Complementary MOS structures for common mode EMI reduction

Similar documents
Design and Realization of Autonomous Power CMOS Single Phase Inverter and Rectifier for Low Power Conditioning Applications

A New Approach to Modeling the Impact of EMI on MOSFET DC Behavior

Power- Supply Network Modeling

Wireless Energy Transfer Using Zero Bias Schottky Diodes Rectenna Structures

A 100MHz voltage to frequency converter

INVESTIGATION ON EMI EFFECTS IN BANDGAP VOLTAGE REFERENCES

A high PSRR Class-D audio amplifier IC based on a self-adjusting voltage reference

Floating Body and Hot Carrier Effects in Ultra-Thin Film SOI MOSFETs

A Busbar Like Power Module Based On 3D Chip On Chip Hybrid Integration

MODELING OF BUNDLE WITH RADIATED LOSSES FOR BCI TESTING

Design of Cascode-Based Transconductance Amplifiers with Low-Gain PVT Variability and Gain Enhancement Using a Body-Biasing Technique

On the role of the N-N+ junction doping profile of a PIN diode on its turn-off transient behavior

Gate and Substrate Currents in Deep Submicron MOSFETs

A new inductorless DC-DC piezoelectric flyback converter

Computational models of an inductive power transfer system for electric vehicle battery charge

RFID-BASED Prepaid Power Meter

High efficiency low power rectifier design using zero bias schottky diodes

Mitigation of Common mode Noise for PFC Boost Converter by Balancing Technique

A STUDY ON THE RELATION BETWEEN LEAKAGE CURRENT AND SPECIFIC CREEPAGE DISTANCE

Sub-Threshold Startup Charge Pump using Depletion MOSFET for a low-voltage Harvesting Application

Series connection of IGBT

Susceptibility Analysis of an Operational Amplifier Using On-Chip Measurement

A 2.4GHz to 6GHz Active Balun in GaN Technology

L-band compact printed quadrifilar helix antenna with Iso-Flux radiating pattern for stratospheric balloons telemetry

A notched dielectric resonator antenna unit-cell for 60GHz passive repeater with endfire radiation

Electrical model of an NMOS body biased structure in triple-well technology under photoelectric laser stimulation

Power Loss Estimation in SiC Power BJTs

Robustness of SiC MOSFETs in short-circuit mode

Small Array Design Using Parasitic Superdirective Antennas

Optical component modelling and circuit simulation

QPI-AN1 GENERAL APPLICATION NOTE QPI FAMILY BUS SUPPLY QPI CONVERTER

Improving conducted EMI forecasting with accurate layout modeling

A Novel Piezoelectric Microtransformer for Autonmous Sensors Applications

A design methodology for electrically small superdirective antenna arrays

analysis of noise origin in ultra stable resonators: Preliminary Results on Measurement bench

BANDWIDTH WIDENING TECHNIQUES FOR DIRECTIVE ANTENNAS BASED ON PARTIALLY REFLECTING SURFACES

DUAL-BAND PRINTED DIPOLE ANTENNA ARRAY FOR AN EMERGENCY RESCUE SYSTEM BASED ON CELLULAR-PHONE LOCALIZATION

Neel Effect Toroidal Current Sensor

A Comparison of Phase-Shift Self- Oscillating and Carrier-based PWM Modulation for Embedded Audio Amplifiers

Concepts for teaching optoelectronic circuits and systems

Frequency Domain Prediction of Conducted EMI in Power Converters with. front-end Three-phase Diode-bridge

A Low-Profile Cavity-Backed Dual-Polarized Spiral Antenna Array

Electronic sensor for ph measurements in nanoliters

Enhancement of Directivity of an OAM Antenna by Using Fabry-Perot Cavity

Design of EMI Filters for DC-DC converter

Enhanced spectral compression in nonlinear optical

Prediction of Aging Impact on Electromagnetic Susceptibility of an Operational Amplifier

Conducted EMI Issues in a 600-W Single-Phase Boost PFC Design

On the De-embedding of Small Value Millimeter-wave CMOS Inductor Measurements

PARASITIC CAPACITANCE CANCELLATION OF INTE- GRATED CM FILTER USING BI-DIRECTIONAL COU- PLING GROUND TECHNIQUE

A Switched-Capacitor Band-Pass Biquad Filter Using a Simple Quasi-unity Gain Amplifier

A Low-cost Through Via Interconnection for ISM WLP

Development and Performance Test for a New Type of Portable Soil EC Detector

A technology shift for a fireworks controller

STUDY OF RECONFIGURABLE MOSTLY DIGITAL RADIO FOR MANET

On the Use of Vector Fitting and State-Space Modeling to Maximize the DC Power Collected by a Wireless Power Transfer System

Design of an Efficient Rectifier Circuit for RF Energy Harvesting System

Reconfigurable antennas radiations using plasma Faraday cage

Development of an On-Chip Sensor for Substrate Coupling Study in Smart Power Mixed ICs

Towards Decentralized Computer Programming Shops and its place in Entrepreneurship Development

Study on a welfare robotic-type exoskeleton system for aged people s transportation.

Better understanding EMI generation of power converters

Ironless Loudspeakers with Ferrofluid Seals

Antenna Ultra Wideband Enhancement by Non-Uniform Matching

Influence of ground reflections and loudspeaker directivity on measurements of in-situ sound absorption

Assessment of Switch Mode Current Sources for Current Fed LED Drivers

New Structure for a Six-Port Reflectometer in Monolithic Microwave Integrated-Circuit Technology

COTS-Based Modules for Far-Field Radio Frequency Energy Harvesting at 900MHz and 2.4GHz

An improved topology for reconfigurable CPSS-based reflectarray cell,

PIEZOELECTRIC TRANSFORMER FOR INTEGRATED MOSFET AND IGBT GATE DRIVER

LISN UP Application Note

Robust Optimization-Based High Frequency Gm-C Filter Design

SUBJECTIVE QUALITY OF SVC-CODED VIDEOS WITH DIFFERENT ERROR-PATTERNS CONCEALED USING SPATIAL SCALABILITY

A Modified Single Phase Inverter Topology with Active Common Mode Voltage Cancellation

3D MIMO Scheme for Broadcasting Future Digital TV in Single Frequency Networks

NOVEL BICONICAL ANTENNA CONFIGURATION WITH DIRECTIVE RADIATION

Low temperature CMOS-compatible JFET s

Research Paper ELECTROMAGNETIC INTERFERENCE REDUCTION IN CUK CONVERTER USING MODIFIED PWM TECHNIQUES

Application of CPLD in Pulse Power for EDM

A Passive Mixer for 60 GHz Applications in CMOS 65nm Technology

CHAPTER 2 EQUIVALENT CIRCUIT MODELING OF CONDUCTED EMI BASED ON NOISE SOURCES AND IMPEDANCES

Gis-Based Monitoring Systems.

S-Parameter Measurements of High-Temperature Superconducting and Normal Conducting Microwave Circuits at Cryogenic Temperatures

A simple LCD response time measurement based on a CCD line camera

CHAPTER 1 INTRODUCTION

Dynamic Platform for Virtual Reality Applications

A 180 tunable analog phase shifter based on a single all-pass unit cell

Design and Simulation of Low Cost Pure Sin Wave Inverter Through Multivibrator Technique by Using Multisim

Compound quantitative ultrasonic tomography of long bones using wavelets analysis

Reconfigurable Patch Antenna Radiations Using Plasma Faraday Shield Effect

Analysis of the Frequency Locking Region of Coupled Oscillators Applied to 1-D Antenna Arrays

100 µw Coreless Flyback Converter for microbial fuel cells energy harvesting

A synchronized self oscillating Class-D amplifier for mobile application

High linear low noise amplifier based on self- biasing multiple gated transistors

WIRELESS CHIPLESS PASSIVE MICROFLUIDIC TEMPERATURE SENSOR

Modeling of Conduction EMI Noise and Technology for Noise Reduction

Direct optical measurement of the RF electrical field for MRI

Application Note AN- 1094

Process Window OPC Verification: Dry versus Immersion Lithography for the 65 nm node

ENERGY CABLE MODELING UNDER POWER ELECTRONIC CONVERTER CONSTRAINTS

Transcription:

Complementary MOS structures for common mode EMI reduction Hung Tran Manh, Jean-Christophe Crébier To cite this version: Hung Tran Manh, Jean-Christophe Crébier. Complementary MOS structures for common mode EMI reduction. EPE 09, Sep 2009, Barcelone, Spain. 2009. <hal-00422524> HAL Id: hal-00422524 https://hal.archives-ouvertes.fr/hal-00422524 Submitted on 7 Oct 2009 HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Complementary MOS structures for common mode EMI reduction Manh Hung TRAN, Jean-Christophe CREBIER Grenoble Institute of Technology, Grenoble Electrical Engineering Lab (G2ELab) BP 46-38402 Saint-Martin-d'Hères Cedex, FRANCE Tel: +33 / (0) - 4 76 82 64 77 Fax: +33/ (0) - 4 76 82 63 00 E-Mail: tranm@g2elab.grenoble-inp.fr, crebier@g2elab.grenoble-inp.fr Keywords EMC/EMI, efficiency, MOSFET Abstract The paper deals with the management and the reduction of conducted common mode EMI noise in power converters. Especially, it is presented how complementary P-N MOSFET structures, used in a specific manner, are able of great and natural common mode current reductions. The advantages obtained by the use of complementary MOSFET topologies are balanced with the additional conduction losses that P MOSFETs are responsible of. The paper presents these issues based on simulation and practical results. I. Introduction In power electronic, the common mode conducted noise in switched mode power supplies may cause radiated noise emissions and unexpected disturbances to the other equipments [1][2][3]. Therefore, it is important to reduce this noise and standard have been established to limit radiated and conducted EMI [4]. For that purpose, many works have been carried out to develop power converters that meet standard regulation limits. Indeed, the passive filters [5][6], active filtering [7][8], dynamic reduction of noise sources and symetrization of conducted current s propagation paths [9][10] have been studied. The common feature of these works is to add active or passive components in the conventional structure in order to block or to recycle the propagation of conducted EMI. This is usually done by increasing the amount of components and the global cost of the converter. This paper proposes to investigate the EMI signature of complementary topologies based on the utilization of N and P types MOSFETs. In some cases, it will be shown that complementary structures, by a full symetrization of the, are able of great and natural reductions of the conducted common mode disturbances. As an example, a half-bridge complementary inverter is implemented in this paper and its EMI operation is compared with the one produced by a comparable converter based on N type transistors. The mechanism of common mode noise reduction is explained. Since the introduction of the P type MOSFET in the converter will introduce additional ON state losses, the performances of the conventional and complementary half bridge topologies are also studied and compared thanks to simulation and experimental results. II. Introduction to the concept of complementary inverter circuits In power electronics inverter topologies, the main propagation paths of common mode current harmonics are through the parasitic connected between the middle point of each inverter leg and the frame ground. This parasitic is usually formed by the copper plate on which is soldered the backside of the power dies in isolated module or even the case of the chip (connected to the backside electrode of the power die) in the non isolated packages (ex. TO220 or TO247). In most cases, the heat sink being connected to ground, a large parasitic is created between each power die backside and the ground, offering a common mode current propagation path. Then, as a function of the voltage variations appearing on these parasitic s, large common mode EPE 2009 - Barcelona ISBN: 9789075815009 P.1

disturbances may be generated. Fig. 1 gives an illustration of the parasitic s physical locations whereas these s are added to the schematic on Fig. 2. Bonding wire Power devices die connection Gate connection connection Copper circuit layer Dielectric Layer Cb1 Cb2 Parasitic Capacitors Heat sink Aluminium Fig.1: Main common mode parasitic s in a conventional power electronic structure Cb1 Bipolar gate driver Rg high high-side Bipolar gate driver Rg low low-side Common mode current Cb2 Fig.2: Conventional half-bridge inverter - schematic and electrical location of common mode parasitic s. The conventional inverter structure is depicted on Fig. 2. It is a half-bridge where the main parasitic s relating to the drain of the two MOSFET transistors are Cb1 and Cb2, of which the Cb2 is connected to middle point to the inverter leg. In switched mode power supplies, the middle point whose is floating exhibits large voltage variations with respect to time (dv/dt) each switching transition. Therefore, a large current pulse flows through the parasitic capacitance Cb2. Consequently, this leads the large common mode conducted noise current. On the other hand, the parasitic s connected to the DC bus of the converter are a lot less excited by large dv/dt. In this context, we propose to investigate how a complementary half-bridge inverter, as shown in Fig. 3, could be able to reduce conducted common mode disturbances. In complementary topologies, each leg is made with a series association of a N type and a P type MOSFET transistors. This which is interesting for EMI reduction is not the classical inverting CMOS structure with the P type transistor as the "high side" transistor and the N type transistor as the "low side" transistor. It is the with the complementary transistors associated in series but having their sources connected at the same as it is presented in Fig. 3. This structure is fully symmetrical and presents the main parasitic s (Cb1 & Cb2 located at the back side of the power dies) located mainly on the DC bus whose s are more stable. Therefore, in this configuration, the common mode conducted current levels are decreased significantly. The confirmation of common mode conducted reduction will be presented in the next section. EPE 2009 - Barcelona ISBN: 9789075815009 P.2

Cb1 Rg Bipolar gate driver Common source MOS P Cb2 Fig.3: Complementary half-bridge inverter III. Simulation and experimentation results A. Case study In order to analyze the common mode conducted noise evolution as a function of the, two half-bridge inverter structures were simulated thank to SIMPLORER and were implemented on metal isolated substrates (Fig.4). The conventional being made for reference is based on N type MOSFET devices and two bipolar isolated gate drivers. The complementary structure was built thanks to a pair of N and P type MOSFET transistors and only one isolated bipolar gate driver. This unique driver is able of driving both transistor gate electrode with the same control signal since the power devices are complementary N and P type transistors. Moreover, the two inverters were also designed to operate in a similar manner and to exhibit identical functionalities. Especially, in order to keep EMI sources and switching losses comparable in both cases, the switching transition dynamics were set to be alike. The parameters of the studied topologies are listed in the Table I. In a first approximation, the parasitic s can be deduced by: Sb Cb = ε. (1) e where ε, e, S b are respectively the dielectric permittivity, the dielectric thickness and the surface of the copper layer related to the backside of the transistor s package. Conventional Complementary Input voltage 150 V Power devices MOSFET N: IRFP240 Package TO247AC MOSFET N: IRFP240 Package TO247AC MOSFET P:IRFP9240 Package TO247AC Gate resistances Rg high = 47 ohm Rg low =47 ohm Rg = 22 ohm Input Capacitances 2 470µF 250V Switching frequency 36 khz Table I: Global characteristics of the two inverters topologies EPE 2009 - Barcelona ISBN: 9789075815009 P.3

(a) (b) Fig.4: Two half-bridge inverters implemented. (a) Conventional. (b) Complementary B. Simulation and experimentation results LISN L1 LISN L1 C1 C1 L1 C2 50Ω Z1 Z2 50Ω C2 high-side low-side Cb1 Cb2 C1 C1 L1 C2 50Ω Z1 Z2 50Ω C2 MOS P Common Cb1 Electrode Icom Electrode Icom Ground plane Icom Cb2 Icom N - MOSFET Complementary Fig.5: Half bridge inverter showing the common mode propagation paths The topologies that were simulated are depicted on Fig.5. Besides the main parasitic s approached in previous section, the parasitic of the source electrodes which is formed by the source tracks on metal isolated substrate (see Fig. 1) was also estimated and considered. This capacitance can be calculated as: S s Cs = ε. (2) e where S s is the surface of the track. Thus, the proportion between backside capacitance C b and electrode source capacitance depends on ratio of backside s surface and source tracks surface. C b S = b (3) Cs S s The common mode current can be defined by: dv I = m com C. (4) dt A special attention has been paid to the evolution of the middle point voltage V m. Since it is the main common mode source in the half bridge converter considered in this work, the converters were set such that in simulation and in practice, this middle point voltage exhibits comparable dv/dt. Then, the EMI sources are kept comparable for two topologies (Fig.6), making possible the comparison between the two converter topologies. EPE 2009 - Barcelona ISBN: 9789075815009 P.4

Complementary Complementary N-MOSFET N-MOSFET Fig.6: Switching voltage waveforms of the middle point of the inverter leg with respect to ground for both topologies The frequency spectrums of the common mode currents obtained for both topologies and simulated with Simplorer are shown on Fig.7 (a). It is observed that the common mode current generated by the complementary structure is more than 25 dbµa lower than in the conventional over a large frequency spectrum. This is due to the location of the main parasitic s in the complementary. It leads to a considerable reduction of the parasitic located at the middle point of the inverter leg in the complementary structure. Since this is mainly dependent on the size and surface of the copper track necessary to connect together the sources of the transistors and the middle point to the load, its value can be reduced. Ideally, this copper track having a limited or even negligible thermal contribution, it can even be removed from the SMI substrate, leading to a greatly reduced parasitic. Of course, if this is possible with complementary transistors, this could not be made with the use of only N type transistors. I,dBuA 20 8 4 16 28 40 52 64 76 88 1. 10 4 1. 10 5 1. 10 6 1. 10 7 f,hz (a) N-MOSFET I,dB ua 20 8 4 16 28 40 52 64 76 88 1.10 4 1.10 5 1.10 6 1.10 7 f,hz Complementary Fig.7: Frequency spectrums of the common mode current levels for both topologies (measured at ground LISN terminal [4][11][12]). (a) Simulation results. (b) Experimental results (b) A similar investigation was carried out in experiments where the source electrode of the inverter was not connected on the SMI substrate but directly from the pins of the chips. The experimentation results are given on Fig.7 (b). We can see that the simulation and experiment results are comparable (Fig.8), leading to a parasitic ratio between, drain and source in the range of 20 (as estimated in theory). EPE 2009 - Barcelona ISBN: 9789075815009 P.5

0 8 16 24 I,dBuA 32 40 48 56 64 72 80 1. 10 4 1. 10 5 1. 10 6 1. 10 7 f,hz Experiment Simulation SIMPLORER Simulation SIMPLORER Fig.8. Difference of common mode current s frequency spectrum (for the conventional structure) between SIMPLORER and experiment results. In order to show even more clearly the advantages offered by the complementary structure with respect to common mode EMI generation, standard measurements across LISN resistors [4][11][12] were carried out and the passive filters were designed and implemented to ensure that the two structures meet the EN55011 standard Class A. These topologies were set identical to ease the comparisons and it is showed on Fig. 9. L I S N Cx1 L1 7 mh 0.68mF 0.68mF Cx2 Cy1 7 mh L2 Filter Cy2 Electrode Icom Ground plane Fig.9. Conducted EMI noise filter For that purpose, the same differential mode current filters (L1=L2=0,7mH, Cx1=Cx2=0,68mF) were built for two topologies. Indeed, complementary topologies are not acting on differential mode disturbance levels. On the other hand, each converter requires a different common mode filter such as C y1 =C y2 =133nF for the conventional inverter and C y1 =C y2 =10nF for the complementary. Therefore, we can see that a filtering ratio of 13,3 is required between the two common mode filters in order to comply with the regulation standards. The frequency spectrums of both topologies obtained in practice with and without EMI filters are presented on Fig.10. EPE 2009 - Barcelona ISBN: 9789075815009 P.6

90 dbµv 80 70 60 50 40 N-MOSFET Complementary Class A (EN55011) 30 20 10 1.10 4 1.10 5 1.10 6 1.10 7 1.10 8 f, Hz Fig.10: Conducted EMI frequency spectrum measurements with filter for both topologies (the voltage across the impedance Z1 of LISN) with and without EMI conducted filters Since it is not realistic or even possible to implement common mode EMI filter with large Cy s, the common mode EMI filter needed for the conventional would require normally the use of a common mode choke. This would result in an additional cost and additional losses. These results demonstrate the interest in considering complementary topologies as far as common mode EMI reductions are concerned. However, as it is well known, the introduction of P type transistors into a power electronic converter leads usually to the increase of the ON state losses and the reduction of the efficiency levels. Let's investigate this point in the next section of this paper. Please note on Fig. 10 that common mode disturbances are simple to separate to differential mode disturbances. Indeed, in a half bridge inverter, common mode disturbances appear at each switching frequency multiple whereas differential mode noises appear at each multiple of twice the switching frequency. Therefore, it is simple to identify on the frequency spectrum the contribution of the common mode currents added to the differential mode currents on the global disturbance levels. IV. Efficiency analysis The main drawback of this approach corresponds to the extra losses introduced by the implementation of P type transistors with increased ON state voltage drops. This issue is investigated in details in practice in order to evaluate what is the price to pay as a function of the design and the operating conditions of the two types of inverters. Table II below summarizes a set of experimental efficiency measurements for the conventional and the complementary structure. Fig.12 are plotted the levels measured and reported in Table II. It is well known that in a converter, the ratio conduction losses over power flow increases linearly whereas the ratio switching losses over power flow is about stable as a function of power flow. Therefore, it appears that, as expected, when the conduction losses increase due to a larger power flow, the complementary structure is less efficient. As it could be expected, the complementary induces larger conduction losses. When the switching losses are in the range of the conduction losses, the decrease in efficiency is reasonable. However, for larger load levels, the losses generated by the complementary converter are twice larger than in a conventional converter. In Table II, another test is presented where the complementary converter was built with a P type transistor twice as large as before (Fig.11). In this case, the conduction losses are reduced and are closer to the one of the conventional. EPE 2009 - Barcelona ISBN: 9789075815009 P.7

Bipolar gate driver Rg Common source MOS P Fig.11: Complementary implemented with two parallel P-type MOSFETs Input Power, W (at Vin=150V) 50 200 300 400 N-MOSFET 98.85 98.06 96.26 95.07 93.68 Efficiency % Complementary (1 P-MOSFET) 98.04 96.14 92.82 89.37 85.66 Complementary (2 parallel P- MOSFET) 98.45 97.10 94.21 92.49 90.65 Table II: Efficiency table as a function of the, the input power level and the number of P-MOS devices Efficiency,% 98.4 96.8 N-MOSFET Complementary 2 parallel MOSFET P Complementary 1 MOSFET P 95.2 v1 v2 v3 93.6 92 90.4 84 88.8 87.2 85.6 84 0 200 300 400 0 Power, Puissance W 45 Fig.12: Plot of the evolution of the efficiency as a function of the, the input power level and the number of P-MOS Moreover, the structure with two P type MOSFET associated in parallel does not degrade the EMI advantages of complementary structure highlighted in the previous sections because two backside () electrodes parasitic s are connected in the DC bus whose is relatively stable. This can be seen on the Fig.13 where a new frequency spectrum of the conducted EMI is plotted when two P type MOSFET are used for the implementation of the half bridge inverter leg. It is shown that EPE 2009 - Barcelona ISBN: 9789075815009 P.8

the complementary including two parallel P type MOSFETs meet the EMI standard with the same passive filter as for the converter based on one P type MOSFET. 90 80 dbuv 70 60 50 40 30 20 10 1.10 4 1.10 5 1.10 6 f,hz 1.10 7 1.10 8 Complementary 1 MOSFET P Complementary 2 parallel MOSFET P Class A (EN55011) Fig.13: Conducted EMI frequency spectrum measurements with conducted EMI filter for both complementary topologies (the voltage across the impedance Z1 of LISN) Therefore, great reductions on conducted EMI levels are obtained enlarging the size of a power device. In some applications, for low and medium power, replacing passive filtering elements with extra active silicon devices may lead to cost effective solutions and improved integration levels, especially if efficiency is maintained high enough by employing larger active devices. In some high voltage high frequency light load application, the work presented in this paper can offer an interesting answer for the reduction of the common mode current management and reduction. V. Conclusion To conclude on this work, it is shown that common mode conducted EMI could be treated as a function of the of the converter and the type of device used for the implementation. Based on simulation and experimental results, the work detailed in this paper demonstrates that valuable reductions of common mode current levels and propagation paths could be achieved using complementary devices. As a function of the layout design of the converter, great reductions in common mode filtering needs can be obtained. However, this structural advantage must be balanced by the additional conduction losses introduced by the P-MOS transistor implementation. It is also exhibited that the additional conduction losses generated by P-MOS transistor could be limited using a larger power device. This is obtained without altering the improvements obtained on the EMI behavior of the complementary structure. This work shows that possible modifications on active device size and type enable to reduce filtering requirement for standard compliance. This means that in an effort of converter integration, more active devices may be used to limit the sizes of passive devices which are harder to integrate. It is then a question of compromise as a function of efficiency, integration, cost and global performances. In some applications, especially but not only under reduced power levels, the converters implemented thanks to complementary power devices may be preferred. It is now a question of power device optimization. Indeed, it is simple to find any type of N type device at any voltage and power level. It can not be said the same thing for P type power devices. Especially, as far as IGBT are concerned, it could be of interest to develop P type power devices to see if the compromise between the switching and the conduction losses could be set differently. Finally, the global design and optimization of the converter could be based on complementary power devices. EPE 2009 - Barcelona ISBN: 9789075815009 P.9

References [1] C. Paul, A comparison of the contributions of common-mode and differential-mode currents in radiated emissions, IEEE Transactions on Electromagnetic Compatibility, vol. 31, 1989, pp. 189-193. [2] R. Kaires, The correlation between common mode currents and radiated emissions, IEEE International Symposium on Electromagnetic Compatibility, 2000, pp. 141-146 vol.1. [3] Jianqing Wang, O. Fujiwara, et K. Sasabe, A simple method for predicting common-mode radiation from a cable attached to a conducting enclosure, Microwave Conference. APMC 2001.Asia-Pacific, pp. 1119-1122 vol.3. [4] CISPR16-1-2, Specification for radio disturbance and immunity measuring apparatus and methods, Aug. 2006. [5] S. Shahparnia et O. Ramahi, Electromagnetic interference (EMI) reduction from printed circuit boards (PCB) using electromagnetic bandgap structures, IEEE Transactions on Electromagnetic Compatibility, vol. 46, 2004, pp. 580-587. [6] B. Pierquet, T. Neugebauer, et D. Perreault, Inductance Compensation of Multiple Capacitors with Application to Common- and Differential-Mode Filters, Power Electronics Specialists Conference. PESC '06. 37th IEEE, 2006, pp. 1-10. [7] T. Kato, K. Inoue, et K. Akimasa, EMI Reduction Method for a Single-Phase PWM Inverter by Suppressing Common-Mode Currents with Complementary Switching, Power Electronics and Motion Control Conference. IPEMC '06. CES/IEEE 5th International, 2006, pp. 1-5. [8] N. Mortensen et G. Venkataramanan, An Active Common Mode EMI Filter for Switching Converters, Industry Applications Society Annual Meeting. IAS '08. IEEE, 2008, pp. 1-7. [9] Shuo Wang, Pengju Kong, et F. Lee, Common Mode Noise Reduction for Boost Converters Using General Balance Technique, IEEE Transactions on Power Electronics, vol. 22, 2007, pp. 1410-1416. [10] M. Shoyama, Ge Li, et T. Ninomiya, Balanced switching converter to reduce common-mode conducted noise, IEEE Transactions on Industrial Electronics, vol. 50, 2003, pp. 1095-1099. [11] C.R. Paul, "Introduction to Electromagnetic Compatibility", John Wiley & Sons, Inc, 1992. [12] Guy-Gérard Champiot, Compatibilité électromagnétique: Normalisation, réglementation et mesure Techniques de l'ingénieur, May 2000. EPE 2009 - Barcelona ISBN: 9789075815009 P.10