Accurate loss estimation method in IGBT and MOSFET based inverters

• An analytical breakdown of power losses in three-phase SiC and Si-IGBT inverters. • A high-fidelity evaluation of switching and conduction losses under practical operating conditions. • A power-loss estimation framework derived from device datasheets, explicitly accounting for junction temperature, gate resistance, and related parameters. • A detailed analytical model for inductor core losses based on the modified Bertotti method. • Experimental hardware validation of SiC and Si IGBT inverter losses at various operating conditions (power levels, modulation indices, and switching frequencies). • 95 measurements provides, ensuring reliable and consistent performance validation beyond conventional single-point results. • Correlation of measured, simulated, and analytically computed power losses demonstrates that the proposed model maintains over 95 accuracy across variations in modulation index, switching frequency, and power level, with its domain of validity extending from low-power to rated-power operation. Accurate power-loss estimation is vital for high-frequency inverters, where efficiency and reliability are influenced by switching, conduction, and passive component losses. Existing datasheet-based, analytical, and simulation-driven methods often neglect operating-condition dependencies or oversimplify passive losses, leading to significant deviations from real performance. This paper presents a comprehensive loss estimation framework for silicon carbide (SiC) MOSFET and silicon insulated-gate bipolar transistor (Si IGBT) inverters. The proposed approach accounts for junction temperature, gate resistance, and nonlinear device behavior, while also incorporating inductor core losses using a modified Bertotti method. Loss components are analytically modeled under realistic operating conditions and validated against measurements from a hardware prototype. The experimentally measured losses under varying power levels, modulation indices, and switching frequencies showed strong agreement (over 95%) with analytical and PLECS simulation results for both Si IGBT and SiC MOSFET-based inverters. At 4 kW, deviations of 7.6% for Si IGBT and 6.8% for SiC MOSFET-based inverter were observed due to elevated junction temperatures. Overall, the proposed model maintains high accuracy across different operating conditions, validating its applicability from low to rated power.