Thermo-Mechanical Stress Comparison of a GaN and SiC MOSFET for Photovoltaic Applications

Integrating photovoltaic applications within urban environments creates the need for more compact and efficient power electronics that can guarantee long lifetimes. The upcoming wide-bandgap semiconductor devices show great promise in providing the first two properties, but their packaging requires...

Full description

Bibliographic Details
Main Authors: Wieland Van De Sande, Omid Alavi, Philippe Nivelle, Jan D’Haen, Michaël Daenen
Format: Article
Language:English
Published: MDPI AG 2020-11-01
Series:Energies
Subjects:
Online Access:https://www.mdpi.com/1996-1073/13/22/5900
Description
Summary:Integrating photovoltaic applications within urban environments creates the need for more compact and efficient power electronics that can guarantee long lifetimes. The upcoming wide-bandgap semiconductor devices show great promise in providing the first two properties, but their packaging requires further testing in order to optimize their reliability. This paper demonstrates one iteration of the design for reliability methodology used in order to compare the generated thermo-mechanical stress in the die attach and the bond wires of a GaN and SiC MOSFET. An electro-thermal model of a photovoltaic string inverter is used in order to translate a cloudy and a clear one-hour mission profile from Arizona into a junction losses profile. Subsequently, the finite element method models of both devices are constructed through reverse engineering in order to analyze the plastic energy. The results show that the plastic energy in the die attach caused by a cloudy mission-profile is much higher than that caused by a clear mission-profile. The GaN MOSFET, in spite of its reduced losses, endures around 5 times more plastic energy dissipation density in its die attach than the SiC MOSFET while the reverse is true for the bond wires. Potential design adaptations for both devices have been suggested to initiate a new iteration in the design for reliability methodology, which will ultimately lead to a more reliable design.
ISSN:1996-1073