Core Loss Model for Arbitrary Excitations with DC Bias Covering a Wide Frequency Range

Stenglein E, Dürbaum T (2021)


Publication Type: Journal article

Publication year: 2021

Journal

DOI: 10.1109/TMAG.2021.3068188

Abstract

Core loss prediction is quite complex as the excitation of magnetic materials varies depending on the topology and the operating mode of switched-mode power supplies. In particular, the waveform of the magnetic flux density including its dc bias affects core losses. This paper presents the results of an extensive core loss study performed on Mn-Zn ferrites. The investigation comprises core excitations with, among others, sinusoidal and triangular shapes of the magnetic flux density with and without dc bias. Based on the obtained core loss data, an empirical core loss model has been derived for the prediction of specific core losses in Mn-Zn ferrites that is valid for arbitrary excitations including dc bias of the magnetic flux density. Multiplying the quasi-static energy losses depending on the swing and the dc bias of the magnetic flux density by an effective frequency yields the core losses. The proposed model covers a wide frequency range and only requires the quasi-static energy losses at low frequencies as well as two frequency-independent parameters. These parameters can be easily extracted from a limited number of measurements or from the data provided by manufacturers. The verification of the proposed approach demonstrates an adequate accuracy of 15% for all investigated Mn-Zn ferrites as well as for the broad range of studied waveforms and parameters.

Authors with CRIS profile

How to cite

APA:

Stenglein, E., & Dürbaum, T. (2021). Core Loss Model for Arbitrary Excitations with DC Bias Covering a Wide Frequency Range. IEEE Transactions on Magnetics. https://dx.doi.org/10.1109/TMAG.2021.3068188

MLA:

Stenglein, Erika, and Thomas Dürbaum. "Core Loss Model for Arbitrary Excitations with DC Bias Covering a Wide Frequency Range." IEEE Transactions on Magnetics (2021).

BibTeX: Download