A Comparative Study on Reliability and Ruggedness of Kelvin and Non-Kelvin Packaged SiC Mosfets

Abstract

This article evaluates silicon carbide (SiC) <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">mosfet’</small> s reliability and ruggedness comprehensively between two widely employed device package types: common source and Kelvin source. Multiple accelerated lifetime tests are conducted to comprehensively investigate SiC <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">mosfet</small> ’s reliability at both package and die levels. Specifically, a conventional high electric field test is applied to induce gate oxide degradation, whereas package degradation is investigated by the widely adopted power cycling test. To mimic gate oxide degradation patterns closer to realistic cases, an active channel gate bias test is proposed in this study, and a fair comparison is made regarding the device's degradation pattern in the conventional high electric field test. Throughout accelerated lifetime tests, devices’ electrical parameters, consumable lifetime, and end-of-life failure modes are comprehensively evaluated. For ruggedness comparison, a single-shot short-circuit test is conducted on both common-source and Kelvin-source packaged devices with the same die technique and configuration. Other than analytical discussions with experimental waveforms, failure analysis is also conducted to investigate the device's failure mode and root cause. In respect of experimental results in accelerated lifetime tests and short-circuit tests, the existence of the common-source inductance and resistance may mitigate the electrothermal stress applied on the chip with lower gate–source voltage under load current and results in mitigated device instability and longer consumable lifetime.