The LHC beam loss monitoring system's real-time data analysis card

The BLM (Beam Loss Monitoring) system has to prevent the superconducting magnets from being quenched and protect the machine components against damages making it one of the most critical elements for the protection of the LHC. The complete system consists of 3600 detectors, placed at various locations around the ring, tunnel electronics, which are responsible for acquiring, digitizing, and transmitting the data, and surface electronics, which receive the data via 2km optical data links, process, analyze, store, and issue warning and abort triggers. At those surface units, named BLMTCs, the backbone on each of them is an FPGA (field programmable gate array) which treats the loss signals collected from 16 detectors. It takes into account the beam energy and keeps 192 running sums giving loss durations of up to the last 84 seconds before it compares them with thresholds uniquely programmable for each detector. In this paper, the BLMTC's design is explored giving emphasis to the strategies followed in combining the data from the integrator and the ADC, and in keeping the running sums updated in a way that gives the best compromise between memory needs, computation, and approximation error.