Power Semiconductor Failure Analysis To Improve Research

After long-term of development and mass-production, power semiconductor devicemanufactures are dealing with large quantity orders and suffering from high defect risks.Manufactures are also restricted by limited raw material and human resources input dueto low profitability of power semiconductor products because of the intense marketcompetition. It is one of the primary aims for manufactures to seek ways of improveproducts’ reliability and market survival ability to expand market space. The failureanalysis is wide concerned by various manufactures as a steady way of improvingproducts’ reliability.This paper starts from finding ways to ascend the reliability of products, and thenfocus on the Failure Analysis Technology which is one of the most important methodsto ensure and improve products’ reliability. After investigated advance failure analysistechnology over global, then summarize the main operation process, compared withcompany’s existed process. Researches have been established improvement targets over5major aspects which are man, equipment, process, operation and environmentalimpacts. In order to research in an accurate and efficient orientation, those5aspectsdescribed above are analysis feasibility with statistical methods. Phases of present statsanalysis, research direction choosing, theoretically design and practically confirmationare shown in this paper. Achievements of this project are summoned as below.(1) Improve failure analysis hardware’s and human resources’ abilitiesSeveral tools like test circuit and module are designed produced and managed bymeasurement system independently to improve the efficiency and accuracy of products’electronic stats measuring. Established standard procedures of pre-job training, on-jobrecognition and periodic examination are established in company to meet up with job’sand ISO19001:2008and TS16949:2009quality system’s requirement.(2) Improved efficiency of analysis process.Establish efficient process by theoretical analysis and accumulation experiment onkey steps. QC tools which are widely used in manufactures are also used in statistic analysis of index data.(3) Develop failure analysis database.In order to improve accuracy of analysis conclusion, use Excel software developfailure analysis database for engineers to retrieval data from information of historyreports in analysis process.This research has reached the expected results. Not only promoted company passesthe TS16494:2009certification that brought broader market space, but also madeimportant contributions to improve products’ reliability and promote continuallyprogress of company’s operation efficiency.