Conducted EMI Prediction of the PFC Converter ... - IEEE Xplore

3 downloads 0 Views 1MB Size Report
Abstract—A conducted EMI prediction model for an interleaved power factor correction (PFC) converter including the nonlinear- ity of the boost inductors is ...
IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 55, NO. 6, DECEMBER 2013

1107

Conducted EMI Prediction of the PFC Converter Including Nonlinear Behavior of Boost Inductor Yitao Liu, Student Member, IEEE, Kye Yak See, Senior Member, IEEE, and King-Jet Tseng, Senior Member, IEEE

Abstract—A conducted EMI prediction model for an interleaved power factor correction (PFC) converter including the nonlinearity of the boost inductors is presented. The Jiles–Atherton (J–A) technique is adopted to handle the nonlinear behavior of the interleaved boost inductors. The conventional conducted EMI prediction model without including the nonlinear effect of the boost inductors tends to underestimate the true conducted EMI level. With the proposed model, conducted EMI can be predicted with higher accuracy so that the correct EMI filter can be designed to meet the required EMI regulatory limits.

This paper is organized into six sections. Section II describes the mathematical model to handle the nonlinear hysteresis behavior of an inductor. Section III illustrates the configuration of the interleaved PFC converter system and its conducted EMI prediction model. In section IV, conducted EMI prediction results are presented. Section V demonstrates the accuracy of the proposed predication model with simulation and experimental results. Finally, section VI concludes the paper.

Index Terms—EMI filter, EMI/EMC, Jiles–Atherton (J—A) hysteresis model, PFC converter.

II. NONLINEAR INDUCTOR MODELING A. Nonlinear Inductor Model

I. INTRODUCTION FC converters are gaining popularity due to low-frequency power quality regulatory requirement. Besides power quality issue, PFC converters also require compliance with highfrequency conducted emission standards, for the switching frequency of the converter in this paper is 116.5 kHz, the noise limit will be chosen from 100 kHz to 30 MHz according to the EMI standards [1] description. Therefore, a built-in EMI filter is necessary to ensure high-frequency conducted emission compliance [2], [3]. To choose the correct EMI filter during the PFC converter design stage, an accurate conducted EMI prediction model is needed. This paper proposes a conducted EMI prediction model that includes the nonlinear hysteresis behavior of the boost inductors. Conventionally, for ease of modeling, the boost inductor is assumed to be linear and therefore its inductance remains constant. In reality, the inductance of any inductor with magnetic core varies with the current passing through it. There are several methods to handle inductance variation of an inductor with changing current flow in power electronic circuits. The two most established techniques to model magnetic cores are Preisach’s model [4] and Jiles–Atherton’s (J–A) models [5], [6]. As a timedomain and history-dependent hysteresis model, the J–A model has been proven to predict the magnetic hysteresis behavior in the power electronic circuits accurately. Hence, the J–A model is adopted in this paper to account for the nonlinear effect of boost inductor for accurate conducted emission prediction.

P

Manuscript received September 24, 2012; revised February 2, 2013; accepted March 12, 2013. Date of publication April 8, 2013; date of current version December 10, 2013. The authors are with the School of Electrical and Electronic Engineering, Nanyang Technological University, Nanyang 639798, Singapore (e-mail: [email protected]; [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TEMC.2013.2254120

In the PFC converter, the input current variation is mainly dependent on the input voltage, output power, and boost inductors [7]. To include the nonlinear effect of the boost inductor in the PFC converter, the J–A model can be adopted to describe the hysteresis behavior of the magnetic core, which is a part of the boost inductor. The J–A model consists of a group of first-order ordinary differential equations which could be solved numerically to provide the relationship between the magnetic core’s magnetization and magnetic field. The J–A model can be described by Man (He ) − Mirr dMirr = dH k · sig (dH/dt) − α [Man (He ) − Mirr ] dMirr Man (He ) − Mirr dH = · dt k · sig (dH/dt) − α [Man (He ) − Mirr ] dt   dMan dMrev dMirr =c − dH dH dH M = Mirr + Mrev

(1) (2) (3) (4)

where Mirr is the irreversible magnetization, Man is the anhysteretic magnetization, k and c are the wall motion parameters, the function sig(·) is defined as sig(x) = 1 for x ≥ 0 and sig(x) = −1 for x < 0. M is the total magnetization which is the summation of Man and Mirr [8]. He = H + αM B = μ0 (H + M )

(5) (6)

where He is the effective magnetic field, B is the flux density, H is the magnetic field in the core, α is an interdomain coupling factor, and μ0 is the permeability of free space. The anhysteretic magnetization Man is described using Langevin function [9]   ∂ He − Man = Ms coth (7) ∂ He

0018-9375 © 2013 IEEE

1108

IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 55, NO. 6, DECEMBER 2013

TABLE I ADJUSTABLE J–A MODEL CONSTANT

Fig. 1.

Calculated magnetization curve against magnetic field.



Ms dMan = 1 − coth2 dH ∂



He ∂



 +

∂ He

Fig. 2. Equivalent circuit model of the boost inductor with parasitic components.

2  (8)

where Ms is the saturate magnetic moment of the core material, and ∂ is a shape parameter. Using equations (1)–(8), one can easily derive the following equation: δ · (Man − M ) dM = (1 − c) · · dH sig(H ) · k · (1 − c) − α(Man − M ) +c· where

 δ=

dMan dH . if sig H × (Man − M ) ≤ 0

1,

otherwise

.

(10)

Fig. 1 shows the total and anhysteresis magnetization against magnetic field according to aforementioned mathematical derivation. It clearly indicates that the magnetic hysteresis behavior can be described with the J–A model. The hysteresis parameters Ms , k, ∂, α, and c can be obtained from the material specifications and solved by the following equations. The wall motion parameter k can be estimated as Hc , Hc is the coercive force (A/m). From (7), it can be derived that   ∂ αMr − Man (Mr ) = Ms coth (11) ∂ αMr f (α) = Man (Mr ) − Mr +

α 1−c

(14)

From (7) and (14), one can have (1 − c)kXm αXm + 1

(15)

where Mm is the maximum value of magnetization at tip of B– H loop, Xm is the maximum differential susceptibility. Using the secant method as introduced in (13) to solve for (15), shape parameter ∂ can be obtained. Wall motion parameter c can be calculated as follows: c = 3∂Xin /Ms

(16)

where Xin is the initial differential susceptibility. The J–A hysteresis model with these derived parameters can be used to simulate the boost inductor nonlinear behavior. The hysteresis parameters for the magnetic core material which is CTX16–17309 are given in Table I. B. Parasitic Components of the Boost Inductor

Based on the measured frequency response of the boost inductor impedance using a HP 4194A impedance analyzer, the parasitic components of the inductor are extracted [11]. Fig. 2 shows the equivalent circuit model of the boost inductor with its equivalent parallel resistance (EPR) and equivalent parallel capacitance (EPC), it will be applied in the circuit simulation k  of Section IV. Fig. 3 shows the frequency response of the boost  n (M r ) + 1/ Xr − c dM adH inductor from 100 kHz to 30 MHz. (12)

where Mr = Br /μ0 , remanence flux density Br and saturation magnetization Ms can be obtained from the magnetic material specification. Xr is the remanence differential susceptibility. The secant method [10] introduced in (13) can be used to calculate the parameter α. αn = αn −1 −

He = Hm + αMm .

g(∂) = Man (He ) − Mm − (9)

0,

n −1 If α n −α < tolerable error, then α=αn . αn According to (5)

αn −1 − αn −2 f (αn −1 ). f (αn −1 ) − f (αn −2 )

(13)

III. PFC CONVERTER AND THE CONDUCTED EMI PREDICTION MODEL The interleaved PFC converter topology to be implemented is shown in Fig. 4. The input current of the PFC converter is dependent on the output power of the converter, inductance of the boost inductor and input voltage of the mains. Although a cycle by cycle method can be used to determine the PFC converter’s input current waveform for a half period, some form

LIU et al.: CONDUCTED EMI PREDICTION OF THE PFC CONVERTER INCLUDING NONLINEAR BEHAVIOR OF BOOST INDUCTOR

Fig. 3.

Fig. 4.

1109

Fig. 5.

Switching signals of the interleaving topology.

Fig. 6.

High-frequency noise current path in the interleaved PFC converter.

Fig. 7.

Schematic of LISN (Model 3725/2M).

Impedance frequency response of the boost inductor.

PFC converter topology with an average current control method.

of modulation method is employed in the actual circuit and the current passing through the boost inductor depends on the modulation function. Therefore, a simulation model including the controller should be included in the input current calculation. The lower part of Fig. 4 shows the implementation of the PFC controller. There are two control loops in the controller, a slower response loop that controls average output dc voltage and a much faster response loop that shapes the input current. The two average boost inductor currents are controlled to be proportional to the utility line-voltage waveform, and then the summation of the two interleaved inductor currents is the system input current. The PWM switching waveforms generated by the controllers at the two PFC switches, S1 and S2 , are shown in Fig. 5. The boost inductors L1 and L2 are interleaved in the circuit, and the control signal of the switch S1 is delayed by half switching period than another switch S2 . This topology

produces smaller current ripple and EMI noise but at the expense of additional auxiliary circuits and power devices [12]. The nonlinearity of the two boost inductors L1 and L2 are modeled with the hysteresis behavior described in details in the next section. In the power electronic circuits, high-frequency switching of the power semiconductor is the major cause of conducted EMI [13], [14]. The nonidealities of the power transistor switch (IR840), drain-to-source turn-on resistance (Rd = 0.8 Ω), equivalent series inductances of the drain and source terminals (Ld = 3.5 nH and Ls = 7.5 nH), and junction capacitances (Cg s = 185 pF,Cg d = 185 pF, and Cds = 200 pF) are obtained from the power transistor datasheet [15] and included in the simulation model for more realistic prediction of isum , the summation of the two boost inductor currents. Similarly, the equivalent model of the boost inductor with the parasitic components, as shown in Fig. 2, is also included in the simulation model as shown in Fig. 4. As the PFC converter to be designed is a two-wire system (line and neutral wires without earth wire), differential-mode (DM) conducted EMI will be dominated. Fig. 6 shows the highfrequency (DM) noise current path, which superimposes on the power frequency input current of the PFC converter. As the actual test setup uses the EMCO 3721/2M line impedance stabilization network (LISN) for conducted emission measurement, the circuit model of the LISN, as shown in Fig. 7, must be included in the conducted emission prediction

1110

IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 55, NO. 6, DECEMBER 2013

values of the EMI filter components can be chosen based upon the corner-frequency method [17]. According to Amp`ere’s circuital law πDH = ni, one has the relationship between current and magnetic field H=

ni . πD

(17)

Then,

Fig. 8.

DM conducted emission simulation model. TABLE II PFC CONVERTER SPECICATIONS AND PARAMETERS

μ0 n2 A di dM d · · 1+ VL = μ0 nA (H + M ) = . (18) dt πD dt dH Integrating both side of (18) results in t VL dt − H(t) + M0 + H0 M (t) = μ 0 nA t0

(19)

where M0 is the initial magnetization, t0 is the start moment of the integration, VL denotes the boost inductor voltage respectively. (19) can be discretized as follows: M (j) =

j

VL (m) × Δt − H(j) + M0 + H0 . μ0 nA m =1

(20)

From (17) and (18) one can have model. In the actual measurement, the DUT side is connected to the interleaved PFC converter, and conducted emissions on the live and neutral lines are measured with channel 1 (CH1) and channel 2 (CH2) of the digital oscilloscope, respectively. Once the noise current isum is obtained from the circuit simulation as shown in Fig. 4, the DM conducted emission can be predicted using the DM equivalent circuit model shown in Fig. 8. The noise current through the LISN can be calculated and conducted emission noise voltage VEM I can be determined readily. Due to the interleaved topology, both the boost inductors are identical. The fundamental components of the two boost inductors currents are out of phase and therefore cancel each other. For the second harmonic, the two currents are in phase and result in noise current through LISN. For the same reason, odd harmonics of the noise current are negligible and only even harmonics of the noise current stay. IV. SIMULATION RESULTS The interleaved PFC converter topology illustrated in Section III was integrated with the nonlinear boost inductor model described in Section II and the complete interleaved PFC converter system is simulated with MATLAB Simulink. The simulation was conducted based of the PFC converter specifications given in Table II. The calculated boost inductor value is 200 μH if its nonlinear behavior is ignored [16]. With the hysteresis behavior of the boost inductors included in accordance with (17)–(23), the time-domain input current waveform of the PFC converter can be determined. Once the input current waveform is obtained, the DM conducted emission can be predicted with the equivalent noise model given in Fig. 8. The frequency-domain spectrum of conducted emission could be derived through FFT. With the known conducted emission spectrum and specific EMI limit to be complied, appropriate

πD VL πD VL di = · = · . dt μ0 n2 A 1 + dM/dH μ0 n2 A 1 + dM/dH (21) Hence, dH 1 VL 1 VL = · = · (22) dt μ0 nA 1 + dM/dH μ0 nA 1 + dM/dH where i is the boost inductor current, and it is also part of the PFC converter input current. Equation (22) can be discretized as 1 VL (j) H(j + 1) − H(j) = · . Δt μ0 nA 1 + (dM/dH)(j)

(23)

According to the introduction for the magnetic inductor voltage and current relationship, then the nonlinear magnetic inductor electrical model can be built in the embedded MATLAB Function of Simulink using following six numerical procedures. 1) Initialize magnetization, current and set the initial value of dM/dH as zero. 2) Calculate the effective magnetic field He , anhysteretic magnetization Man , and dMan /dH using (5), (7), and (8). 3) The boost inductor voltage VL (m) is captured from the overall circuit simulation at each time step, the value of M (j) can then be calculated using (20). 4) Calculate the sign of dH/dt using H(j) − H(j − 1), determine the sign of Man (j) − M (j) and value of δ, and calculate the value of dM/dH using (9). 5) Calculate the H(j + 1) using (23). Update the current i(j + 1) using (17). 6) Repeat step 2 through 5 in the embedded MATLAB function until simulation ends. Fig. 9 shows the simulated hysteresis curve of the boost inductor based on the J–A hysteresis model. The main loop in power frequency and the minor loop in the switching frequency

LIU et al.: CONDUCTED EMI PREDICTION OF THE PFC CONVERTER INCLUDING NONLINEAR BEHAVIOR OF BOOST INDUCTOR

Fig. 9.

Hysteresis behavior of boost inductor.

Fig. 11.

Fig. 10. 110 V.

1111

Boost inductors current and their zoom in waveforms.

PFC converter input voltage, current, and output voltage, V in −rm s =

show that the J–A model could be applied in the magnetic model at both low and high frequency appropriately. Fig. 10 shows the simulated steady-state output voltage and input current waveforms of the PFC converter with the boost inductors’ nonlinear behavior included, which clearly demonstrates that the J–A model governed by a group of complex differential equations works seamlessly with the various circuit blocks of the PFC converter. From the observed waveforms as shown in Fig. 11, it can be found that the ripple currents in the two inductors cancel each other due to the interleaving topology, and results in much smaller ripple in the total input current. Thus, the interleaved topology has an ability to reduce the inductor magnetic core size compared with the traditional single-boost inductor topology. Fig. 12(a) and (b) shows the boost inductor currents without and with J–A model included. Fig. 12(c) shows the zoom in current comparison when the boost inducotr nonlinearity is ignored and considered, it is clearly observed that the current ripple in Fig. 12(b) is smaller than that shown in Fig. 12(a), a clear indication that the J–A model handles the boost inductor nonlinearity very well in the simulation. The accurate prediction of the input current including the inductor’s nonlinearity will lead to more accurate conducted EMI prediction for the proper selection EMI filter, as we will discuss in the following section.

Fig. 12. Boost inductor current, (a) current without J–A model consideration, (b) current with J–A model consideration, and (c) zoom in boost inductor current comparison.

1112

Fig. 13.

IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 55, NO. 6, DECEMBER 2013

Setup of the conducted noise emission measurement.

V. EXPERIMENT RESULTS A. Conducted Emission Comparison To validate the conducted emission results from the prediction model described earlier, a PFC converter is designed in accordance with the specification given in Table II. It is designed with a UCC28258/UCC28220 dual interleaved PFC preregulator and two boost inductors with toroid core CTX16–17309. The core has a diameter D of 2.3 × 10−2 m, a cross-sectional area A of 1 × 10−4 m2 , and the number of turns n is 100. These geometrical parameters of the magnetic core are applied in the J–A model to describe the nonlinear behavior of the boost inductors. To observe both the time-domain and frequency-domain conducted EMI, a time-domain measurement setup (TDMS) using digital signal oscilloscope (DSO) is used for collecting the conducted EMI measurement results [18]. The TDMS is consists of a 3725/2M LISN and a TDS7054 oscilloscope as shown in Fig. 13. The frequency-domain conducted emission can be obtained using the FFT function easily. One channel of the DSO is connected to line L output and another channel of the same DSO is connected to neutral N output of the LISN. By using the adding or subtracting function of the DSO, the common-mode (CM) or DM components of the conducted EMI can be measured separately. A DM EMI filter will be designed and implemented between the input power and interleaved PFC converter to bring the emission level below the limit. Fig. 14(a) shows the measured conducted emission from 100 kHz to 30 MHz. Fig. 14(b) shows the comparison between simulated conducted emission without the J–A model and measured conducted emission. Fig. 14(c) shows the comparison between simulated conducted emission with J–A model and measured conducted emission. For ease of comparison, Table III indicates the measured and simulated conducted emission of the first few harmonics of the switching frequency. It clearly indicates that the simulated emission without the boost inductors nonlinearity consideration underestimates the emission by 4 to 6 dB. The comparisons clearly show that with the J–A model included, the simulated results agree very well with the measured results. Without the J–A model, the nonlinearity of the boost inductor will not be included in the simulation model, which leads to underestimation of conducted emission and incorrect selection of EMI filter to meet the respective EMI limit. However, at higher frequencies, the parasitic capacitance of the boost inductor begins to dominate and the simulated emissions with and without the J–A model do not show much difference.

Fig. 14. Conducted emission: (a) experimental result, (b) comparison with simulation result with J–A model, and (c) comparison with simulation result without J–A model. TABLE III CONDUCTED EMISSION COMPARISON

Fig. 15 shows the CM measured conducted spectrum from 100 kHz to 30 MHz, as it is a two-wire power electronic system, the CM emission is much lower than the DM emission. Thus, only DM EMI filter is needed in this PFC converter. B. EMI Filter Design For the interleaved PFC converter, which operates at a switching frequency fs = 116.5 kHz, the emission is higher at its second harmonic (233 kHz), as explained in section III. With predicted emission level at 233 kHz and the required limit to be

LIU et al.: CONDUCTED EMI PREDICTION OF THE PFC CONVERTER INCLUDING NONLINEAR BEHAVIOR OF BOOST INDUCTOR

Fig. 15.

1113

Conducted common mode (CM) emission.

Fig. 17. Insertion loss of the EMI filter with parasitic components, (a) L D = 85 μHand (b) L D = 210 μH.

Fig. 16.

EMI filter with parasitic components included.

met, the attenuation required from the EMI filter is given by Areq,DM = VDM − Vlim it + 6 dB = 40 log fDM ,cor =

fDM (24) fDM ,cor

1 1 √ √ = . 2π LDM CDM 2π · 2LD 1 CX 1

(25)

Based on (26), the values of the inductor and capacitor are chosen accordingly to meet the required filter attenuation given in (25). The selected inductor and capacitor with their parasitic components, as shown in Fig. 16, are included to ensure more accurate filter attenuation prediction [19]. With the boost inductor nonlinearity ignored, the required filter attenuation of the DM component Areq,DM is 37 dB. The DM filter should provide Areq,DM at 233 kHz, fDM is the frequency of DM emission at 233 kHz. Based on (24), fDM ,cor is 27.7 kHz. Substituting the filter capacitor CX 1 = CX 2 = 200 nF and fDM ,cor into (25), the needed DM choke LD 1 = LD 2 is about 82.5 μH, choose LD 1 = LD 2 = 85 μH for the EMI filter. With the parasitic component consideration, the attenuation is 31.3 dB according to Fig. 17(a), it can be found that the parasitic component could degrade the attenuation of the EMI filter, so the margin of 6 dB is necessary in the design process. The measured suppressed conducted emission is shown in Fig. 18(b) and it shows that the emission at 233 kHz is still exceeding the limit. With the boost inductor nonlinearity is considered, Areq,DM is 45 dB and fDM ,cor is 17.5 kHz, the needed DM choke is 206.8 μH. Choosing LD 1 = LD 2 = 210 μH for the EMI filter. The chosen inductor has an EPR of 4 kΩ and EPC of 5 pF. The

Fig. 18. DM noise spectrum: (a) without EMI filter, (b) with EMI filter (L D 1 = L D 2 = 85 μH), and (c) with EMI filter (L D 1 = L D 2 = 210 μH).

attenuation is 39.2 dB according to Fig. 17(b). The measured suppressed conducted emission is shown in Fig. 18(c), and now the emission at 233 kHz is below the limit.

1114

IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 55, NO. 6, DECEMBER 2013

VI. CONCLUSION PFC converters improve power quality but they still require an EMI filter to comply with the high-frequency conducted EMI limit. A conducted EMI prediction model with the ability to handle nonlinear behavior of the boost inductor is proposed. The nonlinearity of boost inductor is described by adopting the J–A model, which has been shown to emulate the hysteresis behavior of the boost inductor very well. The conducted EMI prediction model with the J–A model included has been validated with measurement results. We have shown that ignoring nonlinearity of the boost inductors leads to underestimation of conducted EMI and, therefore, the EMI filter designed with the predicted emission will be underdesigned and unable to suppress the conducted EMI below the required limit. The higher prediction accuracy of the proposed conducted EMI model allows correct selection of EMI filter to meet the EMI regulatory requirement at the early product development stage without the risk of unexpected underperformance. REFERENCES [1] Limits and methods of measurement of radio disturbance characteristics of electrical lighting and similar equipment (CISPR 15:2005), EN 55015, 2006. [2] K. Raggl, T. Nussbaumer, and J. W. Kolar, “Guideline for a simplified differential-mode EMI filter design,” IEEE Trans. Ind. Electro., vol. 57, pp. 1031–1040, 2010. [3] V. Tarateeraseth, S. Kye Yak, F. G. Canavero, and R. W. Chang, “Systematic electromagnetic interference filter design based on information from in-circuit impedance measurements,” IEEE Trans. Electromag. Compat., vol. 52, no. 3, pp. 588–598, Aug. 2010. [4] Y. Bernard, E. Mendes, and F. Bouillault, “Dynamic hysteresis modeling based on Preisach model,” IEEE Trans. Magn., vol. 38, no. 2, pp. 885–888, Mar. 2002. [5] J. H. B. Deane, “Modeling the dynamics of nonlinear inductor circuits,” IEEE Trans. Magn., vol. 30, no. 5, pp. 2795–2801, Sep. 1994. [6] L. Huiqi, L. Qingfeng, X. Xiao-Bang, L. Tiebing, Z. Junjie, and L. Lin, “A modified method for Jiles-Atherton hysteresis model and its application in numerical simulation of devices involving magnetic materials,” IEEE Trans. Magn., vol. 47, no. 5, pp. 1094–1097, May 2011. [7] K. Raggl, T. Nussbaumer, G. Doerig, J. Biela, and J. W. Kolar, “Comprehensive design and optimization of a high-power-density single-phase boost PFC,” IEEE Trans. Ind. Electro., vol. 56, no. 7, pp. 2574–2587, Jul. 2009. [8] D. Jiles and D. Atherton, “Ferromagnetic hysteresis,” IEEE Trans. Magn., vol. 19, no. 5, pp. 2183–2185, Sep. 1983. [9] L. A. Righi, P. I. Koltermann, N. Sadowski, J. P. A. Bastos, R. Carlson, A. Kost, L. Janicke, and D. Lederer, “Non-linear magnetic field analysis by FEM using Langevin function,” IEEE Trans. Magn., vol. 36, no. 4, pp. 1263–1266, Jul. 2000. [10] P. D´ıez, “A note on the convergence of the secant method for simple and multiple roots,” Appl. Math. Lett., vol. 16, pp. 1211–1215, 2003. [11] C. Henglin and Q. Zhaoming, “Modeling and characterization of parasitic inductive coupling effects on differential-mode EMI performance of a boost converter,” IEEE Trans. Electromag. Compat., vol. 53, no. 4, pp. 1072–1080, Nov. 2011. [12] X. Xiaojun, L. Wei, and A. Q. Huang, “Two-Phase interleaved critical mode PFC boost converter with closed loop interleaving strategy,” IEEE Trans. Power Electro., vol. 24, no. 12, pp. 3003–3013, Dec. 2009. [13] J. Balcells, A. Santolaria, A. Orlandi, D. Gonzalez, and J. Gago, “EMI reduction in switched power converters using frequency Modulation techniques,” IEEE Trans. Electromag. Compat., vol. 47, no. 3, pp. 569–576, Aug. 2005. [14] R. Morrison and D. Power, “The effect of switching frequency modulation on the differential-mode conducted interference of the boost power-factor correction converter,” IEEE Trans. Electromag. Compat., vol. 49, no. 3, pp. 526–536, Aug. 2007. [15] 8 A, 500 V, 0.850 Ohm, N-Channel Power MOSFET. (2002). [Online]. Available: http://www.datasheetcatalog.org/datasheet/fairchild/ IRF840.pdf

[16] T. Nussbaumer, K. Raggl, and J. W. Kolar, “Design guidelines for interleaved single-phase boost PFC circuits,” IEEE Trans. Ind. Electro., vol. 56, no. 7, pp. 2559–2573, Jul. 2009. [17] S. Fu-Yuan, D. Y. Chen, W. Yan-Pei, and C. Yie-Tone, “A procedure for designing EMI filters for AC line applications,” IEEE Trans. Power Electro., vol. 11, no. 1, pp. 170–181, Jan. 1996. [18] M. Kumar and V. Agarwal, “Power line filter design for conducted electromagnetic interference using time-domain measurements,” IEEE Trans. Electromag. Compat., vol. 48, no. 1, pp. 178–186, Feb. 2006. [19] W. Shuo, C. Rengang, J. D. Van Wyk, F. C. Lee, and W. G. Odendaal, “Developing parasitic cancellation technologies to improve EMI filter performance for switching mode power supplies,” IEEE Trans. Electromag. Compat., vol. 47, no. 4, pp. 921–929, Nov. 2005. Yitao Liu (S’11) received the B.Eng. degree in electrical engineering from Wuhan University, Wuhan, China, in 2008 and the M.Sc. degree from Nanyang Technological University, Nanyang, Singapore, in 2010, where he is currently working toward the Ph.D. degree in the School of Electrical and Electronic Engineering.

Kye Yak See (SM’02) received the B.Eng. degree from the National University of Singapore, Singapore, and the Ph.D. degree from Imperial College London, London, U.K., in 1986 and 1997, respectively. Between 1986 and 1991, he was with Singapore Technologies Electronics as the Head of Electromagnetic Compatibility (EMC) Centre. From 1991 to 1994, he held the position of Lead EMC Design Engineer in ASTEC Custom Power, Singapore. He is currently an Associate Professor in the School of Electrical and Electronic Engineering, Nanyang Technological University, Nanyang, Singapore. He also holds concurrent appointments as the Head of the Circuits and Systems Division and the Director of Electromagnetic Effects Research Laboratory (EMERL). His research interests are EMI filter design, signal integrity, and EMC measurement techniques. Dr. See is the Founding Chair of the IEEE Singapore EMC Chapter and a Technical Assessor of Singapore Accreditation Council. He was also the Organizing Committee Chairs for the 2006 EMC Zurich Symposium and 2008 Asia Pacific EMC Conference in Singapore. Since January 2012, he has been the Technical Editor of the IEEE EMC Magazine. King-Jet Tseng (S’85–M’88–SM’98) was born and educated in Singapore. He received the B.Eng. (First Class) and M.Eng. degrees from the National University of Singapore, Singapore and the Ph.D. degree from Cambridge University, Cambridge U.K. He is currently an Associate Professor in the School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. He has more than twenty years of academic, research, industrial, and professional experience in electrical power and energy systems. He has undertaken numerous contract research projects for major corporations such as Vestas, RollsRoyce, and Bosch, and has been holding key advisory appointments in both public and private sector in Singapore. He continues to be awarded research grants from industries, research institutes, and funding agencies. He has published more than a hundred technical papers, and actively reviews and edits papers for major international journals and conferences. Dr. Tseng is a Fellow of Cambridge Commonwealth Society and Cambridge Philosophical Society. He is a senior corporate member of the Institute of Engineers Singapore, the Institution of Engineering and Technology (U.K.), and the Institute of Electrical and Electronic Engineers (USA). He is a Chartered Engineer registered in U.K. In 1996, he was awarded the Swan Premium by the IET, for his work in power engineering. He has held a number of major appointments in professional societies including the Chair of IEEE Singapore Section in 2005. He was awarded the IEEE Third Millennium Medal and the IEEE Region Ten Outstanding Volunteer Award for his contributions to engineering education and technologies.