Fast Online Trigger using FPGA-based Event Classification for the COMET Phase-I

Abstract

The COMET Phase-I experiment searches for a neutrinoless muon-to-electron conversion which has never been observed yet. The world's highest intensity muon beam is applied, and it leads to an unacceptable trigger rate of O(10 6) Hz. For stable data collection, the trigger rate must be reduced to O(10 3) Hz. This requirement is met using online event classification in the detector system which holds 99% of signal events. This classification is performed by an FPGA-based trigger system, and its processing time is set to less than 5 µs by a buffer size of the detector readout electronics. A prototype board for the trigger system was developed, and communication systems for related electronics devices were also constructed. From test results, the total processing time is estimated to be 2.8 µs , which meets the requirement. We have also developed an online self-trigger system for cosmic-rays and confirmed the feasibility of this hardware logic. The trigger electronics were installed in a setup for cosmic-ray measurement, and the data acquisition was successfully done using the self-trigger system.

An FPGA-based Trigger System with Online Track Recognition in COMET Phase-I

“…The COTTRI MB has been reported in Ref. [7]. In this study, the COTTRI FE has been designed and produced, and the functions related to the online trigger algorithm have been implemented on both the COTTRI FE and MB.…”

“…The data acquisition (DAQ) system, however, requires a trigger rate of less than 26 kHz; the conversion-electron event can be missed due to the DAQ dead time. To solve this issue, we have developed an online trigger system called the COMET trigger (COTTRI) system, where the CDC-hit information is utilized with machine learning (ML) technique [5]- [7]. The low-energy electrons making CTH hits hardly reach the CDC due to the magnetic field.…”