To attend via videoconference, email andrew.b.nadler.th@dartmouth.edu.

"Analysis, Optimization, and Design of Small-Scale Hybrid-Core Inductor Designs"

Abstract

As the size of power converters and other electronics has shrunk over time, miniaturization of passive magnetic components, namely inductors and transformers, has fallen behind. Much of this is due to magnetic scaling laws, which degrade inductor performance at reduced sizes, in contrast to other common power converter components, like transistors and capacitors, which are conducive to construction from smaller parallel cells. This does not mean high performance and small sizes are unachievable with magnetic components, but rather intelligent and creative magnetics design is increasingly critical.

Typical inductors utilize a conductive winding wrapped around a magnetic core, with loss occurring in both places. The heat released from this loss often sets minimum inductor size and, consequently, maximum energy density. However, much of what prevents reducing loss, and specifically winding loss, are core material saturation considerations that set the minimum winding turns, since the best low loss core materials, such as ferrites, have low saturation flux densities while high saturation flux density materials, such as steels, have significant core loss, especially at high frequencies.

This work explores two novel hybrid core inductor types, a hybrid steel core and a hybrid permanent magnet (PM) core, that achieve small size and increased energy density by strategically utilizing the strengths and mitigating the weaknesses of their core materials. Theory behind these two hybrid core designs is established with reluctance models, and these models are used in an optimization routine to develop designs and characterize spaces where hybrid cores provide the most benefit. Finite element simulation procedures unique to each hybrid core are used to verify designs and refine the reluctance model. Simulation results show a hybrid steel core inductor at a frequency of 1.5 MHz in a 1210 package increases energy density 600% compared to a ferrite core inductor at the same DC resistance while a hybrid PM core inductor increases energy density 200% at 4 MHz. The successful completion of a hybrid PM prototype in a 1210 package is also discussed, including a new winding construction method implementing electroplating to form a 3D winding.

Thesis Committee

• Charles Sullivan (chair)
• Jason Stauth
• William Scheideler
• David Perreault (MIT)

Contact

For more information, contact Theresa Fuller at theresa.d.fuller@dartmouth.edu.