Screening Attenuation of Long Cables

Transcription

1 Screening Attenuation of Long Cables Carl W. Dole, John W. Kincaid Belden Electronics Division Richmond, Indiana Abstract The characteristics of a triaxial test fixture, which has been developed for screening attenuation measurements on long shielded cables, are described. Screening attenuation performance of a selection of coaxial CATV cable braid and multi-foil/braid shield designs are presented. Measured screening attenuation results for eter long cables covering the frequency range - 00 MHz are compared with results from measurements on.5 meter long cables that were measured over the frequency range of -000 MHz. The comparison shows that the estimation or deduction of long length, low frequency performance from high frequency performance can have significant error. The paper concludes that worse case or minimum screening attenuation performance should be assessed for cable lengths and frequencies that are pertinent to usage of the cable. Keywords Attenuation; braid; CATV; coaxial; effectiveness; foil; IEC; screen; SCTE; shield; test. Introduction Screening attenuation is a practical shield effectiveness parameter, and the test method is being standardized in IEC 696- [], pren and SCTE test method IPS-TP-3B []. The test was originally developed for short length copper coaxial cables involving a resonant triaxial test fixture length of - meters [3], and the frequency range of interest was 00 MHz to 3 GHz. The minimum screening attenuation corresponds to the condition of maximum power transfer that occurs at resonant frequencies within the test fixture. With a.5 meter long fixture, the lowest resonant frequency is about 00 MHz. Fixture length and the fixture TEM mode cut-off frequency primarily determine the frequency range. A complete test system is available commercially [4], and test fixture implementations involving standard commercially available components for fixture lengths of.5 and 6.7 meters have been reported [5,6]. Preliminary results obtained with a eter long fixture have also been reported [7]. Applications such as CATV return path and data I/O have stimulated interest in screening attenuation performance where the frequency may be as low as 5 MHz and cable lengths are tens of meters long. However the chief obstacle to implementing the screening attenuation test at these frequencies is the required fixture length. A 6.7 meter length is limited to about 0 MHz, and a eter length is required to arrive at about 4 MHz, which corresponds to about one-half wavelength electrical length in the sample under test. Section of this paper presents the mechanical, electrical and installation aspects of the triaxial fixture. Sections 3 and 4 cover the test procedure and the test program, including samples tested and results. The conclusions are given in section 5.. Triaxial Test Fixture. Overview The triaxial test fixture is sketched in Figure. The main components are five sections of six-meter length, 76.5 mm diameter rigid transmission line segments (without center conductor and end disc supports) that are cascaded together with gas tight coupling flanges. The segment joints are fitted with metallic spacer rings to maintain 50 ohm impedance across the joint. A continuous length of low loss dielectric tubing material is used to support the sample under test from end to end. The tubing is centered with disc spacers that are randomly located along the eter length. An end plate closes the housing at one end and an end cap transition encloses the other. A feed-through connector is mounted in the center of the end plate. The end cap is fitted with a 50-ohm type N jack connector. The center pin of the N jack connects to an anchor connector socket. The sample termination shield is made from a split anchor connector that is mated to the end cap anchor connector socket. One side of the split anchor connector is fitted with a feedthrough connector, and the mated pair encloses and shields the cable sample termination. This arrangement provides for terminating the cable sample under test with a shielded resistive load and for connecting the shield to the center pin of the N connector. The cable sample under test is fitted with a plug type connector at each end and is connected between the feed-through connectors, respectively located in the end plate and the split anchor connector.. Electrical Characteristics The shield of the cable under test (CUT) defines the boundary between the two electrical regions within the fixture. As shown in Figure, these are () the resonant region, which is located between the rigid metallic cylindrical housing and the cable sample shield, and () the cable sample (CUT). The diameter ratio of the housing and sample shield diameters is important for determining the impedance and percent velocity of the resonant region, which are approximately 50Ω and 90% respectively.

2 End Plate End Cap 0 Cascaded Test Fixture (5 x 6m) Dielectric tube support measured theoretical Figure 3. RL of test chamber: measured and theoretical Dielectric disk supports (random) Termination shield Cable Under Test F Plug Feed through In the resonant region of the test chamber, the termination at the end plate (left hand side in Figure ) is a short circuit, while, at the end cap (right hand side in Figure ), the termination is a 50 ohm coaxial interconnect. Thus the chamber is mismatched at both ends with respect to 50 ohms. The return loss (RL) measured at the end cap (shown in Figure 3) is a result of the termination mismatches and the uniformity of the eter long coaxial test chamber. Here the center conductor is the shield of the sample under test, and the dielectric is made up of the sample jacket, dielectric support material, and air. The measured RL is in close agreement with the theoretical values shown in Figure 3. Figure. Triaxial test fixture overview P () Resonant test chamber () CUT shield load RL P Installation of eter test fixture Metallic cylinder Figure. Electrical regions of triaxial test fixture Figure 4. Open corridor installation of eter test fixture

3 .3 Installation The fixture was installed, as shown in Figure 4, above an open corridor that was located on a mezzanine in the Belden Engineering Center. Individual transmission line segments were suspended. meters above the floor and aligned with commercial Clevis hangars. This arrangement provided relatively easy access to both ends of the fixture; nonetheless it was necessary to run approximately 35 meters of low loss 50 Ohm coax to provide the input power P at the end plate. The network analyzer, amplifier, and data acquisition equipment were located at the far end in Figure 4. This provided for direct connection of the low-level power output P to either a preamplifier or the network analyzer input. 3. Test Procedure 3. Equipment setup The equipment setup is given in Figure 5, and the equipment used is listed in Table. Matching Pad Amplifier Test Fixture (5 x 6m) Network analyzer 3. Normalization Screening attenuation is derived from the difference in power levels (insertion loss) between the end plate P, (energized sample input) and the end cap P, (test chamber output). The measured power ratio is normalized [] with the following formula. The screening attenuation curve is then obtained by drawing the envelope (not shown) formed by connecting the resonant peaks in the power ratio versus frequency plot. a n = Where: a n = a meas = Z = Z S = ε = ε = ε 3 = ameas + 0 log0 Z S Z + 0 log0 normalized screening attenuation (decibels). ε ε ε 3 ε measured screening attenuation of sample in normalized and calibrated setup (decibels). impedance of cable under test (ohms). normalized impedance of screening attenuation fixture (50 ohms). relative dielectric permittivity of cable under test. relative dielectric permittivity of the environment of the cable. relative dielectric permittivity of screening attenuation fixture outer circuit with respect to a 0% velocity difference. Splitter Figure 5. Equipment setup for screening attenuation measurement Table. Test equipment Network Analyzer: HP8753ES, +0dBm, 0 Hz res. bandwidth Power Splitter: HP850C, 9.5 db loss nominal, DC-3 GHz Matching Pad: HP85B, 5.7 db loss nominal, DC-3 GHz Power Amplifier: HP8347A, 5 db gain nominal, 00 khz-3 GHz Resistive Termination on sample under test: 75Ω type F for CATV applications 4. Test program 4. Samples tested Several constructions of RG-6 type and RG-59 type coaxial cable were tested and are designated -4. ( is RG-59 type and -4 are RG-6 type cables). The details are given in Tables and 3. Table. Cable shield design data # Foil (inner) Braid/ Angle (inner) Foil (outer) Braid Angle (outer) Shield DCR mω/m a 3 a 4 a 95% b. c. /3 % Al 7 % Al /7 % Al /7 b a % Al /

4 Table 3. Shielding tape design data Layer Thickness (mm) Foil Type Al Foil Polyester Al Foil Width (mm) a b meter fixture test results Test results are plotted versus frequency in Figures 6 and 7 for the MHz and -00 MHz frequency ranges respectively. The test program involved testing multiple samples of the designs given above. However, only single sample results are shown for clarity. The statistical characteristics of screening attenuation are beyond the scope of this paper. In Figure 6 the lowest frequency resonance peak (not shown) is just below 00 MHz. Resonance peaks out to 000 MHz are connected by envelopes, as shown. The.5 meter fixture length is long enough to produce response peaks, which correspond to maximum power transfer as well as minimum screening attenuation performance at specific frequencies. The exact frequency location of the peaks will vary as the length is varied, but the peak amplitude will follow the envelope. Envelopes for designs and 3 are approximately flat, whereas envelopes for designs and 4 show an upward slope. At these frequencies the cable length under test is electrically long envelope eter fixture test results Figures 8- present screening attenuation results for eter lengths of designs -4 respectively. Results for.5 meter are also shown Figure 7. Screening attenuation versus frequency m envelope 0 00 Figure 8. Screening attenuation versus frequency Design Figure 6. Screening attenuation versus frequency In Figure 7 only one resonance peak is shown, which is at about 95 MHz. The fixture length of.5 meters is too short to allow resonance to occur at lower frequencies. Therefore, power transfer does not reach a maximum and the measured screening attenuation performance is consequently not minimal. Designs -4 show a tendency to converge below MHz. At these frequencies the cable length under test is electrically short m envelope 0 00 Figure 9. Screening attenuation versus frequency Design 4

5 00 0 envelope.5 m 0 00 For each design the eter length produces minimal screening attenuation, as shown by the envelope curve. Table 4 summarizes the relationship between.5 meter and 30 meter lengths for 0 MHz and 00 MHz results. For example, there is approximately a 0 decibel difference in performance at 0 MHz between the.5 meter and eter sample lengths. Alternatively, it can be seen that the difference in performance (error) is approximately decibels (designs 3 and 4) if it is assumed the eter performance at 0 MHz should be approximately equal to that of the.5-meter performance at 00 MHz. For design the eter performance at 0 MHz is about the same as the.5 meter performance at 00 MHz. For design the difference is about 30 decibels Figure 0. Screening attenuation versus frequency Design 3.5 m Figure. Screening attenuation versus frequency Design 4 envelope Conclusions The paper has described a test methodology that is in the process of being standardized internationally. The test has been applied to 30 meter and.5 meter long cable samples, and a comparison of the measured screening attenuation performance of a selection of braid and multi-foil/braid shield designs has been presented. The performance of eter and.5 meter long samples has been compared and differences as high as decibels were noted depending on frequency. The worse case or minimum screening attenuation performance of a particular shield design can depend on the length of the cable and the frequencies involved. In this work the test fixture has been normalized to 50 Ω and 90 %VP, but in actual usage the electrical environment around the cable can also influence the actual screening attenuation performance. Cable specifiers and system designers should be sure to specify screening attenuation requirements, which take into account the frequency band of operation and the cable lengths utilized. 6. Acknowledgements The authors are grateful to the Belden Electronics Division for the support extended to develop screening attenuation measurement technology. Thanks to Benjamin Willett for assistance with the laboratory measurements. Table 4. Screening attenuation estimates versus test length Design # db at 0 MHz.5 meter eter db at 00MHz.5 meter References [] 696- Amendment Radio-Frequency Cables Part : Generic Specification General, definitions, requirements and test methods. [] SCTE (Society of Cable Telecommunications Engineers, Inc.) IPS-TP-3B (preliminary-/8/000), Test Method for Shield Effectiveness: Screening Attenuation of Coaxial Cable. [3] O. Breitenbach, T. Hähner, and B. Mund, Screening of Cables in the MHz to GHz Frequency Range Extended Application of a Simple Measuring Method, IEE Colloquium on Screening Effectiveness Measurements, Savoy Place, London, May 6,

6 [4] bedea/rosenberger, CoMeT Coupling Measuring Tube. bedea BERKENHOFF & DREBES GMBH, Herborner Straβe Aβlar Germany [5] J. Kincaid, and C. Dole, Test Fixture Design and Shielded Screening Attenuation Performance of CATV Coaxial Cable, IEE Colloquium on Screening Effectiveness Measurements, Savoy Place, London, May 6, 998. [6] J. Kincaid, and C. Dole, Implementation of IEC 696- Shielded Screening Attenuation Test Method, International Wire and Cable Symposium, Philadelphia, Pennsylvania, November 9, 998. [7] J. Kincaid, and C. Dole, Shielded Screening Attenuation Test Method Down to 5 MHz, International Wroclaw Symposium and Exhibition on Electromagnetic Compatibility, Wroclaw, Poland, June 7, 000. Carl Dole Carl is a Product Engineer and has been with the Belden Engineering Center of Belden Electronics Division since 990. He currently holds one U.S. patent. His academic achievements include a B.S. degree in Electrical Engineering Technology ( With Highest Distinction ) from Purdue University. Prior to joining Belden, Carl worked 0 years in television broadcast engineering. He is a Certified Senior Broadcast Engineer and has a lifetime FCC General Class Radiotelephone License. His areas of responsibility include developing improved electrical test methodologies, writing technical papers, and working on new product development. He is a member of SMPTE, IEEE, and SBE. John Kincaid John is a Senior Product Engineer at the Belden Engineering Center. He holds BSEE and MSEE degrees from the University of Oklahoma and has over 5 years experience with Belden. His experience encompasses engineering management and product development positions in the USA as well as in Europe. He holds nine patents. John is a member of the IEEE and is active in IEC and TIA cable standardization activities. He is the US Technical Advisor to IEC SC 46A on coaxial cables, and is Convenor of IEC SC 46A/WG3 on data and CATV cable. He is also an expert on working groups 5 and 7 dealing with shielding and premises cabling issues. Carl W. Dole originally presented this paper November 4, 000 at IWCS 000, Atlantic City, New Jersey, USA. 6