Abstract
Lifetime prediction for microelectronic devices and systems is complicated by many factors including the validity of linear acceleration, choice of extrapolation model, presence of multiple failure mechanisms with common driving forces, correlation between failure mechanisms, time-variant loading (voltage pulses) etc… With real-time prognostics and health management coming up as a useful alternative to conventional post-failure reliability data analysis, significant progress has been made in estimating the individual lifetime of microelectronic devices/systems during operation (real-time). In this study, we present a case study of decoding the contributions of the bias temperature instability (BTI) and hot carrier injection (HCI) mechanisms to the overall time-dependent threshold voltage (VTH) shift observed in real-time during a conventional HCI stress test applied to a single NMOS device. Assuming no prior knowledge of the time exponents for VTH degradation for both the BTI and HCI mechanisms, our methodology enables us to deconvolute the overall VTH data signal, predict the remaining useful life (RUL) for the device (given a threshold failure criterion) and extract the distribution of the power-law exponents for the pure-HCI and pure-BTI mechanisms. We used the particle filter based sequential Monte Carlo (SMC) technique here for the prognostic study and the advantage of our approach is its generic use for non-linear systems and non-Gaussian noise trends. The impact of prognostics based data-driven algorithms in dynamic lifecycle estimation of microelectronic devices is evident in this work and such an approach can be handy in high-power space electronics applications when the reliability (health/robustness) of a single device (integrated in the satellite) needs to be studied (under normal operating conditions) and there is no large sample size population of similar devices available for a conventional accelerated stress test exercise off-field. To our knowledge, this is the first study applying the particle filter technique for a multiple failure mechanism scenario. The accuracy of our RUL estimates was compared with real data extracted from past experimental studies.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.