ONLINE DIAGNOSTICS AND MONITORING FOR THE TRIGGER SELECTION AT THE CDF EXPERIMENT

You are viewing the theme
[Voti: 0    Media Voto: 0/5]

The Collider Detector at Fermilab (CDF) is a general purpose experiment for the study of collisions at the Tevatron Collider, located in Batavia (Illinois, USA). CDF consists of many different detector subsystems placed around the collision point, on the Tevatron beam-line. The detectors observe the collisions taking place at their centers and record all the available information for later analysis. Since the collision rate is much higher (7.6 MHz) than the rate at which data can be stored on tape (~100Hz), a particularly difficult task at CDF is the selection of interesting events that are to be recorded permanently. This daunting task is performed by the "Trigger". This is a very complex system of dedicated powerful processors which work together to analyze events and make difficult decisions on every collision in a very short time. Even though the CDF trigger is the most powerful existing today, it needs continuous upgrades since the Tevatron working conditions become each day more demanding, requiring more powerful processing units. During my thesis work I have contributed to a particularly important trigger upgrade that will allow the CDF experiment to be much more competitive in the search for the Higgs particle. This upgrade achieves supercomputer-class performance using a mixture of technologies (FPGA-based boards and modern CPUs), applied to different parts of the CDF detector. The system has been designed, approved, produced, tested and will be installed in just six months from the approval. Though I contributed to all phases of the project, my main activity was the development, deployment and operation of a monitoring and diagnostic system for this new component of the CDF Trigger. Each phase of the project (prototyping, production and commissioning) needed specific diagnostic/monitoring functions and I developed and operated all these tools, to fix problems that appeared during tests and to provide a sound on-line monitoring of the system during data acquisition.