Study and improvement of firmware for advanced trigger algorithms, real-time data compression, and pile-up in neda (neutron detector array)

• Autores: Jose Manuel Deltoro Berrio
• Directores de la Tesis: Vicente González Millán (dir. tes.)
• Lectura: En la Universitat de València ( España ) en 2024
• Idioma: español
• Tribunal Calificador de la Tesis: Enrique Sanchis Peris (presid.), A. Pérez Pascual (secret.), Diego Barrientos Turrión (voc.)
• Programa de doctorado: Programa de Doctorado en Ingeniería Electrónica por la Universitat de València (Estudi General)
• Menciones:
  • Internacional
• Calificación: Sobresaliente Cum Laude
• Materias:
  • Ciencias tecnológicas
    • Ingeniería y tecnología eléctricas
• Enlaces
  • Tesis en acceso abierto en: webges.uv.es (pdf)
• Resumen

  • Nuclear physics is a branch of physics that studies atomic nuclei, exploring their structure, properties, interactions, and the forces that bind them together. This discipline is fundamental to understanding the universe at the microscopic and macroscopic levels.

    The importance of nuclear physics is enormous and extends to various areas of science and technology. For instance, in medicine, imaging techniques such as Positron Emission Tomography (PET) and radiation therapy for cancer treatment are based on nuclear physics principles. In astrophysics, nuclear physics helps explain processes occurring in stars, including how elements are formed in nuclear reactions. Nuclear energy is also essential for energy production. Nuclear power plants, which provide a significant energy source worldwide, operate using controlled nuclear reactions.

    In nuclear physics research, the nuclear structure is studied using high-resolution gamma-ray spectroscopy. This method accurately recovers the energy and angular distributions of gamma photons emitted in the transition between nuclear states.

    The electronics associated with the radiation detectors used in experiments are crucial for high-resolution gamma-ray spectroscopy. These instruments must be capable of detecting and measuring subatomic particles and radiation with high precision, which requires advanced electronics.

    The current trend is towards the use of digital electronics, which offer greater flexibility and the ability to implement more complex data analysis algorithms. However, for certain high-resolution measurements, analogue electronics outperform digital electronics. Therefore, one of the challenges in this field is achieving the same performance level with digital systems.

    One of the latest-generation instruments that uses the gamma-ray spectroscopy technique is the AGATA (Advanced GAmma Tracking Array) detector array. This instrument can have ancillary detectors to assist in gamma-ray spectroscopy. One of the ancillary detectors developed is the Neutron Detector Array (NEDA), a neutron detection system designed to be versatile, with high detection efficiency, excellent neutron discrimination, and high-rate capabilities.

    \newpage This thesis focuses on improving the data acquisition and analysis capabilities of NEDA through the following enhancements.

    First, developing a lossless data compression method that allows the signals acquired by NEDA to be compressed individually. Applying this technique allows for increasing the counting rate in the experiments and reducing the storage size required for each event and, therefore, the total data acquired.

    Secondly, a double trigger condition system is developed to improve neutron-gamma discrimination. This involves an in-depth study of the electronics and firmware already implemented in the digitiser of NEDA called NUMEXO2. The system enhances the acquisition capabilities by combining two independent trigger signals based on Charge Comparison (CC) and Time-Of-Flight (TOF) measurements. Implementing this system in the Virtex-6 FPGA includes signal processing, baseline correction, and various trigger logic blocks. Evaluations using data from the E703 experiment at GANIL show that the AND trigger mode is effective for experiments requiring a low level of trigger requests. In contrast, the OR mode is beneficial for minimising neutron loss when the system can handle higher trigger requests.

    Thirdly, the thesis addresses the issue of pulse pile-up in nuclear spectroscopy and reaction studies, where overlapping pulses degrade energy and timing information. The proposed solution uses a one-dimensional convolutional autoencoder (1D-CAE) to reconstruct pile-up events. The method reconstructs pile-up signals and evaluates them by comparing the reconstructed signals to the original ones, achieving very high levels of similarity. This solution allows for the correct analysis of reconstructed signals using Pulse Shape Analysis techniques and provides correct neutron/gamma discrimination.