Development and Evaluation of a Test System for the Quality Assurance during the Mass Production of Silicon Microstrip Detector Modules for the CMS Experiment

The Compact Muon Solenoid (CMS) is one of four large-scale experiments that is going to be installed at the Large Hadron Collider (LHC) at the European Laboratory for Particle Physics (CERN). For CMS an inner tracking system entirely equipped with silicon microstrip detectors was chosen. With an active area of about 198 m2 it will be the largest tracking device of the world that was ever constructed using silicon sensors. The basic components in the construction of the tracking system are approximately 16,000 so-called modules, which are pre-assembled units consisting of the sensors, the readout electronics and a support structure. The module production is carried out by a cooperation of number of institutes and industrial companies. To ensure the operation of the modules within the harsh radiation environment extensive tests have to be performed on all components. An important contribution to the quality assurance of the modules is made by a test system of which all components were developed in Aachen. In addition to thorough tests of the readout electronics and it enables the detection of many faults of the silicon sensor it is connected to. It is used in more than 20 different institutes in Europe and the USA which participate in the module production. The application of the test system for automated tests of modules requires a dedicated readout and analysis software. The software and all components of the test system will be explained in detail in this thesis. Different types of faults on a module show a significant behaviour in particular tests. A dedicated fault finding algorithm searches for these signatures. A safe identification of faulty channels and a reliable information on the respective type of fault is important. It facilitates the reparation and enables the assessment of the appropriateness of a module for the insertion into the tracker. It will be explained which approach was used to find the most appropriate tests for that purpose and how the fault finding algorithm was optimized to give reliable results independent of the specific test setup. Finally, the algorithm is used for the qualification of more than 500 repeatedly measured modules with known failures to verify its suitability. All faulty channels are found and more than 90% of the faults are correctly identified. At the same time less than 0.02% of good channels are wrongly flagged as faulty. The assessment of the module quality is nearly independent of the particular setup and can be reproduced in about 96% of the cases of repeatedly measured modules. In total, less than 0.3% of the channels are faulty and about 98% of all modules are suited for the construction of the tracker