Development of a Test System for the Quality Assurance of Silicon Microstrip Detectors for the Inner Tracking System of the CMS Experiment

The inner tracking system of the Compact Muon Solenoid (CMS) experiment at the Large Hadron Collider (LHC) which is being built at the European Laboratory for Particle Physics CERN (Geneva, Switzerland) will be equipped with two different technologies of silicon detectors. While the innermost tracker will be composed of silicon pixel detectors, silicon microstrip detectors are envisaged for the outer tracker architecture. The silicon microstrip tracker will house about 15,000 single detector modules each composed of a set of silicon sensors, the readout electronics (front end hybrid), and a support frame. It will provide a total active area of 198 m2 and ten million analogue channels read out at the collider frequency of 40 MHz. This large number of modules to be produced and integrated into the tracking system is an unprecedented challenge involving industrial companies and various research institutes from many different countries. This thesis deals with the physics of silicon sensors and the preparation of the large-scale production of front end hybrids and modules which will span over one year. During this period it is essential to assure efficient and reproducible manufacturing as well as diagnostic procedures among all centres. This demands a test environment capable of supervising the complete production phase while providing a reliable quality assurance of front end hybrids and modules. To meet these requirements, the test setup ARC (APV Readout Controller) was developed in Aachen and distributed among all collaborating institutes. In this thesis, the ARC system is introduced and, in particular, the development of test procedures implemented in the corresponding readout and analysis software (ARCS) are described. Based on the characterization of a pre-series of 21 silicon modules, called express-line, built in the year 2002 by the tracker end cap collaboration, the functionality and suitability of the ARC system could be demonstrated. The measurements of these modules, performed at ambient and at low temperatures, allowed the improvement of existing test procedures as well as the development of new tests incorporated in the ARCS environment to guarantee a comprehensive and redundant failure determination. Three types of faults showed up during the different test procedures: open bonds, shorts, and pinholes. In total only 29 channels (of 21 x 512 channels) were identified as faulty channels affected by one of the failures mentioned above. This leads to a failure rate of less than 0.3% which emphasizes the good quality of the components used and the accurate manufacturing methods. Especially the reproducibility and the comparability of the test results are encouraging with respect to the enormous challenge of the large-scale production the CMS experiment is facing.