A BAYES APPROACH AND CRITICALITY ANALYSIS FOR RELIABILITY PREDICTION OF AlGaInP LIGHT EMITTING DIODES

While use of LEDs in fiber optics and lighting applications is common, their use in medical diagnostic applications is very rare. Since the precise value of light intensity will be used to interpret patient results, understanding failure modes is very important. We used the Failure Modes and Effects Criticality Analysis (FMECA) tool to identify the critical LED failure modes. Once the critical failure modes were identified, the next step was the generation of time to failure distribution using Accelerated Life Testing (ALT) and Bayesian analysis. ALT was performed on the LEDs by driving them in pulse mode at higher current density J and higher temperature T. This required the use of accelerating agent modeling. We have used Inverse Power Law model with J as the accelerating agent and the Arrhenius model with T as the accelerating agent. Such power law dependence originates directly from the electromigration assumption of the failure mechanism, The Bayesian modeling began by researching published articles that can be used as prior information for Bayesian modeling. From the published data, we extracted the time required for the optical power output to reach 80% of its initial value (our failure criteria). Analysis of published data for different LED Materials (AlGaInP, GaN, AlGaAs), the Semiconductor Structures (DH, MQW) and the mode of testing (DC, Pulsed) was carried out. This data was converted to application conditions of the medical environment. Many of the LED degradation mechanisms occur simultaneously. The weakest link causes the actual failure. This leads us to believe that Weibull distribution is the most suitable distribution for time to failure of the LEDs. We used this rationale to develop the Bayesian likelihood function. In this study, we report the results of our ALT and develop the Bayesian model as an approach for analyzing LED suitability for numerous system applications.