These slides were presented at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

This study proposes a systematic and automated methodology for optimizing MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) power modules, essential components in high-power electronic systems requiring efficient operation and robust thermal management. The approach integrates multi-domain simulations into a unified optimization framework, enabling simultaneous improvements in electrical and thermal performance, and addressing key challenges in power module design, such as managing high voltages, currents, and power consumption. The optimization targets critical design parameters, including MOSFET placement, wire bond diameter, and substrate layout, to minimize parasitic inductance and thermal resistance while maximizing power-handling capacity. The process begins with schematic-driven design entry, utilizing tools like Cadence Allegro PKG Designer, which ensures accurate structural definition and seamless integration with subsequent analysis steps. Parasitic extraction is conducted using Clarity 3D quasi-static solvers, and the results are incorporated into PSPICE time-domain simulations to evaluate switching behaviors, voltage and current waveforms, and power losses. Thermal analysis using Cadence Celsius assesses junction temperatures, current density, and heat dissipation to identify and mitigate potential thermal bottlenecks, ensuring reliable operation under high-power conditions. Moreover, S-parameter extractions identify electromagnetic coupling between power paths and gate paths, uncovering potential resonances and electromagnetic compatibility issues that could degrade module performance, particularly at higher switching frequencies. By integrating these analyses into an iterative optimization loop, the framework achieves measurable performance improvements. Automated workflows implemented using Optimus significantly streamline the design process, enabling efficient exploration of design trade-offs. The optimization of wire bond parameters, MOSFET positioning, and substrate layout enhances both electrical and thermal performance, leading to reduced power losses, improved thermal dissipation, and increased operational reliability. This work demonstrates the advantages of a holistic, simulation-driven approach for power module design. The proposed framework balances electrical and thermal considerations, offering a scalable and efficient methodology applicable to automotive, telecommunications, and industrial electronics. The findings underscore the potential of automated design optimization to meet the rigorous demands of modern high-power electronic systems, advancing the development of high-performance power modules.