Date of Award

5-2018

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical and Computer Engineering

Committee Chair

Scott Sudhoff

Committee Member 1

Dionysios C. Aliprantis

Committee Member 2

Steven D. Pekarek

Committee Member 3

Dan Jiao

Abstract

The objective of this work is to propose a design methodology for a high-frequency transformer in the context of an isolating DC-DC converter. The design challenges relating to high-frequency operation include transformer parasitics, high-frequency loss mechanisms, and their coupled effect on the performance of the transformer and DC-DC converter. These issues are addressed in this work by incorporating high-frequency effects into the relevant magnetic, electrical, and thermal analyses needed for multi-objective optimization based design.

First, an accurate and computationally-efficient method for the transformer magnetic analysis is proposed. Leakage inductance is calculated using a procedure based on the Biot-Savart law and the method of mirror images. The method is validated for a prototype high-frequency transformer. Another key issue addressed is high-frequency transformer loss estimation. In this regard, the winding loss due to proximity effect is analyzed. A simplified time-domain model of the DC-DC converter is set forth to determine the transformer currents necessary for loss estimation. Next, analytical methods are used to estimate the parasitic capacitances in a transformer. The high-frequency transformer common-mode and differential mode impedances are derived. A transformer thermal analysis is set forth to estimate the temperature rise in the windings and to include the thermal effect on conductor material parameters. Finally, these analyses are coupled using a multi-objective optimization in order to create a new comprehensive and automated high-frequency transformer design paradigm.

Download

Share

COinS