1 December 2014 in IEEE Transactions on Power Electronics
The design of optimal electromagnetic compatibility filters for converters used in electrical networks requires the precise knowledge of input and output terminal impedances. For this reason, the knowledge of the network impedance, including all connected converters, is determined. This paper proposes a study of both the common mode (CM) and the differential mode (DM) impedances of modern embedded networks including cables and several converters. Using a dc network for illustration, we determine the impedance taking into account all operating states of a converter (online or offline operations). Without any restrictions, a theoretical model is developed to express the CM and DM impedances using a switching function and a Fourier series decomposition. An experimental validation has been carried out in a frequency range from a few Hertz up to 80 MHz. Our model has then been used in a simple network, so as to compute the network impedance. Measurements have confirmed the proposed model accuracy.
30 July 2021 in Advanced Electromagnetics
In this paper, an optimization method of toroidal core-based Hybrid common-mode chokes (HCMCs) for the design of an Electromagnetic interferences (EMI) filter is proposed. A dedicated algorithm is developed using MATLAB to characterize compact HCMCs that exhibit effective CommonMode (CM) chokes with optimized leakage inductances by systematic variations in the winding patterns and geometric dimensions of their magnetic cores. It takes into consideration the physical limitations of these components as well as the constraints related to the design of EMI filters. The proposed algorithm allows through a small computational task to propose a variety of configurations for optimal HCMCs. Finite element method (FEM) simulations are conducted on the HCMCs to extract their CM and leakage inductances. The results are compared with those calculated analytically and yield a good match. The performances of optimized HCMCs are evaluated through their implementation in the designed filter. All the cases of HCMCs including the smallest one allow the EMI filter to easily qualify a power converter to an electromagnetic compatibility (EMC) standard.
1 January 2019 in Progress In Electromagnetics Research B
MBMohamed Amine Cheurfi Belhadj
Mohamed Amine Cheurfi Belhadj
Since electromagnetic compatibility studies intend to predict the compliance with electromagnetic standards, an accurate computation of both common and differential mode conducted noises is necessary. Modern networks such as in automobiles that are known for supplying many electrical actuators-include many power converters and long cables (conductors) to efficiently manage power transfer. However, the presence of both converters and cables creates new electromagnetic compatibility issues. For example, the interaction between cables and converters becomes a noise source. For this reason electromagnetic compatibility study becomes more complex. Therefore, the purpose of this paper is an attempt to propose an analytical model that computes noise sources by generating conducted signals within the network at any site, meaning all along the cable according to the CISPR16 standard. Our approach primarily consists of modeling conducted noise sources generated by converters connected to the DC-network which are extracted and identified in both frequency and time domains. The electromagnetic compatibility modelling of converter's behaviour is performed by defining a mathematical switching function. The model is assessed with time domain simulations and identified by experimental measurements. Secondly, the extracted converter's model, based on equivalent noise sources, is used to predict the conducted noise inside a defined network at any location of the cable. The process of the network's modelling is realised through using the Multi-Transmission Line Method of lossless lines. This network's model is crucial for EMC analysis in order to evaluate the interaction degree between noise sources and cable parameters.