CHARACTERIZATION OF THE GEM CHAMBERS FOR THE SBS/BB FRONT TRACKER AT JLAB HALL A

A new Large-Acceptance Forward Angle Spectrometer (Super Bigbite Spectrometer-SBS) is under development for the upcoming experiments in Hall A at the Thomas Jefferson National Accelerator Facility (Jlab in Virginia-USA), where a longitudinally polarized (up to 85%) electron beam up to 12 GeV energy is now available. The excellent beam intensity (up to 100 μA), combined with innovative polarized targets, will provide luminosity up to 10^39/(s·cm2), opening interesting opportunities to investigate unexplored aspects of the inner structure of the nucleons. In its full configuration, the new spectrometer will consist of a dipole magnet, three charged particle trackers, two identical proton polarimeters and a segmented hadron calorimeter. The main requirements for the SBS tracking system come from the upcoming experiments devoted to the measurement of the nucleon electromagnetic form factors at high momentum transfer and more generally from the experiments at high luminosity and with high energy beam. The front tracker, placed just after the dipole magnet, consists of six layers of large area GEM (Gas Electron Multiplier) chambers; each chamber is made by three adjacent GEM modules of 40x50 cm^2 active rectangular area. The GEM technology may sustain the expected high hit rate (~100 MHz/cm^2), providing adeguate spatial resolution (~80 μm), and excellent radiation hardness, at relatively small cost. We present the main features of the SBS front tracker and its GEM detectors and we will discuss the first results of the tracker commissioning at JLab. Finally, we present an method to remove noise/background data from the real-time data stream, to get a sustainable data rate. In fact, the quantity of collected data may become problematic and a data reduction that preserve useful physics information is desirable. Therefore, we introduce the possible exploitation of the Brain Project tool, an AI-based technique that can produce robust discriminating functions, which can be implemented in firmware to efficiently discriminate noise and background from the signals of physical interest. In order to fulfill this goal, we describe a detailed study of the APV signals coming out from the GEM modules strips.