10],               # Energy deposition(s) in keV
    tN=1e-6,            # Total simulation duration (s)
    arrival_times=[1e-9],  # Fixed arrival time; if False, times are drawn from Poisson process
    lambda_=1e4,        # Input count rate (s⁻¹), used when arrival_times=False
    fS=1e9,             # Sampling frequency (Hz)
    # Scintillation parameters
    nChannel=1,         # Number of detector channels
    tau1=4.6e-9,        # Prompt fluorescence decay time (s)
    tau2=120e-9,        # Delayed-component characteristic time (s)
    quenching=False,    # Enable Birks ionisation quenching of the prompt yield
    kB=0.01,            # Birks constant (cm/MeV), used if quenching=True
    TTA_yield=True,     # Use the dE/dx-dependent TTA delayed-yield model
    TTA_kinetics=True,  # Use Voltz bimolecular kinetics for the delayed-component time shape
    Sd=0.005,           # TTA efficiency factor, used if TTA_yield=True
    kd=0.01,            # TTA saturation constant (cm/MeV), used if TTA_yield=True
    p2=0.1,             # Fraction of delayed component, used only if TTA_yield=False
    nE=100,             # Discretization points for the quenching/TTA integrals
    F=1,                # Fano factor of the scintillator
    L=5,                # Light yield (photons/keV)
    # PMT parameters
    C=5e-12,            # PMT capacitance (F)
    G0=20e6,            # PMT gain
    sigma_G=0,          # PMT gain fluctuation (std dev)
    I=-1,               # Voltage inverter (+1 or -1)
    tauS=2.23e-9,       # Charge spreading time (s)
    afterPulses=False,  # Enable after-pulses
    pA=1e-3,            # After-pulse probability
    tauA=5e-6,          # Mean after-pulse delay (s)
    sigmaA=1e-6,        # Std dev of after-pulse delay (s)
    darkNoise=False,    # Enable thermionic dark noise
    fD=1e-4,            # Dark count rate (s⁻¹)
    # Analog electronics
    electronicNoise=False,  # Enable Johnson-Nyquist noise
    sigmaRMS=0.01,      # Electronic noise RMS (V)
    pream=False,        # Enable RC preamplifier
    G1=1,               # Preamplifier gain
    tauRC=1e-3,         # Preamplifier RC time constant (s)
    ampli=False,        # Enable fast amplifier (shaper)
    G2=1,               # Amplifier gain
    tauCR=2e-6,         # Amplifier CR time constant (s)
    nCR=1,              # Order of the CR filter
    # Digitization
    digitization=False, # Enable ADC simulation
    fc=0.4e9,           # Anti-aliasing filter cut-off frequency (Hz)
    R=10,               # ADC resolution (bits)
    Vs=0.5,             # ADC voltage range (V)
)

# Plot the PMT transimpedance output (v4)
plt.figure(figsize=(8, 3))
plt.plot(t, v4, "-", alpha=0.7, label="Transimpedance output")
plt.xlabel(r"$t$ /s")
plt.ylabel(r"$v$ /V")
plt.title("Simulated Scintillation Pulse")
plt.legend(loc="upper right")
plt.grid(True)
plt.tight_layout()
plt.show()

📡 The Signal Chain (Outputs)

The function returns a 12-element tuple. This allows inspection of the signal at every stage of the processing chain.

⚙️ Configuration Parameters

1. Source & Timing

2. Physics & Scintillation

See docs/photophysics_model.md for the full mathematical model, including the electron stopping-power model and the discrete-time hazard functions used for the stochastic photon simulation.

3. Photodetector (PMT)

4. Analog Electronics

5. Digitization (ADC)

📚 Citations

If you use scintiPulses in your research, please cite:

Simulation of scintillation detector signals EPJ Web of Conferences (2025) DOI: 10.1051/epjconf/202533810001

You can also look at:

Simulation of stochastic jittering in liquid scintillation counters RAD Conference - abstract (2026) DOI: 10.21175/rad.abstr.book.2026.28.5

⚖ License

This project is licensed under the MIT License.