This work presents a comparative analysis among four power MOSFET technologies: conventional Silicon (Si), Superjunction (SJ), Silicon Carbide (SiC) and Gallium Nitride (GaN), indicating the voltage, current and frequency ranges of the best performance for each technology. For this, a database with 91 power MOSFETs from different manufacturers was built. MOSFET losses are related to individual characteristics of the technology: drain-source on-state resistance, input capacitance, Miller capacitance and internal gate resistance. The total losses are evaluated considering a drain-source voltage of 400 V, power levels from 1 kW to 16 kW (1 A–40 A) and frequencies from 1 kHz to 500 kHz. A methodology for selecting power MOSFETs in power electronics applications is also presented.

A design methodology for minimum volume of LCL filters applied to grid connected converters is proposed. It combines the determination of filter parameter values (inductances and capacitances) to hardware design (component technology and inductor construction). Using the proposed strategy, different combinations of L-C-L that meet standard restrictions are determined. The influence of the harmonic content that results from filter design is considered in order to estimate component losses, and filter and bus capacitor lifetimes. Results are presented for filters that lead to the smallest volume, highest capacitor lifetime, or a compromise between both. A case study with three different magnetic material technologies and two types of capacitors is done. The design methodology is experimentally validated for a 1 kW converter. Step-by-step procedures for the determination of filter parameters and inductor hardware design are provided.

No abstract

A simple and accurate analytical model for the estimation of switching losses on power MOSFETs is proposed. It consists of simplifying the non‐linear behaviour of Miller capacitance as a function of voltage. Experimental results are used to validate the model in the 5–500 kHz range. The proposed analytical model is compared to other frequently used methods. Results confirm the accuracy of the proposed model in different voltage levels, using four different MOSFET part numbers, spanning three technologies: SiC, superjunction, and conventional silicon. Because of its simplicity of implementation, it is especially recommended for applications that design converters by evaluating a large database of transistor part numbers.