Modeling and Validation of Common-Mode Emissions in Wide Bandgap-Based Converter Structures
Modeling and Validation of Common-Mode Emissions in Wide Bandgap-Based Converter Structures
复制标题
基于宽带隙的转换器结构中共模发射的建模和验证
DOI：
--
复制
发表时间：
2020
影响因子：
6.7
通讯作者：
Brian T. DeBoi
中科院分区：
文献类型：
--
作者：
A. Lemmon;Aaron D. Brovont;Christopher D. New;Blake W. Nelson;Brian T. DeBoi
已知设计有宽带隙（WBG）半导体的现代功率转换器产生大量传导电磁干扰（EMI）作为高边沿速率和高频开关的副作用。本文提供了一个统一的处理方法来推导共模等效模型，这是有用的分析这种增加的EMI签名。还包括这种方法的一步一步的示范应用于原型转换器，证明了许多实际的电力电子应用。对于开发的模型的经验验证，该转换器是在一个定制设计的EMI特性测试平台，这是类似于由MIL-STD-461 CE-102规定的设置操作。在时域和频域的分析预测和经验结果之间取得了良好的一致性。频谱比较被证明是特别准确的频率范围在10 kHz和30 MHz之间，这是感兴趣的主要范围控制在基于WBG的系统中的传导发射。为了实现所演示的模型协议，一组寄生参数值，通过阻抗表征的系统下的研究。识别的关键寄生元素，必须量化，以实现良好的预测能力，为所提出的模型代表了本文的具体贡献之一。
Modern power converters designed with wide band-gap (WBG) semiconductors are known to generate substantial conducted electromagnetic interference (EMI) as a side effect of high-edge-rate and high-frequency switching. This article provides a consolidated treatment of an approach to the derivation of common-mode equivalent models that are useful for analyzing this increased EMI signature. Also included is a step-by-step demonstration of this approach applied to a prototype converter that exemplifies many practical power electronic applications. For empirical validation of the developed model, this converter is operated within a custom-designed EMI characterization testbed that is similar to the setup specified by MIL-STD-461 CE-102. Good agreement is achieved between analytical predictions and empirical results in both the time domain and the frequency domain. Spectral comparisons are shown to be especially accurate in the frequency range between 10 kHz and 30 MHz, which is the principal range of interest for controlling conducted emissions in WBG-based systems. In order to achieve the demonstrated model agreement, a set of parasitic parameter values were obtained through impedance characterization of the system under study. Identification of the critical parasitic elements that must be quantified to achieve good predictive capability for the presented model represents one of the specific contributions of this article.