Reliability Assessment of Telecommunications Equipment

This thesis studies the reliability of telecommunications equipment, its components, and the systems made using those components. Special attention is paid to creating stronger links between the reliability analyses performed at different hierarchy levels. The thesis starts with a temperature derating study. It is found out that the generic handbook based procedures may not always be very attractive, as they do not take satisfactorily into account the actual lifetime requirements. An alternative approach is proposed as a remedy to the current situation. Thermal cycling requirement handbooks are surveyed, and based on the findings some enhancements are proposed. Next, a component and product specific approach to create thermal cycling requirements is suggested. When applying the new approach several factors can be taken into account: the product’s lifetime requirement, the field environment, the reliability test result, and the statistical distribution of the component population. A new method of how to predict the reliability of a component population that is addressed to several, different field environments is presented. Ceramic, leadless components are studied by testing and by utilizing Engelmaier’s analytical solder fatigue model and Finite Element (FE) simulations. A new approach to interpret the solder joint height in conjunction with solder castellations is introduced. Based on this, a very good correlation between the test results and the predictions based on Engelmaier’s model can be obtained. The parameter sensitivity of both the Engelmaier’s model and the FE analysis are studied and compared. Error margins based on the parameter sensitivity studies are given. Time-averaged hazard rate functions are studied in order to be able to use component level test data in simplistic parts-count method type reliability predictions. Finally, the availability of a full 3 generation telecommunications network is studied.