Search for: All records

This paper describes challenges to automated bulk collection of temperature-controlled magnetic core loss data in the 1-20 MHz regime. Oil immersion is shown to alter the small signal impedance of ML91S and FR67 ferrite cores by more than 5% over part of their rated frequency ranges which prevents accurate estimation of core loss. Air-based thermal forcing is shown to be a viable alternative to oil for core temperature regulation in high frequency core loss testers. Temperature regulation to 25 ± 3°C is demonstrated on FR80 at 1 MHz up to 1.25 W of dissipation. 

Magnetic core loss measurement methods suitable for high-frequency sinusoidal excitations are currently time intensive or inherently suffer from flux drive limitations due to the costly and ill-suited radio frequency (RF) equipment utilized in the measurement. The recent development of automated testers has enabled the collection of large core loss data sets across a broad range of operating frequencies and flux densities. Improper loading of the RF power amplifiers utilized in these measurements makes the collection of accurate core loss data impractical above certain drive levels. In this paper, we develop and demonstrate an automatable core loss testing method which replaces the commonly employed RF power amplifier with a high frequency switching inverter. An analytical framework for assessing flux harmonics in the core is also presented. An experimental demonstration is performed in the range of 1-7MHz for the following materials: (1) Fair-Rite 67, (2) Fair-Rite 80, and (3) Proterial ML91S. 

This paper summarizes the main results and contributions of the MagNet Challenge 2023, an open-source research initiative for data-driven modeling of power magnetic materials. The MagNet Challenge has (1) advanced the stateof-the-art in power magnetics modeling; (2) set up examples for fostering an open-source and transparent research community; (3) developed useful guidelines and practical rules for conducting data-driven research in power electronics; and (4) provided a fair performance benchmark leading to insights on the most promising future research directions. The competition yielded a collection of publicly disclosed software algorithms and tools designed to capture the distinct loss characteristics of power magnetic materials, which are mostly open-sourced. We have attempted to bridge power electronics domain knowledge with state-of-the-art advancements in artificial intelligence, machine learning, pattern recognition, and signal processing. The MagNet Challenge has greatly improved the accuracy and reduced the size of data-driven power magnetic material models. The models and tools created for various materials were meticulously documented and shared within the broader power electronics community.