Research On Life Evaluation Methods Of Typical Integrated Circuits Based On Accelerated Degradation Analysis

The IGBT(Insulated Gate Bipolar Transistor)module is called an insulated gate bipolar transistor.It has the characteristics of low energy consumption,easy drive,large conduction current,and large withstand voltage,and it is widely used.With the gradual improvement of the manufacturing industry,system equipment and its components in various fields gradually show the characteristics of long life and high reliability.Traditional performance degradation failure data is becoming more and more difficult to obtain.How to perform degradation modeling and life prediction for an IGBT module,which is one of the equipment components,has become a research hotspot in the field of reliability.This article selects the IGBT module as the research object to study the law of its performance degradation during the working process,so as to carry out life prediction.The research content of this topic includes:1?Design and carry out the performance degradation test scheme based on the failure mechanism of the tested object.The IGBT module called IGRBC30 K is selected as the research object,and its common failure mechanism is analyzed.In view of the difficulty in obtaining failure data and degradation data under normal stress,the environmental stress that has a greater impact on its failure is selected to design the performance degradation test program under high stress.And complete the construction of the test platform according to the design plan.Perform a preliminary analysis of the data results detected during the test and summarize the data characteristics.2? Life prediction method based on degradation data.Analyze the relationship between performance degradation data and time under single stress.Research the commonly used life prediction methods based on performance degradation data.Aiming at the high overlap of degradation trajectories between samples,an improved Weibull function-based rapid life evaluation model is proposed.The right approximation method is used for parameter estimation and simulation verification of the method..Combining the characteristics of the performance degradation data in Chapter 2,an improved Weibull function-based rapid life assessment model is selected for modeling,and the life prediction results under various stresses are given,and the model method is verified by comparing the results with the linear model prediction results Reliability.3?Research on the relationship between life characteristics and different stresses.Analyze the relationship between failure data and stress under multiple sets of stress.The commonly used acceleration models in the research include the Arrhenius model related to temperature stress.The Weibull distribution is used to characterize the life characteristics of the failure data,and the construction method and steps of the acceleration model based on the Weibull distribution are given.Combining part of the performance degradation experiment failure data in Chapter 2,use the Arrhenius model for modeling,use the best linear unbiased estimation to obtain the model parameters,build an acceleration model at any two temperatures,and use the failure data that did not participate in the model construction Validate the acceleration model.Reveal the relationship between IGBT module life and temperature stress.4?The design of life prediction software.The overall framework and work flow of the software design are elaborated in detail,and the software development is realized based on the MATLAB platform.The monitoring data of the performance degradation test in Chapter 2 is used to verify the software functions and results.