A Novel Active Common-mode EMI Filter for PWM Inverter

A Novel Active Common-mode EMI Filter for PWM Inverter Yo-Chan Son and Seung-Ki Sul School of Electrical Engineering & Computer Science #024, ENG420, Seoul National University Kwanak P.O.BOX 34, Seoul, Korea (ZIP 151-742) Abstract – This paper presents a novel active common-mode EMI filter for the PWM inverter application. The proposed filter is based on the current sensing and compensation circuit and it utilizes a fast transistor amplifier for the current compensation. T
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A Novel Active Common-mode EMI Filter for PWM Inverter Yo-Chan Son and Seung-Ki Sul School of Electrical Engineering & Computer Science #024, ENG420, Seoul National UniversityKwanak P.O.BOX 34, Seoul, Korea (ZIP 151-742)  Abstract  – This paper presents a novel active common-modeEMI filter for the PWM inverter application. The proposedfilter is based on the current sensing and compensation circuitand it utilizes a fast transistor amplifier for the current compen-sation. The amplifier utilizes an isolated low-voltage dc powersupply for its biasing and it is possible to construct the activefilter independent of the source voltage of the equipment. Thusthe proposed active filter can be used in any application re-gardless of its working voltage. The effectiveness of the pro-posed circuit is verified by experimental results. I. I NTRODUCTION The PWM inverter as shown in Fig. 1, is the inherent noisesource that makes abrupt voltage transitions (high dv/dt  )accompanied by switching actions [1, 2]. Coupled with thestray capacitance of the load machine the high frequencycurrent is generated, which can affect nearby equipments dueto the conducted and radiated electromagnetic interference(EMI) [3, 4]. As the semiconductor technology progressespower devices are getting faster to reduce the switching lossand to increase the controllability of the system, but the in-crease of  dv/dt  is accompanied by the increase of the EMIlevel. Especially the leakage current by the common-modevoltage of the PWM inverter is the primary concern of theconducted and radiated EMI. Because the high frequencyleakage current at the motor returns through the earth ground,its circulation loop is relatively large compared with that of the normal-mode current. This large circulation loop plays asa role of the antenna for the radiated EMI. These days manycountries limit the EMI emission under their regulations[3~6]. To meet such regulations many efforts to EMI mitiga-tion have been concentrated and most of them have beenrelated to the design of input/output EMI filters [1~3, 7].Most of conventional filters were passive filters. In designingpassive filters the obtainable insertion loss is limited by theavailable LC product that should be increased to meet therequirement. In case of the common-mode EMI, one cannotincrease the total capacitance for the safety reason [3] andshould increase the inductance of the common-mode chokein order to increase the insertion loss, which makes the com-mon-mode filter bulky. Many researchers have tried to de-velop some active filtering techniques in order to overcomethis limitation [8~12].In this paper, a novel active common-mode EMI filter isintroduced in order to mitigate the conducted common-modeEMI. It can provide the sufficient attenuation under the limit-ed LC product and there is a promising possibility of its ap-plication regardless of the working voltage of the system. Itsanalysis and experimental verification will be given.II. A CTIVE C OMMON - MODE EMI F ILTER Fig. 2 shows the basic concept of the proposed ActiveCommon-mode EMI Filter (ACEF). The circuit is based onthe topology using the current sensing and compensation [8,9, 11, 12]. The noise source is the PWM inverter in Fig. 2and the input filter can be an additional passive filter, whichgives additional insertion loss. The series-connected com-mon-mode choke  L CM  works as the common-mode currentsensing element by the additional winding. The high frequen-cy current that passes through  L CM  generates the high fre-quency flux in the common-mode choke, which makes thehigh frequency voltage at the input terminal of the trans-conductance amplifier. The output of the amplifier is con-nected to the output capacitor C  o that is used for the currentinjection to the earth ground. In this filter circuit, the injectedcurrent cannot be circulated within the system without usingthe coupling capacitor C  c because the closed loop cannot bemade. Thus C  c is used to provide the low-impedance path of the high frequency common-mode current for the internalcirculation. The supply voltage of the filter circuit V  c is usedto give the bias voltage to the amplifier. At low frequency theimpedance of  C  c is large enough to isolate the bias voltagefrom the main voltage.The required bias voltage of ACEF can be calculated as in(1). For example, if 0.5[  A ] of the high frequency current at    s Motor  Input  Filters PWM Inverter  LISN  1  g  i  g  i 2  g  i Fig. 1. PWM inverter system using single-phase ac input.  1[ MHz  ] should be supplied, then the bias voltage should belarger than 7.96[ V  ] in order to drive 10[ nF  ] of  C  o . ooc  fC iV  π 2max ≥ (1)Fig. 3 shows the application of the proposed ACEF. In thisexample the single-phase, three-phase and dc applications areshown. As introduced in [12], a push-pull amplifier is usedand C  c is connected to the dc-bus of the PWM inverter or anydc load system. In case of the PWM inverter supplied by thecommercial 60[  Hz  ]/220[ V  rms ] utility line, the dc-bus voltageis about 300[ V  ]. As in [11], if the push-pull amplifier shouldbe placed across the dc-bus capacitor, then each transistorshould be able to handle the full dc-bus voltage. In this casethe high-voltage/high-current  pnp -transistor is hardly avail-able, that limits its application. Moreover the required volta-ge of the push-pull amplifier is very small compared to thedc-bus voltage as calculated in (1) and it can make the ineffi-cient voltage usage. However in the proposed circuit theadditional low voltage supply can be used to drive the push-pull amplifier. This enables the use of low voltage devicesand extends the application of the proposed ACEF. Fig. 3(a)shows its application when coupling capacitors are placedbetween the dc-bus voltage V  dc and the filter supply voltage V  c . While coupling capacitors in Fig. 3(a) isolate the dc-busand the ACEF at low frequency, they make the low impedan-ce path at high frequency, which allows the internal circula-tion of the high frequency leakage current between the sys-tem and the ACEF as introduced in [12]. Coupling capacitorsalso can be connected to ac input lines of the system asshown in Fig. 3(b) and it is possible to construct a separateinput filter stage. The same idea can be extended to the 3-phase applications, which is shown in Fig. 3(c) and (d).The proposed circuit works as follows. In this analysisthere is an assumption that coupling capacitors have suffi-ciently low impedance at the frequency band of interest. If the inductance of the common-mode choke seen at the pri-mary winding is  L CM  without the consideration of the secon-dary winding, then the relation between the input common-mode current i  g  and the base current of the push-pull ampli-fier can be found as (2). 1 ω +⋅−=  s s N ii  g b and CM in  Lr  N  21 = ω ,(2)where  N  is the turn ratio between the primary winding andthe secondary winding, and r  in is the input impedance of thepush-pull transistor amplifier including the additional resistor.When the leakage inductance of the common-mode choke,the common-mode inductance  L CM  can be represented as (3). eeCM  l  A N  L µ 2 = ,(3)where µ is the relative permeability,  A e is the effective areaand l  e is the effective length of the magnetic core. Thus thecutoff frequency ω 1 can be represented as (4) regardless of the number of turns  N  . eein  Al r  µω = 1 .(4)If the bandwidth of the transistor amplifier is ω T  and theamplifier is modeled as a first-order system, then the injecti-on current i o of the amplifier can be derived as (5). Becausethe input common-mode current i  g  is the sum of the injectioncurrent i o and the load leakage current i  g  1 , then the relationbetween i  g  and i  g  1 can be found as (6).  g T  febT  feo i s s s Nhi shi 1 11 ωωω +⋅+−=+= ,(5)where h  fe is the ac current gain of the transistor. in r  o C  c C  CM   L m  g   Input  Filter  NoiseSource c V  c C   g  i o i 1  g  i  s  z  Fig. 2. Basic concept of proposed ACEF.(a) o C  CM   L c C  −+ dc V   Rectifier or Converter dc Load System c V  o i 1  g  i  g  i (b) o C  CM   L c C  ac or dc Load System c V  b i o i 1  g  i  g  i (c) o C  CM   L c C  −+ dc V   Rectifier or Converter dc Load System c V   g  i 1  g  i o i (d) o C  c C  3-phase Load System c V  CM   L  g  i 1  g  i o i Fig. 3. Configurations of proposed ACEF.(a) ACEF using dc-bus coupling, (b) ACEF using ac line coupling for sin-gle-phase application, (c) ACEF using dc-bus coupling, (d) ACEF using acline coupling for three-phase application.  ( )( ) 112111 11111  g T  feT T  g o g  i s s Nh s siii ωωωω ωω +      +++++=+= .(6)If the bandwidth of the transistor amplifier and  Nh  fe arehigh, then (6) can be simplified as (7). ( )( ) 1321 1111  g T  g  i s s s si ωωωω ++++≈ ,(7)where ω 2 = ω 1  /(1+  Nh  fe ) and ω 3 =  Nh  fe ω T  . If  ω T  is muchhigher then ω 1 then (7) is simplified as (8). 121 11  g  g  i s si ωω ++≈ ,(8)Fig. 4 shows the frequency response of the proposedACEF without including the effect of coupling capacitors.Parameters used in Fig. 4 are listed in Table I. As can be seenfrom the figure, the approximation (8) holds in the conductedEMI frequency band (0.15 ~ 30[ MHz  ]). But the attenuationperformance is degraded in the radiated EMI frequency band(30 ~ 300[ MHz  ]) due to the limited bandwidth of the tran-sistor amplifier. The maximum attenuation of 1/(1+  Nh  fe ) isachieved between ω 1 and ω T  . Simple passive filters can beinserted at the input or output of the transistor amplifier toreduce the effect of the limited performance in the radiatedEMI frequency band. Also as concerned with the inflow of the low-frequency leakage current, it cannot be attenuatedbecause of the nature of the sensing circuit of (2).The equivalent inductance of the common-mode chokedue to the ACEF can be calculated using the relation shownin (2) and it is given as (9). Combined with (7) and (9), theinput common-mode current is effectively suppressedwithout increasing the low frequency leakage current. 1 11' ω  s L L CM CM  +⋅= .(9)III. E XPERIMENTS In this section, several experimental results are shown.First, Fig. 5 ~ Fig. 9 show the evaluation of the proposedACEF using ac line coupling for a single-phase applicationand Fig. 10 shows the evaluation of the proposed ACEF for athree-phase application using ac line coupling.Fig. 5 shows the configuration of the experimental system.Detailed parameters of the experimental system are listed inTable I. A 3.7[kW] induction motor is used as a load machineof the PWM inverter and a dc smoothing reactor is used inorder to reduce the input harmonic current. An input filterprinted circuit board (PCB) is used separately from the PWMinverter PCB. The input filter is composed of the ACEF andadditional passive filtering elements. A single-phase LISN isused to provide the stable source impedance at the high fre-quency as shown in Fig. 1 and the peak detector is used in themeasurement of the conducted EMI spectrum [3~5].Waveforms of leakage currents and the conducted EMIspectrum of the system without any EMI filter are shown inFig. 6. Because there is no Y-capacitor, the motor leakagecurrent generated by the switching of the PWM inverter isdirectly reflected to the input common-mode current asshown in Fig. 6(a). The conducted EMI spectrum includesboth of the common- and normal-mode EMI. Although theyshould be separately considered, the normal-mode EMI willnot be discussed in this paper with the assumption that someappropriate normal-mode filtering elements are installed foreach design stage. After sufficient normal-mode filtering, thecommon-mode EMI plays as the bottleneck of the total con-ducted EMI [12, 13].Fig. 7 shows leakage current waveforms and the conductedEMI spectrum when the proposed ACEF is added in the sys-tem of Fig. 6. Low-voltage complementary transistors ( V  CEO-max = 30[ V  ]) are used in this filter circuit and a 12[ V  ] dc pow-er supply is used as the bias voltage V  c . Actually this voltageis coming from the control power supply in the inverter sys-tem. In this experiment C   x 1 and C   x 2 are installed along withthe ACEF. Using parameters listed in Table I, corner fre-quencies, ω 1 and ω 2 , can be calculated as 11.2[ MHz  ] and14[ kHz  ] respectively when r  in and h  fe are assumed to be  s 3.7kW  InductionMotor  o C  c C  c V  o i 1  x C  2  x C  2  y C  1  y C  1 CM   L 2 CM   L  AC Source 1 Q 2 Q  PWM  Inverter  Input Filter  1  g  i  g  i Fig. 5. Configuration of experimental system. Frequency (Hz)     M   a   g   n     i    t   u     d   e    (     d    B    ) -60-50-40-30-20-10010 4 10 5 10 6 10 7 10 8 ConductionRadiation Fig. 4. Frequency response of proposed ACEF.  100[ Ω ] and 100, respectively. Between ω 2 and ω 1 , the pro-posed ACEF attenuates the conducted EMI according to (7)as shown in Fig. 4. Above ω 2 the attenuation by the proposedACEF is effective as shown in Fig. 7(b). Most of the highfrequency leakage current i  g  1 from the motor is absorbed bythe proposed ACEF and only the low frequency ripple can beseen from the waveform of  i  g  as shown in Fig. 7(a). Conse-quently the high frequency input common-mode current ismuch reduced compared with that of Fig. 6(a). Because itspeak magnitude is considerably reduced and this may helppreventing the saturation of the magnetic core in the addi-tional filter, which helps improving the performance of theadditional filter stage. Thus one can get the enhanced attenu-ation performance of the additional filter, or make the size of the additional magnetic core smaller.In Fig. 8 an additional filter stage is added to the system of Fig. 7. In Fig. 8(a) an additional Y-capacitor C   y 2 is installedat the input terminal along with an additional common-modechoke  L CM  2 . With the aid of the additional filter stage theconducted EMI is suppressed about 20[ dB µ V  ] until 2[ MHz  ]but the level of the EMI spectrum is increased especiallyabove 2[ MHz  ]. Although the entire level is placed under thegiven limit line it is not favorable to have the small margin.In Fig. 8(b) C   y 1 is installed instead of  C   y 2 . The entire EMIlevel is much attenuated and the conducted EMI above5[ MHz  ] is not increased as in the case of Fig. 7(a).The proposed ACEF was adopted for the commercial airconditioner and Fig. 9 shows its evaluation result. The circuitof Fig. 3(a) was used in this experiment. Fig. 9(a) shows theconducted EMI spectrum of the system with 2 stage cascadedpassive filters. An ACEF was implemented using the existingcommon-mode choke of filter stages of Fig. 9(a) and theresult is shown in Fig. 9(b). As can be seen from Fig. 9, theEMI spectrum of 150 of 500[ kHz  ] is much attenuated so it is (a) 1  g  i  g  i (b)Fig. 6. No EMI filter installed.(a) Leakage currents (200[ mA/div ], 1[ µ  s/div ]), (b) conducted EMI spectrum.(a) 1  g  i  g  i (b)Fig. 7. Proposed ACEF installed.(a) Leakage currents (200[ mA/div ], 1[ µ  s/div ]), (b) conducted EMI spectrum.(a)(b)Fig. 8. Additional passive EMI filter installed.(a) C   y 2 installed, (b) C   y 1 installed.(a) Quasi-peak Measurement  Average Measurement  (b) Quasi-peak Measurement  Average Measurement  Fig. 9. Application example of ACEF for commercial air conditioner.(a) 2 stage passive filters. (b) 1 passive filter + 1 ACEF.
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks