2007

Poznańskie Warsztaty Telekomunikacyjne Poznań 6 - 7 grudnia 2007

P O Z NPOZNAN A N U N IUNIVERSITY V E R S I T Y OOF F TTECHNOLOGY E C H N O L O G YACADEMIC A C A D E MJOURNALS IC JOURNALS No Electrical Engineering 2008

Łukasz SKIBA Piotr RYDLICHOWSKI*

DISTANCE TO FAULT AND VECTOR MEASUREMENTS LABORATORY EXERCISE PROPOSAL In the article, the Distance To Fault (DTF) and vector measurements using spectrum analyzer are proposed. Presented measurement schemes can be used as a laboratory exercise at Technical Universities - Telecommunication Faculty. Three measurement setups are presented: distance to fault in coaxial cables, vector measurements of a cable television filter and coupled microstrip lines. Experimental results are compared with analytical and numerical modeling of considered problems. Presented results have proven being valuable teaching aid for modern spectrum analyzer operation.

Keywords: Distance to fault, Vector measurements, laboratory exercise, spectrum analyzer

1. INTRODUCTION The modern telecommunication systems are becoming more and more complex and require appropriate hardware for service and maintenance. Today spectrum analyzers have increasing number of features, options and predefined measurements options for specific telecommunication systems – both wired and wireless. Modern spectrum analyzers can be bought with useful function of DTF and vector measurements. It is possible because of a built in tracking generator which is used as a signal and reference source. Formerly these kind of measurements required a special and expensive dedicated hardware which however, still presents today better accuracy and functionality. DTF measurements allow to find exact location of discontinuity, connection or failure in a specific cable system. Vector measurements allow to find transfer function for a specific device under test - amplitude and phase characteristics. To conduct DTF or vector measurements properly a basic theoretical and practical background knowledge is required. The article presents laboratory exercise proposal where students and future engineers can become familiar with specificity of DTF and vector measurements. It __________________________________________ * Poznań University of Technology.

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

1/8

Łukasz Skiba, Piotr Rydlichowski

can be learned how to setup hardware and interpret obtained results. Rohde & Schwarz FSH-3 [1] spectrum analyzer was used, it requires VSWR bridge and calibration cable for DTF measurements. Vector measurements require calibration only. Little can be said about exact algorithm and method for DTF and Vector measurements implemented in considered spectrum analyzer. Exact details are protected by manufacturer or patents and not available for public. In general an internal signal coupler is used with digital signal processing for timing purposes. In DTF measurements spectrum analyzer measures waves reflected at discontinuities and using that data calculates reflection coefficients. In vector measurements signal received by analyzer from device under test is compared with reference output signal from tracking generator. Amplitude and phase characteristic is calculated using compared signals.

2. MEASUREMENTS SCHEMES The following measurements were done: DTF in an YWlek and RG 058 coaxial cables, vector measurements of a cable TV filter (“channel 30 trap”) and coupled microstrip lines. Figures 1 and 2 present measurement setups.

Fig. 1. Measurement setup for DTF.

VSWR bridge, combined with spectrum analyzer measures reflection coefficients. Test cable can be replaced by system of interconnected cables.

Fig. 2. Setup for vector measurements.

During vector measurements one port is connected to tracking generator output and the second port is connected to spectrum analyzer input. For DTF measurements spectrum analyzer requires cable parameters – attenuation value, propagation velocity and approximate length (needed for scaling purposes only). Analyzer needs to be calibrated for DTF and vector measurements – for each frequency, display parameters (filters bandwidths) and temperature. Figure 3 present measurement hardware photograph.

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

2/8

Distance to fault and vector measurements, laboratory exercise proposal

Fig. 3. Measurement hardware photograph.

3. DTF MEASUREMENTS Various options for DTF measurements are possible. The simplest configuration is one with cable terminated with short or specific load. This case allows us to find exact analytical solution for reflection coefficient [2] and compare it with experimental results. Another DTF measurement configuration is with interconnected cables with splitters. Such option models more actual and real life conditions. One of the analyzed configurations consists of YWlek 50-2,25/7,25 cable with 22,36 meters of length connected with 1,5 m of 75 Ohm cable shorted at the end. This cable is typically used for CB radio applications. Detailed measurement parameters and figures with results can be found in [3]. Obtained results are in good agreement with theoretical reflection coefficient values for shorted cable (Γ=1) and connection between 50 Ohm and 75 Ohm cables (Γ=0.2). Spectrum analyzer only measures first reflection waves, it is unable to process multiple reflections. Measurements have shown that chosen probing frequency and bandwidth has significant impact on achieved accuracy. Higher frequency can give more precise location of discontinuity in cable. In presented configuration for lower frequency and narrow bandwidth it is not possible to observe reflections for second short 75 Ohm cable. It should be noted that each cable is designed for specific frequency range for which all given parameters (attenuation and propagation constant) are valid. Choosing different frequency and higher bandwidth than specified for a given cable results in a condition where reliance on not verified parameters and significant characteristic impedance mismatch between spectrum analyzer output and cable input is observed. Higher

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

3/8

Łukasz Skiba, Piotr Rydlichowski

probing frequency allows for more precise measurements in cables with short interconnects. However long cables require lower probing frequencies due to lower attenuation for such frequencies. Vector measurements accuracy is limited to calibration quality only. In specific cases influence of connecting cables needs to be also considered (especially connectors).

4. VECTOR MEASUREMENTS In first configuration, channel filter is analyzed and amplitude with phase characteristic is compared with SPICE simulations of filter theoretical circuit model. Figure 4 present filter circuit model for spice simulations. Figures 5 and 6, show theoretical amplitude and phase characteristic respectively. Figure 7 present experiments results. Measurements have shown acceptable agreement between experimental and theoretical results. Observed differences between characteristics (especially phase) are result of incomplete theoretical circuit model - circuit elements values.

Fig. 4. SPICE circuit model for channel filter.

Fig. 5. Theoretical amplitude characteristic for channel filter.

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

4/8

Distance to fault and vector measurements, laboratory exercise proposal

Fig. 6. Theoretical phase characteristic for channel filter.

Fig. 7. Experimental amplitude and phase characteristic for channel filter.

Another device chosen for vector measurements were coupled microstrip lines (Figure 8). Theoretical results were obtained using 3D MoM Electrmagnetic Full Wave Solver – IE3D from Zeland Corporation [3]. Transfer function for ports 1-2 and 3-4 and coupling 1-4, 1-3 was obtained. Results have shown significant differences for amplitude and phase characteristic (Fig. 9,12 and 13). Coupling was in good agreement (Fig. 10 and 12). Observed differences are result of simplified device model in field solver (especially influence of port connectors). Because analytical model for such device is complicated to define this element is not recommended for laboratory exercise for students at beginner level. However, coupling have shown good agreement with theory and could be beneficial experiment for students. In general coupled microstrip lines require more precise simulation model for reference results.

Fig. 8. Coupled microstrip lines.

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

5/8

Łukasz Skiba, Piotr Rydlichowski

Fig. 9. Numerical and experimental amplitude characteristic for microstrip lines (1-2,3-4).

Fig. 10. Numerical and experimental amplitude characteristic for microstrip lines (13,31,24,42).

Fig. 10. Numerical and experimental amplitude characteristic for microstrip lines (14,41,23,32).

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

6/8

Distance to fault and vector measurements, laboratory exercise proposal

Fig. 11. Numerical and experimental phase characteristic for microstrip lines (1-2,3-4).

Fig. 12. Numerical and experimental phase characteristic for microstrip lines (13,31,24,42).

Fig. 13. Numerical and experimental phase characteristic for microstrip lines (14,41,23,32).

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

7/8

Łukasz Skiba, Piotr Rydlichowski

7. CONCLUSIONS Article presents exemplary DTF and vector measurements using spectrum analyzer. Three configurations were considered, DTF in coaxial cables (with optional interconnects), vector measurements of cable TV channel filter and coupled microstrip lines. Coaxial cables and filter case showed satisfactory results. Experiments were in good agreement with theory. Results for coupled microstrip lines case mostly did not agree with numerical model. Only coupling (amplitude characteristic), showed expected values. Presented results are of significant importance for students and are good proposal for laboratory exercise. Students can learn how to setup hardware and how to interpret obtained results, practical knowledge can be gained. Coupled microstrip lines case require further research to construct appropriate analytical model.

REFERENCES [1] [2] [3] [3]

Rohde & Schwarz FSH-3 operating manual, Rohde & Schwarz Corporation, 2006 Jerzy Osiowski: Zarys Rachunku Operatorowego, NT, Warszawa 1972 Łukasz Skiba: „Measurement and identification of characteristics of the selected transmission channels”, praca magisterska, Politechnika Poznańska 2007 IE3D manual, Zeland Corporation, www.zeland.com, 2004

PWT 2007 - POZNAŃ 6-7 GRUDNIA 2007

8/8