1、ANSI C63.13-1991American National Standard Guide on the Application and Evaluation of EMI Power-Line Filters for Commercial UseAccredited Standards Committee On Electromagnetic Compatibility, C63accredited by theAmerican National Standards InstituteSecretariatInstitute of Electrical and Electronics
2、Engineers, Inc.Approved June 28, 1991American National Standards InstituteAbstract: A basic understanding of the application, evaluation, and safety considerations ofelectromagnetic interference (EMI) power-line filters used in both ac and dc applications is provided. Theconstruction of an EMI power
3、-line filter and its functions in providing suppression of conducted noise aredescribed. The functions and performance of the filter components, particularly the capacitors andinductors, are discussed. It is explained why seemingly identical filters may not give the same performancein a particular a
4、pplication. No-load insertion-loss test methods are presented. Proper installation of thefilters in equipment is discussed. Safety regulations are briefly addressed.Keywords: capacitors, common-mode noise currents, differential-mode noise currents, electromagneticinterference power-line filters, ind
5、uctors, no-load insertion lossThe Institute of Electrical and Electronics Engineers, Inc.345 East 47th Street, New York, NY 10017-2394, USACopyright 1991 by the Institute of Electrical and Electronics Engineers, Inc. All rights reserved. Published 1991.Printed in the United States of America.ISBN 1-
6、55937-138-2No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without theprior written permission of the publisher.American National StandardAn American National Standard implies a consensus of those substantially concerned with its scope and p
7、rovisions.An American National Standard is intended as a guide to aid the manufacturer, the consumer, and the general public.The existence of an American National Standard does not in any respect preclude anyone, whether he has approved thestandard or not, from manufacturing, marketing, purchasing,
8、or using products, processes, or procedures notconforming to the standard. American National Standards are subject to periodic review and users are cautioned toobtain the latest editions.CAUTION NOTICE: This American National Standard may be revised or withdrawn at any time. The procedures ofthe Ame
9、rican National Standards Institute require that action be taken to reaffirm, revise, or withdraw this standard nolater than five years from the date of publication. Purchasers of American National Standards may receive currentinformation on all standards by calling or writing the American National S
10、tandards Institute.iiiForeword(This Foreword is not a part of ANSI C63.13-1991, American National Standard Guide on the Application and Evaluation of EMIPower-Line Filters for Commercial Use, but is included for information only.)Many individuals have expressed the need for a document that provides
11、a guide to the basic understanding of EMIpower-line filters.This guide is intended to assist users and designers of electronic equipment in the proper design, construction,installation, and evaluation of power-line filters. Safety considerations are also discussed.At the time this standard was appro
12、ved, the working group that prepared this standard had the following membership:Stephen D. Bloom, Chair Alphonse A. ToppetoAt the time that Accredited Standards Committee C63 approved this standard, Subcommittee One had the followingmembership:Donald N. Heirman, Chair James H. AllenWallace E. AmosDa
13、nte AnoiaStephen D. BloomEdwin L. BronaughW. F. BryantJoseph E. Butler, Jr. Stephen CaineEdward W. ChapinDave CofieldLouis B. CostelloMyron L. CrawfordGlen DashHector DavisRobert H. DavisBill DeveyDennis DzierzawskiFred FriedmanThomas GardnerFrank E. GarlingtonHarold A. Gauper, Jr.William K. HayesJa
14、mes S. HillH. R. HofmannKeith KalanquinAlbert R. KallT. W. KernWarren A. KesselmanJames C. KloudaEugene D. KnowlesNestor KolcioWilliam S. LambdinAl R. LavisJohn LichtigSiegfried LinkwitzRichard MasseyHerbert K. MertelWalter A. PoggiWilliam T. RhoadesPaul RuggeraTerence RybakRichard B. SchulzDavid A.
15、 SegersonNeal H. SheperdRalph M. ShowersLouis SlesinAlbert A. Smith, Jr.Dave StaggsLeonard W. Thomas, Sr.Eb M. TingleyAlphonse A. ToppetoAnatoly TsaliovichArt WallStan XavierAt the time that the Accredited Standards Committee on Electromagnetic Compatibility, C63, approved this standard,it had the f
16、ollowing membership:Ralph M. Showers, Chair Edwin L. Bronaugh, Vice Chair Susan L. Vogel, Secretary ivOrganization Represented Name of RepresentativeAeronautical Radio, Inc Kendall SimmonsAmador Corporation Dan HoolihanJames JohnsonAmerican Council of Independent Laboratories William K. HayesAssocia
17、tion of American Railroads Chris AllmanAssociation of Telecommunications Attorneys Jon CurtisGlen DashAT whereas, in the other line,4 Copyright 1991 IEEE All Rights ReservedANSI C63.13-1991 GUIDE ON THE APPLICATION AND EVALUATION OFit is one-half their difference. This explains why conducted noise m
18、easurements must be made on each line rather thanon only one.6. Operation of an EMI Power-Line FilterThe basic functions of an EMI power-line filter in suppressing noise currents can now be deduced by referring to Figs1 and 2.Consider common-mode suppression first. Since the common-mode currents are
19、 identical in both lines with respect toground, the line-to-line capacitors, C1, will have no effect. The common-core inductors, L1, are wound with identicalwindings having the polarity shown in Fig 1. The reason for this type of winding is to minimize the size of the corenecessary to avoid magnetic
20、 saturation by the differential power currents while providing relatively large values ofinductance.Figure 3 Phasor Diagram of Noise CurrentsFigure 4 Common-Mode Noise Equivalent CircuitCopyright 1991 IEEE All Rights Reserved 5EMI POWER-LINE FILTERS FOR COMMERCIAL USE ANSI C63.13-1991Figure 5 Differ
21、ential-Mode Noise Equivalent CircuitFinally, it is clear that the line-to-ground capacitors will be effective against common-mode noise. These capacitorsbeing connected to ground are limited by safety regulations to be very small in value compared to the line-to-linecapacitors. Therefore, the equiva
22、lent common-mode circuit of the filter in Fig 1 is that shown in Fig 4, where L2inductors are neglected in comparison to L1.The two independent inductors, L2, will also attenuate common-mode noise. However, since each of these inductors isseparately energized by the power frequency current, which is
23、 much higher than the noise currents, their value is muchless than L1in order to prevent core saturation at the rated current of the filter.In Figure 4, the equivalent value of line-to-ground capacitance is 2C2because these capacitors are in parallel for thisnoise mode. The equivalent inductance is
24、approximately L1rather than 2L1, however, because of the common corenature of these coils.The equivalent differential-mode circuit for the filter of Fig 1 may be determined in a similar manner. The line-to-linecapacitors are clearly effective against such noise, and the independent inductors, L2, ar
25、e also active in this case.However, the coupled inductors, L1, are not effective per se because of their polarity, which causes them to cancelaround the power loop in which the differential-mode current flows. The series combination of the line-to-groundcapacitors also provides a path for differenti
26、al-mode current, but again they are very small with respect to the line-to-line capacitors. Their effect at high frequencies can be significant; so, they are included in Fig 5 along with the leakageinductances, Ll, of the two L1s, which do not cancel because they are not coupled.7. Understanding the
27、 Components of an EMI Power-Line FilterLine-to-line capacitors are usually of the metallized film or film/foil structure. Such capacitors have a relatively highcapacitance, e.g., 0.1 m F to 2.0 m F, along with high reliability. These capacitors typically have a self-resonantfrequency of the order of
28、 1 or 2 MHz. Therefore, they are most effective against lower frequency differential-modenoise.Line-to-ground capacitors, as mentioned earlier, must have a very low capacitance for safety reasons. Such capacitorsare usually in the range of 0.001 to 0.01 m F, for example. For this reason, they are mo
29、st effective against higherfrequency common-mode noise. Their structure is often of the ceramic type because of the superior high self-resonantfrequency of ceramic capacitors when compared to wound film capacitors. Ceramic capacitors with very short leadswill resonate at 50 MHz or more, depending on
30、 their geometry. Film-type capacitors, even with short leads, willresonate at frequencies of the order of 10 MHz.6 Copyright 1991 IEEE All Rights ReservedANSI C63.13-1991 GUIDE ON THE APPLICATION AND EVALUATION OFAny capacitor at frequencies higher than its self-resonant frequency behaves as an indu
31、ctor and is, therefore, no longereffective as an EMI power-line filter component. This fact is important in selecting the type of capacitor and its methodof assembly into the filter.Similarly, inductors are not purely inductive. The windings by their very nature will be shunted by distributedcapacit
32、ance. Therefore, inductors too suffer from self-resonant characteristics. Above their self-resonant frequency,the capacitance dominates so that inductors lose their effectiveness at higher frequencies. Depending on the value ofthe inductance, the geometry of the windings, and the core material, coil
33、 self-resonance typically may occur in therange of 150 kHz to 2 MHz.It is quite clear that the design of independent inductors such as L2in Fig 1 must take into account the saturationcharacteristics of the core material due to the current rating of the filter and the turns required en that core to a
34、chievethe desired inductance. Otherwise, the core would be saturated under normal operating conditions and would beineffective as a filter component.It is not so clear, however, that saturation is a very important consideration in the design of common-core inductors.The two windings of such a compon
35、ent are designed with an equal number of turns so that the magnetomotive forcearound the core due to the power frequency current in these windings cancels. In Fig 6, it may be seen that the netmagnetomotive force around the core is zero because of the cancellation of the ampere-turns (NI) associated
36、 with eachwinding. Therefore, if one stops at this point, one will conclude that such a coil structure will never saturate. This is notthe case, however, because of leakage flux, which is not coupled from one winding to another. These independentfluxes can cause the core material to saturate in the
37、regions where they exist. This saturation, even though it islocalized, will have the same effect as the introduction of a large air gap in the core. That is, the inductance of thewindings will decrease dramatically.Figure 6 Common-Core InductorAn EMI power-line filter will most often contain a bleed
38、er resistor to discharge the line-to-line capacitors when poweris disrupted. It has no effect on filter performance.If a ground choke is included in the filter, it will contribute to suppression of common-mode noise, not differential-mode noise.Copyright 1991 IEEE All Rights Reserved 7EMI POWER-LINE
39、 FILTERS FOR COMMERCIAL USE ANSI C63.13-19918. Why Similar EMI Power-Line Filters May Not Perform in the Same WayAt this point, several reasons why two supposedly identical filters from two different manufacturers can be expected toperform differently in the same application become apparent.Inductan
40、ce and capacitance value measurements are normally made at very low values of voltage, current, andfrequency using a bridge instrument. The usual measurement frequency is 1 kHz whereas the frequency range overwhich performance is evaluated in an application is up to and above 30 MHz. Therefore, no i
41、nformation as to the self-resonant frequency of the components is noted.In the case of inductors, no information about the current at which the inductors, both common core and independentcore, will saturate has been gained by the low current measurements. Therefore, depending on the peak currentsthr
42、ough the filters in a particular application, the cores in one may saturate before those in the other, rendering the onewith saturated cores ineffective.Other, more subtle differences between two supposedly identical filters exist and cause significant differences in theirperformance. These effects
43、are parasitic in nature and are not detected by the usual element value measurements. Thereis capacitance within the coil winding, the distributed capacitance, which determines the self-resonant frequency of thecoils. The self-resonant frequency of the capacitive branches is not found in the low-fre
44、quency measurements.The foregoing examples have been implied in previous sections of this document; however, there are others.Core losses of the magnetic pieceparts used in a filter are significant in the neighborhood of coil self-resonance. Highercore losses at this frequency will decrease the atte
45、nuation of the filter in an application in the vicinity of coil resonance;the variation of inductance versus frequency also has not been noted.Above coil self-resonance, the distributed capacitance of the coil forms a capacitive divider with the line-to-groundcapacitors. The higher the value of the
46、distributed capacitance, the lower the attenuation of the filter will be.Other parasitic, capacitive coupling will exist between coils and other coils, between coils and capacitors, and betweeninput and output capacitors. These parasitic capacitances all serve to degrade performance of the filter.Mo
47、st filters are potted in order to stabilize the mechanical positions of the components and to provide better thermalconductivity for heat dissipation. Since the potting material must be an insulator, all parasitic capacitances areincreased by the dielectric properties of the particular material.Magn
48、etic couplings also exist to reduce filter performance. Transformer coupling arises between the input and outputterminals of a filter as a function of conductor routings within the filter. Magnetic coupling will also exist between thecoil structures and the metallic enclosures of the filter.In summa
49、ry, many parasitic parameters exist in any filter that are not easily determined by measurements. All of these,plus the properties of the materials in the components, will very likely make two apparently identical filters behavedifferently in any given application.Finally, it must be remembered that insertion-loss data on filters are most often obtained from tests in a 50 W system.Since the impedance presented to the filter at its load terminals in a particular application is unlikely to be 50 W overthe entire frequency range, and since it likely is not purely resistive, th
