Power Efficiency and EMI Attenuation Optimization in Filter Design - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue IEEE Transactions on Electromagnetic Compatibility Année : 2018

Power Efficiency and EMI Attenuation Optimization in Filter Design

Résumé

This paper presents a discrete conducted electromagnetic interference filter optimization procedure, based on a Genetic Algorithms. A macro-modeling technique taking into account the load impedance, the source emissions and the filter parasitic components, including the filter layout, is used to obtain an accurate solution for the optimization process. The latter searches among supplier passive component databases and provides, for a given filter topology, an optimal set of components available on the market. This approach has been applied to a differential Class-D audio amplifier for validation. By considering the EMI, the additional power losses introduced by the filter and the audio gain, two different optimization formulations have been tested. The first corresponds to maximizing the power efficiency of the system while respecting a determined level of electromagnetic emissions. The second corresponds to minimizing the EMI without exceeding a determined level of power loss. The optimized filters are built and measurements are carried out. The results show a remarkable power efficiency improvement and a significant EM emission reduction when compared to a reference filter.
Fichier principal
Vignette du fichier
2017_FERBER_MRAD_IEEE_EMC_04.pdf (1.31 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01887159 , version 1 (15-10-2018)

Identifiants

Citer

Moises Ferber, Roberto Mrad, Florent Morel, Gaël Pillonnet, Christian Vollaire, et al.. Power Efficiency and EMI Attenuation Optimization in Filter Design. IEEE Transactions on Electromagnetic Compatibility, 2018, 60 (6), pp.1811 - 1818. ⟨10.1109/TEMC.2017.2765921⟩. ⟨hal-01887159⟩
164 Consultations
720 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More