Online trigger processing for the NA62 rare kaon decay experiment

The work presented in this thesis covers almost all the aspects of the common Trigger and Data Acquisition of the NA62 experiment that has as main goal the measurement the Branching Ratio of the ultra-rare K+ -> pi+ nu nubar decay, very useful to obtain a stringent test of the Standard Model. This PhD work began with the development and the testing of the firmware of common boards of the NA62 TDAQ system: TDCB and TEL62. The TDCB is a daughter-board of the TEL62 and measures the detector hit times. The TEL62 processes and stores these detector data in a buffer memory; at the arrival of a L0 trigger request, it extracts the data within a programmable time window around the trigger time to send them to the PC farm. The TEL62s of some detectors also take care of producing the L0 trigger primitives that are merged to generate L0 trigger requests. In this thesis is described the significant contribution given to the developing, the testing and the commissioning of the TDCB and TEL62 firmware. Since the 2012 Technical Run to the 2015 Run the system was tested, and evolved to be compatible with the detector input rate and the beam at growing intensity up to nominal. After three main versions the system composed by TDCB and TEL62 manages to cope with the design rate. Once the work on the Data Acquisition system was concluded, I focuses on the Trigger system and the analysis of the L0 and L1 triggers. The goal of this work was to study the detector response and the trigger conditions required to obtain the needed rejection factor with the minimum amount of signal loss. Starting from 13 MHz of event rate, the L0 trigger must provide a factor 13 of rejection to reach the design L0 output rate of 1 MHz; the L1 trigger should provide a rejection of a factor 10 to achieve the goal of 100 KHz of L1 event output rate. The starting L0 and L1 scheme analysed failed to reach the output rate request by a factor 5. This gap could be filled only by using the STRAW spectrometer at the L1. For this reason in the last part of this work the development of a L1 STRAW algorithm is described, starting from a Monte Carlo simulation and validating with the use of real data samples used for the L0 and L1 analysis. The results reached by the STRAW algorithm plus other improvements allows to achieve the required L1 rejection factor.