Authors: Thomas Nirmaier, Infineon Technologies AG, DE; Manuel Harrant, Infineon Technologies AG, DE; Monica Rafaila, Infineon Technologies AG, DE; Georg Pelz, Infineon Technologies AG, DE; Jerome Kirscher; Zhanat Maksut
Abstract:
Abstract
Automotive power micro-electronics complexity is driven by an ever and ever increasing demand for energy effi-ciency and safety. As a result of this complexity the error detec-tion latency exceeds practical limits in standard post-silicon vali-dation for mixed-signal power devices and often results in late detection of device failures in the final automotive application. Car manufacturers therefore push for application robustness beyond classical compliance to specification, which requires methods to assess robustness and enable quantification with an appropriate metric with respect to specific mission profiles.
Though there is a natural understanding that robustness means to be functional under a wide range of operating environments, there is no natural metrics obvious to robustness. We therefore first want to discuss different existing measures of location and spread for use as robustness metric with respect to a device speci-fication and or application mission profile.
The fundamental differences for the assessment of robustness in analog and mixed-signal systems are discussed and robustness metrics for an example battery management system IC with respect to a specification and with respect to a mission profile are calculated. The results show how quantifiable robustness metrics can be derived to assess real application fitness for a semiconduc-tor device or an electronic control unit.
1st RIIF Workshop @ DATE13
Publication Date: 2013/03/22
Location of Publication: 1st RIIF Workshop @ DATE13, DATE Conference, Grenoble, Frankreich
Keyword: Automotive
