Litcius/Paper detail

Active cancellation of periodic EMI of power electronic systems by injecting artificially synthesized signals

Andreas Bendicks, Tobias Dörlemann, Timo Osterburg, Stephan Frei

2020IEEE Electromagnetic Compatibility Magazine20 citationsDOI

Abstract

Active EMI cancellation is an interesting solution to reduce the size of passive filter components. In this work, the already established active EMI filter types are summarized. The systematic limitations are elaborated that result from the feedback or feedforward approach for cancellation signal generation. Afterward, a new method is presented in which the cancellation signal is artificially synthesized and injected into the system. This method is especially well suited for the suppression of periodic EMI since the anti-EMI can conveniently be synthesized from cancelling sine waves. Bothersome effects, like complex frequency characteristics or delays, can be compensated by appropriate amplitudes and phases. By doing so, systematic limitations of active EMI filters can be resolved. In this work, the fundamental concept and possible variants are depicted. Although many limitations of active EMI filters are resolved, also the new method has its limits. These are described and discussed. Furthermore, different methods for determining the cancelling sine waves are summarized. Last, a prototype cancellation system is presented and discussed to suppress the EMI of a GaN-based 48 V/12 V DC-to-DC converter with a switching frequency of 300 kHz. The disturbances are actively suppressed below the strictest limit (class 5) of the automotive standard CISPR 25 in the frequency range from 150 kHz to 30 MHz. The active suppression of the cancellation system and the passive attenuation of the coupling circuits are individually investigated to elaborate the performance of the active cancellation strategy. The power of the generated cancellation signal is estimated.

Topics & Concepts

EMIElectronic engineeringSine waveElectromagnetic interferenceSIGNAL (programming language)Feed forwardFilter (signal processing)Computer sciencePower (physics)EngineeringElectrical engineeringPhysicsVoltageControl engineeringQuantum mechanicsProgramming languageElectromagnetic Compatibility and Noise SuppressionRadio Frequency Integrated Circuit DesignPower Line Communications and Noise