PhD students' seminars - Jaber AL RASHID2:00 pm | POLYTECH ANGERS |Room 10 | 62, avenue Notre-Dame du Lac | ANGERS
The January 12, 2023
Subject : Reliability Assessment Model for Predicting Long-term Electromagnetic Robustness of Integrated Circuits based on the Accelerated Degradation Test
Abstract
Intergrated circuits (IC) in electronic systems demonstrate significant evolution of its electromagnetic compatibility (EMC) performance when subjected to external environmental (i.e. temperature, humidity, electrical stress) conditions. Degradation or failure mechanisms are activated and accelerated during the ageing process, resulting significant impact in the EMC performance and functionality of the IC operating within the electronic system. Hence, integrating IC-EMC immunity model with the predictive reliability model is essential to estimate the long-term electromagnetic robustness of the ICs under various external operating conditions.
The current ongoing research work has focussed to develop the predictive reliability model based on the degradation process observed while conducting the Direct Power Injection (DPI) test. Accelerated degradation test (ADT) plan was designed and the high temperature operating lifetime (HTOL) tests were performed on two different voltage regulator IC references (i.e. UA78L05 and L78L05). The total number of 12 samples (i.e. 6 units) of the considered ICs were subjected to step-stress accelerated degradation test (SSADT) inside the climatic chamber for a total stress duration of 950 hours. The DPI test was performed on two different voltage regulator IC references (i.e. UA78L05 and L78L05) to evaluate and investigate the evolution of its conducted electromagnetic immunity throughout the ageing process. The presentation of this work would present EMC performance degradation results evaluated based on the the DPI immunity test results measured for all these 12 samples under nominal conditions after every 100-200 hours stress intervals.
Mean power injected immunity drift were evaluated at different ageing time to produce degradation curve for each of the 12 IC samples. Based on the measured degradation data, physics based models (i.e. Arrhenius and Inverse-power law) and statistical model (i.e. Gama process) were applied to develop the degradation model for the degradation path of the performed SSADT. Maximum Likelihood estimation (MLE) method was applied to estimate the unknown parameters, followed by developing the analytic cumulative distribution function (CDF) for the defined SSADT performed on these selected ICs to predict the reliability performance parameters (i.e. time-to-failure (TTF), probability of failure) based on the defined degradation or failure threshold criterion. Finally, the reliability function would be developed to predict long-term EMC performance in terms of conducted immunity expressed as power injected (i.e. measured in dB) for any untested stress-conditions (i.e. temperature and voltage). Moreover, this model would also allow to perform predictive assessment to estimate the failure lifetime of these ICs, considering the specified failure criterion defined on the EMC degradation parameter (i.e. power injected drifts) evaluated based on the DPI measurement data (i.e. power injected and frequency).
