Open Collections

Accurate power loss models of complex converters

University of British Columbia

Power loss estimation is essential for the design and optimization of power converters. Traditionally, the power loss estimation is done using datasheet-level information about the switching devices. However, this information is limited to isolated operating points, compromising the accuracy of the estimation. In addition, the representation of PCB effects and gate driver, which cannot be considered by using datasheet parameters, increase the complexity of the tasks by adding more unknown magnitudes and more complex equations without guaranteeing their correct estimation. In this work, a novel power loss estimation method, which can be applied to any topology, is presented. This proposed method utilizes Design of Experiments (DoE) and Response Surface Methodology (RSM) to model the different types of losses in all the elements in a power converter such as power switches, diodes, and inductors. With RSM, simple equations explain the different types of losses as a function of variables that directly affect them allowing the losses estimation under any operating condition accurately. These models are extensively validated to show the advantage of using the proposed method and significant improvement with respect to datasheet calculations is obtained. For complex topologies such as Power Factor Corrector (PFC), accurate power loss estimation is not only necessary but also a challenge. To show the capabilities of the proposed model, it is applied to Universal PFC, which is a complicated converter in terms of loss calculation and analysis. Using the proposed models, the inductor is designed optimally and the optimal switching frequency is selected to minimize the losses in the different input voltage levels. Experimental validation of the proposed method, prediction, and design impact is presented; the proposed method yields a power loss reduction up to 25%, having the greatest improvement at an input voltage of 110 V, and an output power higher than 500 W.

Text

2019-12-16

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/72768

Electrical and Computer Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/