Reduced-Order Ladder Network Modeling for Common-Mode Characterization of Transformers

Abstract

The ladder network with detailed parameters and explicit physical meanings is beneficial for the wideband modeling of transformers. However, due to an excessive number of sections, the high complexity of the model limits its application in the common-mode analysis of transformers working in power converters. In order to improve the applicability, in this article, the correlation between the resonant frequency of each section pair and the validity of the model is analyzed first. A valid frequency range prediction is derived then to evaluate whether the order of the model is reasonable, which forms the foundation of the proposed reduced-order modeling method. This method enables accurate transformer modeling using a ladder network with much fewer parameters. Furthermore, the reduced-order transformer model is also applied to meet the demand in the balance winding techniques and the wideband common-mode noise analysis. Finally, the effectiveness of the proposed method and its applications are verified by the insertion voltage gain of transformers and the common-mode current of the converter.