Study on synthetic system for conducted EMI noise ... - Science Direct

Procedia Earth and Planetary Science 1 (2009) 1593–1598

Procedia Earth and Planetary Science www.elsevier.com/locate/procedia

The 6th International Conference on Mining Science & Technology

Study on synthetic system for conducted EMI noise analysis and suppression+ Zhao Yanga, b,*, Lu Xiao-quana, Dong Ying-huaa, Feng Zhi-mingc, Zhao Boc, Yan Weia a

School of Electrical & Automation Engineering, Nanjing Normal University, Nanjing 210042, China b State Key Laboratory of Millimeter Waves, Nanjing 210096, China c Jiangsu Metrology Institute of Sc`ience & Technology, Nanjing 210007, China

Abstract A synthetic system for conducted EMI noise analysis and suppression is proposed in this paper. With this system, the conducted EMI noise from the equipment under test (EUT) can be collected and analyzed, and its noise suppression scheme can be given by the system automatically. Technical studies and applications prove that the proposed system is an efficient solution to conducted EMI noise analysis and suppression.

Keywords: conducted electromagnetic interference (EMI), synthetic system, noise analysis and suppression

1. Introduction Recently, electromagnetic interference (EMI) problems are more and more serious for power electronic systems[1]-[5]. Therefore, the effective solutions for conducted EMI noise analysis and suppression have become great concerns. The conducted EMI noise is composed of two components, one is CM noise and the other is DM noise. Furthermore, the EMI filter should be designed for CM noise and DM noise, respectively. In order to diagnose and analyze the conducted EMI noise, noise separating networks based on radio-frequency transformers have been proposed to identify the noise modes in reference [6]-[8]. But their high frequency performances are limited by parasitic effects of transformer. Moreover, in order to suppress the conducted EMI noise, the proper design of EMI filter is necessary. Noise source impedance modeling is used to match the load impedance for EMI filter design. An efficient modeling approach for EMI noise source impedance is discussed by using two current probes [9]. After that, the EMI filter can be designed. A modal measurement of EMI filter based on scattering parameters is presented to analyze its characterization with the advantage of simplicity and precision [6] .However, it still lack a synthetic solution scheme including conducted EMI noise analysis and suppression methods. In order to solve this problem, a synthetic system for conducted EMI noise analysis and suppression is introduced in this paper, and its design scheme and major techniques are discussed. Furthermore, the system is applied to diagnose the conducted noises of switching-mode power supply (SMPS), and EMI filter in the system is verified to be able to suppress the unwanted noise of the SMPS efficiently. Applications prove that the proposed system is an efficient tool for noise analysis and suppression.

2. System design scheme

* Corresponding author. Tel.: +86-25-85481170; fax: +86-25-85481170. E-mail address: [email protected].

1878-5220 © 2009 Published by Elsevier B.V. Open access under CC BY-NC-ND license. doi:10.1016/j.proeps.2009.09.245

1594

Z. Yang et al. / Procedia Earth and Planetary Science 1 (2009) 1593–1598

2.1. Hardware design for the synthetic system The hardware schematic of the synthetic system is shown in Fig.1.(a). Line impedance stabilization network (LISN) is placed between the power line and EUT. The noise collected from LISN is fed into the hardware of synthetic system, where it is separated into CM and DM noise by noise separating network. Controlled by FPGA, there is only one signal among CM, DM and the total noise sent through switch K. Then, this signal is amplified by 40dB. According to EMC standards, the observation frequency is between 10kHz to 30MHz, therefore, the amplified signal is sent through a low-pass filter and a high-pass filter to remove the noises outside the observation frequency band. After that, also controlled by FPGA, these analog signals are converted into digital signals by A/D transformer and temporarily saved in RAM, and then sent through the serial port into the computer installed with the software of the synthetic system.

Fig. 1. (a) hardware schematic of the synthetic system; (b) software menu of the synthetic system

2.2. Software design for the synthetic system In the synthetic system for conducted EMI noise analysis and suppression, the main functions of its software are shown by the software menu in Fig.1(b). Firstly, the conducted CM, DM and total noise of EUT collected by the hardware is displayed on the software interface. Secondly, whether these noises comply with EMC standard or not, they are analyzed and presented. Thirdly, if the noise can not meet the EMC standard, the EMI filter for CM and DM noise can be easily obtained by the software. Afterwards, the noise suppression results are predicted to evaluate the capability of the proposed filter. Finally, a noise analysis and suppression report can be printed for the user. Besides, an introduction of the software is provided for the user as a help. 3. Technique studies 3.1. A high-performance noise separating network In order to improve the noise separating performance, a high-performance noise separating network [11] is used in the synthetic system for conducted EMI noise analysis, which is constructed of power splitters shown in Fig.2. (a). There are four parameters defined to evaluate a noise separating network, that is , the common-mode insertion loss (CMIL), differential-mode insertion loss (DMIL), common-mode rejection ratio (CMRR) and differential-mode rejection ratio(DMRR). The transmission coefficient of noise separating network are defined as S21=20log(V2/V1) (dB)

(1)

where V1 and V2 are the input and output of the network. As the transmission coefficient S21 is CMIL/DMIL, the V1 and V2 represent the same mode signals, and as S21 is CMRR/ DMRR, V1and V2 represent different mode signals [12]. It should be mentioned that the insertion loss should be not more than 5dB and the rejection ratio should be not less than 40 dB. A spectrum analyzer and 0o/180o splitters are used to measure the characterization parameters of noise separating network [10]. The CMIL result of the high-performance noise separating network is shown in Fig.2. (b). As the frequency goes up, the CMIL declines slightly but remains above -2dB. The DMRR result of the high-performance noise separating network is shown with a good performance in Fig.2. (c). The DMRR rises with increasing frequency, and at 30MHz, the maximum frequency for conducted EMI noise measurement, the DMRR successful remains below -40dB. The characterization measurement results prove that the noise separating network adopted by synthetic system can separate the conducted EMI noise efficiently.


1595

L N LISN E

CM+DM CM-DM

0o/180o Splitter CM/DM Spectrum Analyzer

Fig. 2. (a) Noise separating network constructed by 0o/180o splitter; (b) CMIL of the noise separating network; (c) DMRR of the noise separating network

3.2. A noise source impedance modeling approach by using two current probes The EMI filter is proposed by the synthetic system as the noise suppression tool. The performance of EMI filter relies on the noise source impedance of the circuit and the noise load impedance on test site. The noise load in the synthetic system is LISN, whose DM impedance is 100Ω and CM impedance 25Ω. Therefore, a noise source impedance modeling approach [4] by using two current probes is adopted in the synthetic system for EMI filter design. The test set-up of the two probes approach is shown in Fig. 3. (a) .and the unknown impedance at b-b’ is given as

Zx =

KVsig Vp 2

− Z in =

Z x = SMPS

Rstd V p 2 Vp2

Z x = Rstd

− Z in

Z x = SMPS

(2)

The CM noise impedance achieved by using two current probes approach is shown in Fig. 3. (b).The CM noise impedance declines as the frequency goes up and can be modeled as a 4.6Ω resistor and a 14000 pF capacity in series. This indicates that the noise source impedance modeling approach by using two current probes can provide the noise source impedance for the synthetic system. 500 450 400

impedance(ohm)

350 300 250 200 150 100 50 0

0

5

10

15 f/(MHz)

20

25

30

Fig. 3. (a)test setup of the two current probes approach; (b)CM noise impedance of a SMPS achieved by using two current probes approach

3.3. A measurement approach for EMI filter’s characterization based on S-parameters The EMI filter proposed by synthetic systems for conducted EMI noise suppression should be evaluated. Its characterization parameters are defined as CMIL, DMIL, CMRR and DMRR, which are similar to those of noise separating network. The traditional measurement approach by using spectrum analyzer (SA) and 0o/180o splitter can only achieve their magnitude. Besides, the 0o/180o splitter is imperfect, and its flaw can not be excluded by calibration of the SA. In order to solve this problem, a measurement approach for EMI filter characterization based on S-parameters is employed for the synthetic system [10]. As illustrated in Fig.4, the S-parameters of EMI filter are collected by a vector network analyzer (VNA). It can not only be used to measure their magnitude but also their phrase. Without the splitter, the negative effect from the measurement interconnects can be excluded by calibration of the VNA. The circuital model of EMI filter is characterized by its S-parameter matrix [S], and the modal one is characterized by its Sparameter matrix [SM]. Their relationship can be expressed as[11]

[ SM ] = [ B ][ S ][ A]

−1

where the conversion matrices [A] and [B] are

(3)

1596


Fig. 4. test setup of the measurement approach for EMI filter characterization based on S-parameters

1 1 0  1 1 −1 0 [ A] = [ B ] =  2 0 0 1   0 0 −1

0 0  1  1

(4)

It is indicated from their definitions that CMIL, DMIL, CMRR and DMRR are S-parameters in matrix [SM].The matrix [S] of EMI filter can be directly measured by a VNA, therefore, [SM] is obtained by equation(3). The characterization of EMI filter proposed by synthetic system can be evaluated by the measurement approach based on S-parameters. 4. Applications 4.1. Experiment #1: the synthetic system applied for conducted EMI noise analysis and suppression of SMPS A commercial SMPS (Keri-Com, AC-DC, 24V/960W) with a 100Ω load is connected to LISN as EUT to the synthetic system, as shown in Fig.1.(a). Its noise analysis and suppression process performed by the synthetic system are described in three major software interfaces. • The noise measurement and analysis interface is shown in Fig. 5. (a). It can be seen that the total noise of the SMPS displayed from 10 KHz to 30 MHz on the interface can not meet the FCC standards shown in the redline. The features of total noise are shown on the right side of interface by the start point (0.00MHz, 71.74dBuV), the stop point (22.42MHz, 70.08dBuV) and the peak point (2.89MHz, 87.13dBuV) of the overshoot noise. These are collected for EMI filter design. • The EMI filter design interface is shown is Fig. 5. (b). Certain parameters should be selected for filter design. The filter type is selected as the CM filtering, its pattern is selected as L configuration, the save margin is selected as 3dB and its inductance L is selected as 0.1mH. Based on these selected information and the noise feature collected in the noise measurement and analysis interface, the filter capacitance C is calculated as 1.30756uF. • The noise analysis and suppression report interface is shown is Fig. 5. (c). Some basic information are printed in it, including the test date (April 12th, 2009), the selected EMC standard (FCC standard) and the information whether the noise complies with the EMC standard or not (fail). The EMI filters proposed as the noise suppression tool by the synthetic system are presented in Fig. 5. (c), the CM filter is in L configuration with a 0.1mH inductance L and a 1.30756uF capacitance, and the DM filter is in π configuration with a 0.215mH inductance L and two21.324uF capacitances. The noise suppression result predication is shown for the original total noises and the filtered total noise as well as their compliance with EMC standard. The maximum noise without filter shown in green line is 80dBuV at 12MHz, exceeding the FCC standard line, and the maximum noise with filter shown in purple line is no more than 48dBuV at 12MHz, reduced by 76%. This is an ideal noise suppression result of the proposed filter. 4.2. Experiment #2: implementation of EMI filter proposed by the synthetic system In order to verify the efficiency of the EMI filter proposed by synthetic system, the EMI filter designed for the commercial SMPS mentioned above is made and implemented between LISN and EUT. The same commercial SMPS adopted in experiment #1 is used as EUT, its total noise without and with the filter between 10 KHz to 30MHz are shown in Fig. 6.(a). The maximum total noise can reach 75dBuV at 6MHz before filtered. With the filter implemented, the total noise decreases in the whole frequency band, and its noise peak is 60dBuV at 6MHz, reduced by 25%, as shown in Fig. 6. (b). Besides, at 17MHz, it declines from 64dBuV to 40dBuV, reduced by 60%. Although the test result of the proposed filter is not as good as the ideal one mentioned above, it can be seen that the total noise of the SMPS can still meet the FCC standard after filtered. It is verified from this experiment that the synthetic system can provide an efficient noise suppression design for EUT.


1597

Fig. 5. (a) noise measurement and analysis interface; (b) EMI filter design interface; (c) noise analysis and suppression report interface

Fig. 6. (a) Total noise of a SMPS without EMI filter proposed by synthetic system; (b) total noise of a SMPS with EMI filter proposed by synthetic system

5. Conclusion A synthetic system for conducted EMI noise analysis and suppression is proposed in this paper. The system hardware and software design scheme and its major techniques are discussed. Moreover, the system is applied to analyze the conducted EMI noises of a SMPS, and EMI filter proposed by the system proved to be able to suppress the unwanted noise of the SMPS efficiently. It is indicated by the applications that the proposed system is an efficient tool for noise analysis and suppression.

Acknowledgements This work is supported by Natural Science Foundation of Jiangsu Province (#BK2008429), Open Research of Key Laboratory of Millimeter waves (#K200803), China Post-doctoral Research (#20080431126).

References [1] H. Chen, Implementation of a Three-Phase Switched Reluctance Generator System for Wind Power Applications. Proceedings of IEEE 14th International Symposium on Electromagnetic Launch Technology, 489-494 (2008). [2] H. Chen, X. Zan, L. Cheng, L. Guo, J. liu. Switched Reluctance Starter/Generator System in Low Voltage. Proceedings of 11th International Conference on Electrical Machines and Systems, 1077-1081 (2008). [3] H. Chen, L. Cheng, X. Qiu, Y. Zhao, Conductive EMI Noise Measurement for Switched Reluctance Drive. Proceedings of 20th International Zurich Symposium on Electromagnetic Compatibility, 241-244 (2009). [4] H. Chen, X. Wang, X. Zhou, L. Guo, L. Cheng, Switched Reluctance Machine System in High Speed Operation. Proceedings of 2009 IEEE International Conference on Networking, Sensing and Control, 97 –101 (2009). [5] H. Chen, X. Wang, H. Zeng. Electromagnetic Design of Switched Reluctance Linear Machine. Proceedings of 6th International Power Electronics and Motion Control Conference, 836-840 (2009). [6] C.R. Paul, Diagnosis and reduction of conducted noise emission, IEEE Trans EMC, 4 (1988) 553-560. [7] M. Mardiguian, Raimbourg, An alternative method for characterizing EMI filter, Proc of IEEE on EMC, 2 (1999) 882-886. [8]S. Wang, Fred. C. Lee, Characterization, Evaluation, and Design of Noise Separator for Conducted EMI Noise Diagnosis, IEEE transactions on Power Electronics, 4 974-982. [9] K.Y. See, J. Deng, Measurement of noise source impedance of SMPS using a two current probes approach, IEEE transactions on Power Electronics, 3 (2004) 862-868.

1598

Z. Yang et al. / Procedia Earth and Planetary Science 1 (2009) 1593–1598 [10] J.R. Regue, M. Ribo, Common and Different Mode Characterization of EMI Power-Line Filters from S-parameters Measurements, Proc. of IEEE trans on EMC, 2 (2004) 610-615. [11] T. Guo, D. Chen, Separation of the common-mode and differential-mode conducted EMI noise, IEEE Trans. on Power Electronics, 3 (1996) 480 – 488. [12] Y. Zhao, K.Y. See, Fundamental of Electromagnetic Compatibility and Application, China Machine Press, 2007.