Effective Emi Filter.pdf

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 25, NO. 11, NOVEMBER 2010

2797

Effective EMI Filter Design Method for Three-Phase Inverter Based Upon Software Noise Separation Po-Shen Chen and Yen-Shin Lai, Senior Member, IEEE

Abstract—This paper presents an electromagnetic interference (EMI) filter design method for a three-phase inverter applied to motor drives. The EMI filter design is based upon the results of software noise-separation method. Therefore, no hardware noise separator is required. Moreover, the presented EMI filter design approach provides systematic design procedure, and therefore, trialand-error overhead is not required, and thereby, shortening the development time and saving the cost. The presented method determines common-mode choke and differential-mode choke based upon the required impedance to attenuate the noise. A three-phase inverter for inductor motor drives is used as the equipment under test (EUT), and the related conductive EMI filter is designed and verified. It will be shown that EMI filter designed by this systematic design method effectively attenuates the conductive EMI components to meet the requirement of EN 55011. Index Terms—Common-mode (CM) and differential-mode (DM) signals, electromagnetic interference (EMI) filter, threephase inverter.

V2,DM AT T -CM AT T -DM Rload

Zi,CM ZL ,CM ZCY Zi,DM ZL ,DM ZC X 1 and ZC X 2

NOMENCLATURE CM DM VX , VY , and VZ

VCM and VDM Areq,CM fCM fCM ,con CY LC Areq,DM fDM fDM ,con CX LD V1,CM V2,CM V1,DM

Common mode. Differential mode. Total electromagnetic interference (EMI) for phase R, S, and T, respectively. CM and DM EMI, respectively. Desired attenuation of CM component. Frequency of CM noise. Corner frequency of CM filter. Y capacitor. CM inductor. Desired attenuation of DM component. Frequency of DM noise. Corner frequency of DM filter. X capacitor. DM inductor. Measured CM voltage with CM filter (in dB·µV). Measured CM voltage without CM filter (in dB·µV). Measured DM voltage with DM filter (in dB·µV).

Manuscript received January 12, 2010; revised March 24, 2010 and May 14, 2010; accepted May 17, 2010. Date of current version October 29, 2010. Recommended for publication by Associate Editor J.-L. Schanen. The authors are with the Center for Power Electronics Technology, National Taipei University of Technology, Taipei 106, Taiwan (e-mail: yslai@ ntut.edu.tw). 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/TPEL.2010.2051459

IS,CM and VS,CM IS,DM and VS,DM

Measured DM voltage without CM filter (in dB·µV). Ratio of CM voltage without the CM filter over its counterpart with CM filter. Ratio of DM voltage without the DM filter over its counterpart with DM filter. Equivalent circuit impedance of line-impedance stabilization network (LISN). Current-source impedance of CM noise source. Impedance of the CM inductor. Impedance of CM capacitance. Current-source impedance of DM noise source. Impedance of the DM inductor. Impedance of DM of the capacitors 1 and 2, respectively. CM noise, current, and voltage sources, respectively. DM noise, current, and voltage sources, respectively.

I. PROBLEM DESCRIPTION AND OBJECTIVES NVERTERS have been widely applied to motor drives, renewable energy conversion, etc. Although giving the specified voltage and frequency of inverter is the main concern for inverter control, EMI issue mainly contributed by pulsewidth modulation (PWM) switching becomes more popular, especially for residential applications. Several methods [16]–[26] have also been presented to cope with the EMI of power electronics. These methods focus on the topics related to parasitic-noise-source cancellation [16]–[19], passive and active EMI filters [21], switching schemes [22]–[24], reduction of DM EMI generation [25], and applications of Wiener filter to EMI estimation [26]. To deal with the EMI issues of three-phase inverter, several methods have been discussed. For three-phase inverters with capacitors, CM filters can be used to reduce the CM EMI components, as reported in [1] and [2]. Noise-separation methods for three-phase inverter EMI signals have been discussed in [3]–[7]. In [8] and [9], EMI filter is installed between the input power and inverter based upon corner-frequency method [8], [9] or Butterworth filter method [10]. For these approaches, systematic design procedure of EMI filter design has not been discussed or high-frequency noise cannot be effectively suppressed [9].

I

0885-8993/$26.00 © 2010 IEEE

2798

Fig. 1.

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 25, NO. 11, NOVEMBER 2010

Attenuation and corner frequency for filter design, earlier method. Fig. 4. Conductive EMI signals and measurement. V X : total EMI (phase “R” side), VY : total EMI (phase “S” side), and V Z : total EMI (phase “T” side).

Fig. 2. Illustration of failure case for EMI noise suppression, earlier filter design method.

The main theme of this paper is to present a systematic approach to the conductive EMI filter design for the three-phase inverter based upon software-noise-separation method [11], [12]. Based upon the presented systematic method, the tolerance of noise from the required standard can be specified and the EMI filter is designed accordingly. Therefore, trial-and-error, which may not be allowed for high-power inverter, is not required for the presented filter design method. Moreover, the characteristics of filter components are investigated in this paper to guarantee effective noise suppression. An induction motor drive using three-phase inverter is the EUT for experimental verification. The induction motor is operated in the rated speed and load, and the inverter switching frequency (SF) is up to 15 kHz. It will be shown that EMI filter designed by this systematic design method effectively attenuates the EMI components to meet the requirement of EN 55011 conducted emission limit [13]. II. SOFTWARE-BASED NOISE-SEPARATION METHOD A. CM and DM Noise

Fig. 3. Failure case for EMI noise suppression, frequency-dependent characteristics of filter components.

Fig. 1 shows the attenuation and corner frequency for previous filter design method. As shown in Fig. 1, the corner frequency of filter is determined by drawing the 40-dB/decade slope line, which is tangent to the required CM or DM components. However, the high-frequency components may not effectively suppress the EMI noise, as illustrated in Fig. 2. For Fig. 2, the high-frequency noise, as indicated by “HF” in Fig. 2, the EMI noise cannot be reduced effectively by the filter with slope of (−40) dB/decade. Moreover, the EMI noise cannot be mitigated to meet the required standard due to the frequency dependency of the characteristic features for filter components. Due to the change of components of filter, noise-attenuation-contributed filter is changed from that for “ideal filter” to “real filter,” as shown in Fig. 3.

Fig. 4 shows the conductive EMI signals and measurement circuit of a three-phase system. The conductive EMI signals are classified as CM noise and DM noise, which can be separated with each other by hardware noise separator [14]. As shown in Fig. 1, the CM noise is defined by VCM =

VX + VY + VZ 3

(1)

and the related DM noise between two phases, say phase “R” and phase “S,” is VDM -RS = |VX − VY |.

(2)

B. Software-Based Noise-Separation Method Fig. 5 shows the connection of the presented softwarebased separation method. The CM-noise and DM-noise separation is performed using software written in LabVIEW, as shown in Fig. 5. As indicated in Fig. 5, the three-phase LISN (NNB-4/200X), which is with single-terminal output socket, is

CHEN AND LAI: EFFECTIVE EMI FILTER DESIGN METHOD FOR THREE-PHASE INVERTER BASED UPON SOFTWARE NOISE SEPARATION

Fig. 5.

2799

Setup of the presented software-based separation method.

Fig. 7.

Three-phase LISN with single-port socket, NNB-4/200X.

TABLE I DATA OF MOTOR AND CABLE

Fig. 6.

Flow chart for software-based separation of conductive EMI signals.

used for EMI measurement and separation. Fig. 6 illustrates the flow chart of the presented software-based noise-separation method. It is worthy of noting that the measured values of “VX ,” “VY ,” and “VZ ” are in decibel microvolts in general. To get the accuracy results using (1) and (2), it is required to convert the values in microvolts before calculating VCM and VDM . After converting VCM and VDM in terms of decibel microvolts, separation of VCM and VDM from the measured values of “VX ,” “VY ,” and “VZ ” can be achieved by (1) and (2). III. EXPERIMENTAL SYSTEM AND NOISE SEPARATION A. Experimental Setup and Standard The measured signals are fed to the notebook computer via an EMI analyzer (ADVANTEST R3132) and general-purpose interface bus interface. A three-phase LISN with single-port socket, as shown in Fig. 7, is used for measurement. The details of the EUT and connection cable are shown in Table I. Fig. 8 shows the block diagram of three-phase inverter. As shown in Fig. 8, an induction motor specified in Table I con-

Fig. 8.

Block diagram of three-phase inverters.

nected with a three-phase inverter forms the EUT. Fig. 9 shows the photo of the experimental setup. B. Measurement and Noise Separation When the inverter is operated in the following conditions: fundamental frequency = 60 Hz, SF = 15 kHz, and rated output power. Figs. 10–12 illustrate the measurement results of VX , VY , and VZ , respectively. As demonstrated in Figs. 13 and 14, the CM and DM components are separated. As shown in Fig. 13, for VCM , the maximum value = 90 dB·µV at 686 kHz, which is far beyond its limit value, 60 dB·µV. Similar remarks can be derived for DM noise by the results shown in Fig. 14. Therefore, the designed EMI filter has been demonstrated to decrease the noise to meet the related standard.

2800

Fig. 9.

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 25, NO. 11, NOVEMBER 2010

Photo of experimental setup.

Fig. 10.

Measurement results, V X , maximum value = 90 dB·µV.

Fig. 11.

Measurement results, V Y , maximum value = 90 dB·µV.

Fig. 12.

Measurement results, V Z , maximum value = 88 dB·µV.

Fig. 13.

Measurement results, V C M , maximum value = 90 dB·µV, at 686 kHz.

Fig. 14. Measurement results, V D M , maximum value = 81.47 dB·µV, at 656 kHz.

IV. EARLIER EMI FILTER DESIGN METHOD AND RESULTS For the development to follow and highlight the special features of the proposed conductive EMI filter design method, a brief review of the earlier method is introduced. Moreover, the measured results for the previous method are presented for comparison and confirmation. For conventional EMI filter design method, as shown in Fig. 1, the corner frequency of filter is determined by drawing the 40-dB/decade slope line, which is tangent to the required attenuation component, as indicated by “C” in Fig. 1.

Earlier EMI filter design is illustrated using CM filter as an example. Fig. 15(a) shows the measured CM noise and the related limited line for the tested inverter and IM motor drives. The required attenuation and corner frequency of CM filter is shown in Fig. 15(b). The corner frequency of CM filter is determined as follows. The maximum value of VCM = 90 dB·µV, which occurs at 686 kHz. And VCM = 78 dB·µV at 150 kHz. The limited value is 60 dB and the required limit value becomes 54 dB as considering the 6-dB tolerance. Therefore, the desired

CHEN AND LAI: EFFECTIVE EMI FILTER DESIGN METHOD FOR THREE-PHASE INVERTER BASED UPON SOFTWARE NOISE SEPARATION

2801

Fig. 15. Measured results for EMI filter design, earlier method. (a) Measured CM noise and limit line. (b) Required attenuation and corner frequency of CM filter. (c) Measured DM noise and limit line. (d) Required attenuation and corner frequency of DM filter.

attenuation amplitude of CM component can be derived by Areq,CM = VCM − Limit = 40 log

fCM . fCM ,con

(3)

where Areq,CM is the desired attenuation of CM component, which is 78 − 54 = 24 dB·µV and fCM is the frequency of CM noise, which is 150 kHz. Therefore, by (3), fCM ,con , i.e., corner frequency of CM filter, is 37.678 kHz.√The CM choke can thus be determined by fCM ,con = 1/[2π LC 3CY ] and LC = 1.265 mH, if CY = 4700 pF. The CM choke is implemented using the FT-3KM K3824G core produced by FINEMET CO., LTD. The required number of turns is ten. Using impedance analyzer (HP 4194 A), the measured CM inductance is 1.448 mH at 150 kHz. Fig. 15(c) presents the measured DM noise and the related limited line for the tested inverter and IM motor drives. The required attenuation and corner frequency of DM filter is shown in Fig. 15(d). The corner frequency of DM filter is calculated as follows. The maximum value of VDM = 81.47 dB·µV, which occurs at 656 kHz; and VDM = 73.3 dB·µV at 150 kHz. The limited value is 60 dB and the required limit value becomes 54 dB as considering the 6-dB tolerance. Therefore, the desired attenuation amplitude of CM component can be derived by Areq,DM = VDM − Limit = 40 log

fDM fDM ,con

(4)

whereAreq,DM is the desired attenuation of DM component = 73.3 − 54 = 19.3 dB·µV and fDM is the frequency of DM noise = 150 kHz.

Fig. 16.

Measured results, total noise.

Therefore, by (4), fDM ,con , i.e., corner frequency of DM filter, is 49.38 kHz. √ The DM choke can thus be determined by fDM ,con = 1/[2π 2LD CX ] and LD = 7.63 µH provided CX = 0.68 µF. Fig. 16 shows the measured results for the designed EMI filter using earlier corner-frequency filter design method. As shown in Fig. 16, the EMI noise can be suppressed at lower frequency range to meet the EN55011. However, at the higher frequency range, the EMI filter cannot suppress the related EMI effectively. As shows in Fig. 17, the choke may become capacitive because of parasitic capacitor when the frequency is higher than a certain value. Therefore, noise attenuation contributed by the EMI filter is changed from that for “ideal filter” to “real filter,” as shown

2802

Fig. 17.

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 25, NO. 11, NOVEMBER 2010

Fig. 19.

Circuit of EMI filter.

Fig. 20.

Design steps of the presented EMI filter design.

Frequency response of CM choke.

Fig. 18. Schematic diagram for explaining the presented filter design. (a) Measured noise and limit line. (b) Required attenuation of EMI filter.

in Fig. 3. This change makes the higher frequency EMI noise not meet the required specifications. V. PROPOSED EMI FILTER DESIGN METHOD Fig. 18 shows the schematic diagram for explaining the presented EMI filter design method. As shown in Fig. 18(a), the attenuation level required for the designed filter is determined by the measured noise, CM, or DM noise, and the related limit line. There exists some tolerance, e.g., 6 dB, between the required attenuation and limit line, as shown in Fig. 18(a). The presented method determines CM choke and DM choke based upon the required impedance to attenuate the noise, as indicated by “C” in Fig. 18(b). Fig. 19 shows the filter circuit that contains CM and DM filters. How to select the circuit components to reduce the EMI

component to meet EN 55011-conducted emission limit, while not invoking any trial-and-error is the main theme of the proposed design method. Fig. 20 illustrates the design steps of the presented filter design. As shown in Fig. 20, the presented method determines CM and DM chokes based on the required impedance of CM (DM) choke to attenuate the CM (DM) noise. As indicated in Fig. 20, the CM and DM components are separated by the software-based scheme. The frequencies of amplitude of CM and DM components are then clearly identified. The amplitudes of the CM and DM components are compared with the related standards to give the attenuation amplitude of the EMI filter. In the following filter design steps, the EMI noise is

CHEN AND LAI: EFFECTIVE EMI FILTER DESIGN METHOD FOR THREE-PHASE INVERTER BASED UPON SOFTWARE NOISE SEPARATION

Fig. 21.

2803

Equivalent circuit of CM noise without filter. Fig. 23.

Equivalent circuit of DM noise without filter.

be derived according to Fig. 22 as follows: V1,CM =

ZC Y IS,CM Zi,CM Rload Zi,CM + Z1 ZC Y + ZL ,CM + Rload

(8)

Z1 = (ZL ,CM + Rload ) // (ZC Y ) .

(9)

where Fig. 22.

Equivalent circuit of CM noise with filter.

Equation (8) can be further reduced, as indicated in (10), provided Zi,CM  Z1

regarded as a current source. The EMI noise may be treated as a voltage source; however, the effect of Y capacitor (X capacitor) cannot be included in the CM filter (DM filter) design for this approach. The deduction process is similar to that shown in this paper for regarding the EMI source as current source. A. CM Filter Design The attenuation amplitude is shown in (5), which is the ratio of CM voltage without over its counter part with CM filter AT T =

Vnoise without filter V2 = Vnoise with filter V1

(5)

where Vnoise with filter is the noise voltage by measuring LISN with filter, in decibel microvolts and Vnoise without filter is the noise voltage by measuring LISN without filter, in decibel microvolts. Fig. 21 shows the equivalent circuit of CM noise without filter. Zi,CM means the current-source impedance of CM noise source. The CM voltage by measuring LISN without filter can be, therefore, derived as follows: V2,CM = IS,CM

Zi,CM Rload . Zi,CM + Rload

(6)

The internal impedances of equipment under test can be neglected (Zi,CM is close to infinity), if current source is used to represent the noise source for the analysis, as shown [9], [15]; further reduction of (6) can be derived, as indicated in (7), provided Zi,CM  Rload V2,CM ≈ IS,CM Rload .

(7)

Fig. 22 illustrates the equivalent circuit of CM noise with filter. In Fig. 22, ZL ,CM indicates the impedance of the CM inductance, and ZCY indicates the impedance of CM capacitance. The CM voltage from measuring LISN with the CM filter can

V1,CM ≈ IS,CM

ZC Y

ZC Y Rload . + ZL ,CM + Rload

(10)

As shown in (10), the measured results indicate the component of the CM noise contributed by the equipment under test with attenuation via the CM filter. Substituting (7) and (10) into (5) yields the impedance of CM inductance and can be derived as follows: ZL ,CM = (AT T −CM − 1) ZC Y − Rload .

(11)

B. DM Filter Design Fig. 23 shows the equivalent circuit of DM noise without filter. In Fig. 23, Zi,DM indicates the current-source impedance of DM noise source. By Fig. 23, the DM voltage by measuring LISN without DM filter can be, therefore, derived as follows: V2,DM = IS,DM

Zi,DM Rload . Zi,DM + Rload

(12)

The internal impedances of equipment under test can be neglected (Zi,DM is close to infinity), if current source is used to represent the noise source for the analysis as shown in [9] and [15]; (12) can be further reduced, as indicated in (13), provided Zi,DM  Rload V2,DM ≈ IS,DM Rload .

(13)

Fig. 24 illustrates the equivalent circuit of DM noise with filter. In Fig. 24, ZL ,DM indicates the impedance of the DM inductance, and ZCX1 and ZCX2 indicate the impedance of the DM capacitance. The DM voltage by measuring LISN with DM filter can, therefore, be derived from Fig. 24 as follows: V1,DM =

ZC X 1 IS,DM ZS,DM ZS,DM + Z2 ZC X 1 + ZL ,DM + Z1 ×

ZC X 2 Rload ZC X 1 + Rload

(14)

2804

Fig. 24.

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 25, NO. 11, NOVEMBER 2010

Equivalent circuit of DM noise with filter.

where Z1 = Rload //ZC X 2 ,

and Z2 = (Z1 + ZL ,DM ) //ZC X 1 .

Further reduction of (14) can be derived, as indicated in (15), provided that Zi,DM  Z2 ZC X 1 ZC X 2 Rload . V1,DM ≈ IS,DM (ZC X 1 + ZL ,DM + Z1 ) (ZC X 2 + Rload ) (15) Substituting (13) and (15) into (5) yields, (16) as shown at the bottom of this page. Given ZC X 1 = ZC X 2 = ZC X and by (16), the impedance of DM inductance can be derived as follows:


(AT T -DM − 1)ZC2 X − 2ZC X Rload


(17) ZL ,DM =

(ZC X + Rload )

Fig. 25.

Implemented EMI filter for three-phase inverter.

TABLE II DESIGNED AND IMPLEMENTATION VALUES OF COMPONENTS

where ZC X =

1 ωCX

and ZL ,DM = ω2LD . other one with low-µ core works in the frequency greater than 1.5 MHz.

C. CM EMI Filter Design From the test results shown in Fig. 13, the CM noise over limit from 150 kHz to 8.5 MHz is calculated at different frequency points using the maximum attenuation amplitude. For example, the maximum amplitude CM noise is equal to 90 dB·µV at 686 kHz. The ATT-CM is 90 − 54 (6 dB·µV tolerance) = 36 dB·µV. ZL ,CM is 559.23 Ω, using (11) assuming the Y capacitor = 4700 pF. In this paper, the selection of Y capacitor only considers the leakage current issue to meet EN 55081–2. Increase of Y capacitor can reduce the size of CM inductance at the risk of more leakage current ZL ,CM

50 1 − = (36 − 1) 2 × π × 0.686 M × 3 × 4700 pF 3

= 559.23 Ω 559.23 = 130 µH. LC = 2 × π × 0.686 M In this paper, two inductors connected in series are used to be the CM choke for CM filter. The inductor with high-µ core function works in the low frequency of 0.15–1.5 MHz, and the

AT T -DM =

D. DM EMI Filter Design By the test results shown in Fig. 14, the DM noise over limit from 0.15 MHz to 1.5 MHz is calculated at different frequency points using the maximum attenuation amplitude. For example, the maximum attenuation amplitude of DM noise is equal to 81.47 dB·µV at 656 kHz. ZL ,DM is 0.6857 Ω using (17) assuming X capacitor = 0.68 µF, LD = 82 nH. VI. MEASUREMENT RESULTS A. Frequency Dependency of Filter Components As shown in Fig. 25, the CM choke consists of two parts. One works in low-frequency range and the other one is for higher frequency range. It can be seen in Table II that the component values for implementation are greater than their designed ones. Table II shows the designed and implemented values of filter components. As shown in Table II, the capacitance values agree with each other quite well. Moreover, the implemented

IS,DM Rload . IS ,DM ((ZC X 1 ZC X 2 Rload )/(ZC X 1 + ZL ,DM + Z1 )(ZC X 2 + Rload ))

(16)

CHEN AND LAI: EFFECTIVE EMI FILTER DESIGN METHOD FOR THREE-PHASE INVERTER BASED UPON SOFTWARE NOISE SEPARATION

Fig. 26. Inductance value of CM choke versus frequency during 0.15– 8.5 MHz.

Fig. 28.

2805

Measurement results, total noise.

VII. CONCLUSION

Fig. 27. Inductance value of DM choke versus frequency during 0.15– 1.5 MHz.

The main theme of this paper has been to present a systematic and effective design procedure to design the EMI filter of threephase inverter for motor drives applications. The contributions of this paper are summarized as follows. 1) Presented software-based noise-separation method. 2) Proposed a systematic and effective EMI filter design method for three-phase inverter. 3) Investigated the frequency dependency of EMI filter components. 4) Designed the EMI filter and confirm its effectiveness by measurement results. REFERENCES

inductance is higher than its design values. Therefore, the filter can be effectively suppressed by the EMI noise even though the frequency varies in a very wide range. Fig. 26 shows the comparison results of designed inductance and measured values. The measured results agree very well with its designed values in wide frequency range. To save the cost, the leakage inductance of CM choke, as shown in Fig. 27, is used as the DM choke. As shown in Fig. 27, the leakage inductance is much higher than the designed values in a very wide frequency range. B. EMI Measurement Results Fig. 28 shows the measurement results for the EMI filter designed by the earlier method and the presented method. As shown in Fig. 28, the earlier filter design method cannot provide satisfied results at high-frequency range. In contrast, the proposed method, indeed, suppresses the EMI noise to meet the conducted emission standard of EN 55011. These results confirm the claims and the effectiveness of the proposed EMI filter design method. Moreover, for the proposed method, the attenuation at low frequency is significant, as shown in Fig. 28. This is contributed by the DM inductance LD . In the implementation, the leakage inductance of CM inductance is used as LD, which is much higher than its designed value, as shown in Table II. Therefore, the increase in LD results in relevant attenuation at low frequency.

[1] N. Hanigovszki, J. Landkildehus, G. Spiazzi, and F. Blaabjerg, “An EMC evaluation of the use of unshielded motor cables in ac adjustable speed drive applications,” IEEE Trans. Power Electron., vol. 21, no. 1, pp. 273– 281, Jan. 2006. [2] S. Ogasawara and H. Akagi, “Analysis and reduction of EMI conducted by a PWM inverter-fed ac motor drive system having long power cables,” in Proc. IEEE Power Electron. Spec. Conf., 2000, vol. 2, pp. 928–933. [3] Q. Liu, W. Shen, F. Wang, D. Borojevich, and V. Stefanovic, “On discussion of motor drive conducted EMI issues,” in Proc. IEEE Power Electron. Motion Control Conf., Aug. 2004, vol. 3, pp. 1515–1520. [4] W. Shen, F. Wang, and D. Boroyevich, “Conducted EMI characteristic and its implications to filter design in 3-phase diode front-end converters,” in Proc. IEEE-IAS Ann. Meeting, Oct. 2004, pp. 1840–1846. [5] W. Shen, F. Wang, D. Boroyevich, and Y. Liu, “Definition and acquisition of CM and DM EMI noise for general-purpose adjustable speed motor drives,” in Proc. IEEE Power Electron. Spec. Conf., 2004, vol. 2, pp. 1028– 1033. [6] L. Ran, J. C. Clare, K. J. Bradley, and C. Christopoulos, “Measurement of conducted electromagnetic emissions in PWM motor drive systems without the need for an LISN,” IEEE Trans. Electromagn. Compat., vol. 41, no. 1, pp. 50–55, Feb. 1999. [7] M. L. Heldwein, T. Nussbaumer, F. Beck, and J. W. Kolar, “Differential mode EMC input filter design for three-phase ac-dc-ac sparse matrix PWM converters,” in Proc. IEEE Power Electron. Spec. Conf., 2004, vol. 1, pp. 284–291. [8] Y. S. Lee, T. T. Yang, and M. W. Cheng, “Measurement and mitigation of conducted emission for voltage phase controlled capacitor-run single phase induction motor,” in Proc. IEEE Power Electron. Drive Syst., Nov. 2005, vol. 2, pp. 1622–1627. [9] F. Y. Shih, D. Y. Chen, Y. P. Wu, and Y. T. Chen, “A procedure for designing EMI filters for AC line applications,” IEEE Trans. Power Electron., vol. 11, no. 1, pp. 170–181, Jan. 1996. [10] C. Khun, V. Tarateeraseth, W. Khan-ngern, and M. Kando, “Design procedure of common-mode filter for induction motor drive using Butterworth function,” in Proc. IEEE Power Convers. Conf., Apr. 2007, pp. 417–422.

2806

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 25, NO. 11, NOVEMBER 2010

[11] P. S. Chen and Y. S. Lai, “Software-based separation of conductive EMI signals,” in Proc. IEEE-IAS Ann. Meeting, Oct. 2006, pp. 1209–1214. [12] P. S. Chen and Y. S. Lai, “New EMI filter design method for single phase power converter using software-based noise separation method,” in Proc. IEEE-IAS Ann. Meeting, Sep. 2007, pp. 2282–2288. [13] Industrial, Scientific and Medical Radio-Frequency Equipment—Radio Disturbance Characteristics—Limits and Methods of Measurement, IEC Standard EN 55011, 1998. [14] M. L. Heldwein, T. Nussbaumer, F. Beck, and J. W. Kolar, “Novel threephase CM/DM conducted emissions separator,” in Proc. IEEE Appl. Power Electron. Conf., 2005, vol. 2, pp. 797–802. [15] H. J. Chiu, T. F. Pan, C. J. Yao, and Y. K. Lo, “Automatic EMI measurement and filter design system for telecom power supplies,” IEEE Trans. Instrum. Meas., vol. 56, no. 6, pp. 2254–2261, Dec. 2007. [16] P. Kong, S. Wang, and F. C. Lee, “Common mode EMI noise suppression for bridgeless PFC converters,” IEEE Trans. Power Electron., vol. 23, no. 1, pp. 291–297, Jan. 2008. [17] P. Kong, S. Wang, F. C. Lee, and C. Wang, “Common-mode EMI study and reduction technique for the interleaved multichannel PFC converter,” IEEE Trans. Power Electron., vol. 23, no. 5, pp. 2576–2584, Sep. 2008. [18] S. Wang, F. C. Lee, and J. D. van Wyk, “A study of integration of parasitic cancellation techniques for EMI filter design with discrete components,” IEEE Trans. Power Electron., vol. 23, no. 6, pp. 3094–3102, Nov. 2008. [19] H. F. Chen, C. Y. Yeh, and K. H. Lin, “A method of using two equivalent negative inductances to reduce parasitic inductances of a three-capacitor EMI filter,” IEEE Trans. Power Electron., vol. 24, no. 12, pp. 2867–2872, Dec. 2009. [20] H. Akagi and T. Shimizu, “Attenuation of conducted EMI emissions from an inverter-driven motor,” IEEE Trans. Power Electron., vol. 23, no. 1, pp. 282–290, Jan. 2008. [21] J. Biela, A. Wirthmueller, R. Waespe, M. L. Heldwein, L. Raggl, and J. W. Kolar, “Passive and active hybrid integrated EMI filters,” IEEE Trans. Power Electron., vol. 24, no. 5, pp. 1340–1349, May 2009. [22] K. Mainali, R. Oruganti, K. Viswanathan, and S. P. Ng, “A metric for evaluating the EMI spectra of power converters,” IEEE Trans. Power Electron., vol. 23, no. 4, pp. 2075–2081, Jul. 2008. [23] A. A. Fardoun and E. H. Ismail, “Reduction of EMI in ac drives through dithering within limited switching frequency range,” IEEE Trans. Power Electron., vol. 24, no. 3, pp. 804–811, Mar. 2009. [24] R. Yang, B. Zhang, D. Qiu, and Z. Liu, “Time–frequency and wavelet transforms of EMI dynamic spectrum in chaotic converter,” IEEE Trans. Power Electron., vol. 24, no. 4, pp. 1083–1092, Apr. 2009. [25] P. Zumel, O. Garcia, J. A. Oliver, and J. A. Cobos, “Differential-mode EMI reduction in a multiphase DCM flyback converter,” IEEE Trans. Power Electron., vol. 24, no. 8, pp. 2013–2020, Aug. 2009.

[26] P. Musznicki, J. L. Schanen, P. Granjon, and P. J. Chrzan, “The Wiener filter applied to EMI decomposition,” IEEE Trans. Power Electron., vol. 23, no. 6, pp. 3088–3093, Nov. 2008.

Po-Shen Chen received the B.S. and M.S. degrees in electrical engineering from the National Taipei University of Technology, Taipei, Taiwan, where he is currently working toward the Ph.D. degree. His research interests include motor drivers, inverter technologies, and mitigation of EMI for driver systems.

Yen-Shin Lai (M’96–SM’01) received the M.S. degree from the National Taiwan University of Science and Technology, Taipei, Taiwan, and the Ph.D. degree from the University of Bristol, Bristol, U.K., both in electronic engineering. In 1987, he joined the Department of Electrical Engineering, National Taipei University of Technology, Taipei, Taiwan, as a Lecturer, where he has been a Full Professor since 1999 and was the Chairperson during 2003–2006. Since 2006, he has been a Distinguished Professor. His research interests include design of control IC, circuit design of dc/dc converter, and inverter control. Dr. Lai received several national and international awards, including the John Hopkinson Premium for the session 1995–1996 from the Institute of Electrical Engineers (IEE), the Technical Committee Prize Paper Award from the IEEE Industry Application Society (IAS) Industrial Drives Committee for 2002, the Best Presentation Award from IEEE IECON, 2004. He was the Secretary of IEEE IAS Industrial Drives Committee during 2008–2009. He is currently the Vice Chair, IEEE IAS Industrial Drives Committee. He is also the Associate Editor of the IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS and IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS.