Litcius/Paper detail

Improved Methodology for Parasitic Analysis of High-Performance Silicon Carbide Power Modules

Brian T. DeBoi, Andrew N. Lemmon, Brice McPherson, Brandon Passmore

2022IEEE Transactions on Power Electronics14 citationsDOI

Abstract

The high edge rates of wide bandgap (WBG) semiconductors are known to excite resonances within device packaging, resulting in increased voltage overshoot and electromagnetic signatures. Simulation tools based on finite-element analysis (FEA) are commonly used during design to optimize device geometry and minimize parasitics. However, developing an accurate model to extract these parameters is challenging and requires an extensive understanding of the simulation environment. This article presents three custom power module geometries that are measured empirically and evaluated in COMSOL multiphysics. The included sensitivity analysis demonstrates that even a well-designed FEA model may offer poor predictions of module parasitics compared to empirical measurements. A virtual compensation procedure is then proposed that greatly improves the accuracy of the COMSOL FEA predictions by removing the contribution of the return path. This approach does not require empirical measurements to implement and can be applied generally to a variety of geometries. Finally, the technique is applied to a high-performance WBG power module, for which good agreement is demonstrated between empirical measurements and simulation predictions.

Topics & Concepts

MultiphysicsParasitic extractionFinite element methodSilicon carbideElectronic engineeringPower (physics)ChipComputer scienceMechanical engineeringEngineeringElectrical engineeringMaterials sciencePhysicsQuantum mechanicsMetallurgyStructural engineeringSilicon Carbide Semiconductor TechnologiesElectromagnetic Compatibility and Noise SuppressionAdvanced DC-DC Converters
Improved Methodology for Parasitic Analysis of High-Performance Silicon Carbide Power Modules | Litcius