1、AEROSPACE INFORMATION REPORTAIR1700REV.AIssued 1986-01Revised 1998-12Reaffirmed 2009-11Upper Frequency Measurement Boundary forEvaluation of Shielding Effectivenessin Cylindrical SystemsFOREWORDChanges in this revision are format/editorial only.INTRODUCTIONElectrical energy leaks through a shield bo
2、th by diffusion, as Schelkunoff (Reference 1) has described; and by aperture coupling, as described by Marcuvitz (Reference 2). At low frequencies, diffusion dominates, and at high frequencies, aperture coupling dominates. A convenient way to measure this leakage is Transfer Impedance, Zt, and Trans
3、fer Admittance, Yt, because they characterize the shield leakage independent of the source (or object) of the interfering signal, and because they are also independent of the number and arrangement of conductors inside the shield and their termination. In other words, Ztand Ytare characteristics of
4、the shield alone.Since sources and receivers of electrical energy cover a frequency range of DC to over 100 GHz, shielding is likely to be needed over that range also. If shielding is required, some means of measuring its Ztand Ytare necessary. A number of methods have been proposed (References 3 to
5、 11), most of which are variations of the method of Zorzy and Muehlberger (Reference 12). Most of these methods have been limited to frequencies below 1 GHz, some considerably below. The limitation has in some cases (References 3 and 10) been instituted to simplify the data analysis, although ways a
6、round this difficulty have been suggested (References 4 and 8).Aside from limits of convenience, a transition frequency, ft, exists, beyond which the variance in measurements of Ztand Ytcan become large. The existence of this transition frequency has been recognized in military specifications (Refer
7、ences 5 and 6), where Ztmeasurements are terminated before the transition frequency is reached. In the H revision of MIL-C-38999 (Reference 6), a stirred mode technique (Reference 13) is called out for frequencies over 1 GHz, which generally includes frequencies above the transition frequency. Howev
8、er, stirred mode testing does not evaluate Zt. Methods which do not evaluate Ztor Ytbring into the measurement both the nature of the source (or object) of interference and the nature of the conductor arrangement inside the shield, as well as the SAE Technical Standards Board Rules provide that: “Th
9、is report is published by SAE to advance the state of technical and engineering sciences. The use of this report is entirely voluntary, and its applicability and suitability for any particular use, including any patent infringement arising therefrom, is the sole responsibility of the user.” SAE revi
10、ews each technical report at least every five years at which time it may be reaffirmed, revised, or cancelled. SAE invites your written comments and suggestions. Copyright 2009 SAE International All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or trans
11、mitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE. TO PLACE A DOCUMENT ORDER: Tel: 877-606-7323 (inside USA and Canada) Tel: 724-776-4970 (outside USA) Fax: 724-776-0790 Email: CustomerServicesae.org SAE WE
12、B ADDRESS: http:/www.sae.org SAE values your input. To provide feedbackon this Technical Report, please visit http:/www.sae.org/technical/standards/AIR1700ASAE AIR1700 Revision A- 2 -method of terminating the conductors. Such measurements are situation specific, i.e., they evaluate the shielding in
13、the configuration of, and under the conditions of, the measurement but are not generally transportable to other conditions and configurations. A non-transportable measurement may be suitable for engineering reference or to evaluate an assembled system, but it is not suitable as an acceptance test on
14、 a system component.1. SCOPE:This AIR points out that above a frequency called the “transition frequency,” variances associated with the shielding effectiveness measurements can become large. It includes the derivations to demonstrate this. This fact should be taken into account when designing shiel
15、ding for use above the transition frequency.2. REFERENCES:1. Schelknuoff, S. A., “The Electromagnetic Theory of Coaxial Transmission Lines and Cylindrical Shields,” Bell Syst Tech J., Vol 13, 532-579, Oct. 1934.2. Marcuvits, N., Waveguide Handbook, New York, McGraw-Hill, 1951.3. Publication 96-1A, I
16、nternational Electrotechnical Commission, Geneva, Switzerland, 1976.4. Madle, P. J., “Cable and Connector Shielding Attenuation and Transfer Impedance Measurements Using Quadraxial and Quintaxial Test Methods,” In IEEE 1975 Electromag. Compat. Symp. Record, 75CH1002-5 EMC, pp. 4B1b1-5.5. MIL-C-39012
17、B Amendment 1, Military Specification, Connectors, Coaxial, Radio Frequency, General Specifications for, 18 Oct. 1972.6. MIL-C-38999H Amendment 1, Military Specification, Connectors, Electrical, Circular, Miniature, High Density, Quick Disconnect, Environment Resistant, Removable Crimp Contacts.7. M
18、artin, A. R. and S. E. Emert, “Shielding Effectiveness of Long Cables,” in IEEE 1979 Int. Symp. on Electromagnetic Compatibility, San Diego, California, Oct 9-11, 1979, pp. 13-18.8. Martin, A. R. and S. E. Emert, “The Shielding Effectiveness of Long Cables, II: LTand GTR,” IEEE Trans. on Electromag.
19、 Compat., Vol. EMC-22, pp. 269-275, Nov. 1980.9. Merewether, D. E., and T. F. Ezell, “The Effect of Mutual Inductance and Mutual Capacitance on the Transient Response of Braided-shield Coaxial Cables,” IEEE Trans. on Electromag. Compat., Vol EMC-18, pp. 15-20, Feb 1976.10. Oakley, R. J. “Surface Tra
20、nsfer Impedance Measurements - a Practical Aid to Communication Cable Shield Design,” In 18th Intl Wire and Cable Symp., Atlantic City, NJ, Dec. 1969.11. Knowles, E. D., and L. W. Olson, “Cable Shielding Effectiveness Testing,” IEEE Trans. on Electromag. Compat., Vol EMC-16, pp. 16-23, Feb 1974.SAE
21、AIR1700 Revision A- 3 -2. (Continued):12. Zorzy, J., and R. F. Muehlberger, “Rf Leakage Characteristics of Popular Coaxial Cables and Connectors, 500MC to 7.5GC,” Microwave J., Nov. 1961.13. MIL-STD-1344A Notice 1, Method 3008, Military Standard Electrical Connectors, Test Methods for.14. Grivet, P.
22、, The Physics of Transmission Lines at High and Very High Frequencies, Vol. 1, New York, Academic Press, 1970.15. IEEE Standard Dictionary of Electrical and Electronic Terms, IEEE Std 100 - 1972, Definition #E151-0.16. Vance, E. F., “Shielding Effectiveness of Braided-Wire Shields,” IEEE Trans. on E
23、lectromag. Compat., Vol. EMC-17, May 1975, pp. 71-77.17. Luca, V. A., and P. I. Pressel, “The Radiation Efficiency Technique for Measuring the Shielding Effectiveness of Cylindrical Connectors at Microwave Frequencies,” In Proc. 13th Annual Connector Symp, Philadelphia, PA, 1980, pp. 366-377.3. THE
24、EXISTENCE OF A TRANSITION FREQUENCY:Any system of two or more conductors can form a transmission line. As long as the resistance of the conductors is low (e.g., a metal) and the electrical energy flowing on the line propagates in the TEM mode, the electric and magnetic field vectors are both contain
25、ed in a plane perpendicular to the conductors, and current flow is along the axis of the conductors.Now let us consider a two conductor transmission line in which the outer conductor surrounds the inner one. This arrangement is often encountered in Ztmeasurements, where the inner conductor is a shie
26、ld under test, and the outer conductor is used to establish a known field. The conductors of this line can have any shape in cross-section but are restricted to having the same cross-section throughout their entire length.1Let us examine current flow on the outer conductor. For simplicity, we will a
27、ssume that the conductor is a circular cylinder, although the following discussion applies equally well to a conductor of irregular cross-section. As we have said previously, the flow of current on this conductor is axial when the propagating mode is TEM.Now let us raise the frequency of the propaga
28、ting energy to the point where the perimeter of the conductor under consideration just equals a wavelength. This situation for a circular cylinder is illustrated in Figure 1.1. For the purpose of this discussion, a non-uniform conductor can be replaced by a uniform conductor having the largest cross
29、-section of the non-uniform conductor. Also, one of the two conductors could be geometrically complex; e.g., the skin of a vehicle, or the remaining N-1 conductors in an N conductor cable.SAE AIR1700 Revision A- 4 -FIGURE 1 - Voltage and Current Distribution on a Cylinder in a PlaneTransverse to the
30、 Axis of Propagation, When the Perimeter of theCylinder is One Wavelength Long3. (Continued):A standing wave can now exist which appears as a dipole, with a positive voltage at a, and a negative voltage at c. This dipole drives a current back and forth around the circumference of the cylinder from a
31、 to c. The vector sum of this current and the axial current is a resultant current flowing at some angle (depending on the relative strengths and phasing of the components) with respect to the Z axis. Since the current flow is no longer axial, the mode or propagation is no longer TEM, and the transm
32、ission line is said to have “moded.”The frequency at which this moding becomes possible can be calculated from the relation (Reference 14)(Eq. 1)where f = frequency in Hertz, = velocity of propagation in meters/second, and = wavelength in meters. The velocity of propagation can be calculated from th
33、e relation (Reference 14)(Eq. 2)f-=3108m/s-=SAE AIR1700 Revision A- 5 -3. (Continued):where = relative magnetic permeability of the medium surrounding the inner conductor (which is the shield being examined), and is the relative dielectric constant. For a cylinder of diameter D, the circumference is
34、 equal to a wavelength when(Eq. 3)The frequency at which the circumference (or periphery) of the conductor is just equal to a wavelength is technically defined (Reference 15) as the “cut-off frequency.” However, the use of the term “Transition Frequency” is preferred, as it is more descriptive of th
35、e process of transitioning from a frequency region where the assumption of TEM propagation is generally valid, to a region where the assumption is generally invalid. The transition frequency is designated by ftand can be found by substituting Equations 2 and 3 into Equation 1:(Eq. 4)or(Eq. 5)If D is
36、 in inches instead of meters,(Eq. 6)As an example, let us calculate the transition frequency for a cable shield having an outside diameter of 1 inch, and having a jacket with a relative magnetic permeability of 1 and a dielectric constant of 2.6. If this cable is laid on a ground plane (so that the
37、largest characteristic dimension is that of the cable itself), but is otherwise surrounded by air, the effective dielectric constant will be closer to 2. LettingD = 1, = 1, and = 2 in Equation 6, we find that ft= 2.66 GHz. If the cable is put inside a tube of 2 inches I.D. (for example, to measure i
38、ts shielding effectiveness) and assuming is still about 2, ftdecreases to 1.33 GHz, because the largest diameter in the system is now the inside diameter of the tube.D=ft3108D- Hz=ft0.095D-GHz=ft3.76D-GHz=SAE AIR1700 Revision A- 6 -4. CONSEQUENCES OF MEASURING ABOVE ft:Let us consider two cases. In
39、Case I the shield is a solid tube so that leakage of electrical energy through the shield is by diffusion only. In Case II the shield has one or more holes or cracks in it so that leakage is a combination of diffusion and aperture coupling. Case II includes among other things, braids, connector-shie
40、ld interfaces, and connectors with imperfections.In Case I the leakage through the solid shield can be calculated using Schelkunoffs (Reference 1) diffusion formula, provided that the current flow is axial only. If the current flow is other than axial, then a much more complex expression must be use
41、d to describe the shield leakage. This expression involves the relative magnitudes of the axial and circumferential currents; and since these are generally unknown, the leakage is unpredictable. Because the leakage is extremely small at the transition frequency for a tube of any practical thickness,
42、 Case I is of academic interest only.In Case II the leakage can be determined by adding the aperture leakage calculated from Marcuvitz (Reference 2) equations to the diffusion leakage calculated from Schelkunoffs (Reference 1) formula. As was pointed out in Case I, diffusion leakage is negligible at
43、 the transition frequency. Thus for Case II, leakage of the transition frequency is adequately described by aperture coupling alone.The calculation of aperture coupling is based on Marcuvitz (Reference 2) formula for the coupling between two concentric coaxial lines via an aperture. In the case of b
44、raid cables (Reference 16), it is usually convenient to consider a set of elliptical apertures with their major axes parallel to the axis of the cylinders. Assuming current flow is axial (as it would be for a TEM mode), the amount of coupling is approximately proportional to the number of apertures
45、and the cube of their minor axis. Marcuvitz formulas have been extended to non-coaxial systems, but no further insights would be gained for the purposes of this AIR. Marcuvitz work can be applied directly to single cracks or holes, such as might exist in a connector or a connector-cable interface. M
46、ultiple holes can be treated by summing vectorally the leakage from individual holes. These assumptions are valid below the transition frequency. But above that frequency we have previously shown that the current flow may not be parallel to the axis of the shield. In that case, the current flow will
47、 also not be parallel to the major axis of the braid holes; and as Marcuvitz more general analysis shows, the shield leakage will depend on the orientation of the (elliptical) braid holes with respect to the current flow. The same observation is true for a crack in a shield such as might be formed b
48、y a connector-cable interface. Since, as previously noted, the direction of current flow is not known above the transition frequency, the amount of aperture coupling, which depends on the direction of current flow relative to the axis of the apertures, is also not known. The result is the shield lea
49、kage above the transition frequency is generally not measurable in a way that yields repeatable results independent of the measurement setup. For example, above the transition frequency, Ztnumbers obtained by a vendor in his laboratory may not be the same as those obtained by a system designer, when the part is installed in his system.SAE AIR1700 Revision A- 7 -4. (Continued):Similar difficulties apply to the measurement of transfer admittance Ytwhich
