INFLUENCE OF EUT S VERTICAL CABLES TERMINATION TO DIFFERENT RESULTS OF RADIATED EMISSIONS IN FULLY AND SEMI- ANECHOIC CHAMBER

Transcription

1 INFLENCE OF ET S VERTICAL CABLES TERMINATION TO DIFFERENT RESLTS OF RADIATED EMISSIONS IN FLLY AND SEMI- ANECHOIC CHAMBER Gregor Kovac M.Sc. B.Sc. E. Eng. Head of EMC laboratory SIQ- Slovenian institute of Quality and Metrology Trzaska c., SI-1000 Ljubljana, Slovenia Abstract - Influence of ET s vertical cables termination in fully and semi-anechoic chamber has been investigated. Numerical calculation and experimental results are presented for end-driven wire model (EDW) of equipment under test (ET) with one cable. The investigation shows comparable results in semi and fully-anechoic chamber for ETs with vertical cables terminated with absorbing clamps. I. INTRODCTION Many reasons encourage people in academic and standardisation world to proposed new methods for measuring radio-frequency emission in fully anechoic chamber (FAC). There are two main reasons: Smaller and cheaper room for testing and Shorter time of measuring caused by no need of antenna scan. Standards prepared by ETSI already allowed both test sites for a subtitution measurement method. Many authors [3] [4] [5] [6] performed comparison of results in SAC and FAC. The calculation of the correction factor, which should be used for the same limit, is based on different normalised site attenuations (NSA) in both types of chambers. The NSA for i polarisationin OATS or semi anechoic chamber is described by the equation. [1] d i [ R + ( h h ) ] 1/ = (3) 1 The difference in NSAs could be used as a correction factor. A = E = A ( db) A ( db) (4) ioats A = (5) max 0log( d i ) EDi It is obvious that we can get different results for real ET s than from antennas. Common-mode cable currents are generally the dominant source of the radiated electric fields from electrically small tabletop sources. If we make contact of ET s vertical cable to the horizontal reference ground plane, we ll get completely different radiated pattern of ET (on OATS), than those in free space or fully iac A ioats = ( i = v, h) (1) max 0log f M EDi For the free space or FAC we get NSA: A ifac = () 0log f M log( d i ) () d i is a distance between transmit antenna (or ET) and receive antenna in FAC : Fig.1 End-driven wire model (EDW) of small ET with power cable in SAC with additional floor absorbers

2 anechoic chamber, where such contact is not possible. The ETs, which have a contact with the ground plane, represent the antenna structure with different resonance frequencies than those without contact to the ground plane. EDW as a simulating tool has been used for modeling the EM radiation from electrically small table-top equipment for more than fifteen years []. This paper introduces a similar model as Ω 10 db junction Fig. End-driven wire model (EDW) of small ET with power cable. II. INFLENCE OF VERTICAL CABLES Influence of vertical cables of small ETs are predominant in the low frequency range (up to 0 MHz). The cables with mirror pictures in a ground plane are long enough to represent antennas with relatively high gain in the above mentioned frequency range. All calculations and measurements, we made, are in this problematic frequency range. II.1 model The experiment with end-driven wire We make a simple end-driven wire (EDW) model, which represents small ET with one cable (e.g. power cable). The shield of a thin coaxial cable represents vertical cable of the ET. The cable is on the uper side of the table terminated with 10 db attenuator, Ω load and 10cm upstraight long wire (See figure ). The current can flow along the cable shield, which is terminated to the metal floor with different impedances. Measured E of grounded WEM comparative tool for diferent test methods. Calculated E of grounded WEM Fig.3 Results of radiation emission from EDW terminated to the ground plan: numerically and measured. radiated emission The calculation of radiation from EDW has been performed by method of moments (MoM). Fortran program Numerical electromagnetics code (NEC d) has been used for simulations. Two calculations have been made: calculation of radiation from EDW with contact to ground plan and calculation of radiation from the same EDW in free space. II. Calculations and measurements for different cable terminations First calculation is made for ET with cable connected to the ground plane. ET, which has contact with a ground plane, represents much bigger radiating object than ET without contact to the ground plane. Low impedance between cable and a ground plane is the main reason why enclosure and cables of small ET radiate high disturbances even at low frequencies. The first

3 resonance of our EDW can be observed at MHz. EDW for our experiment is 1. m long eccentrically driven monopole antenna. On the according to mentioned standard. Second measurement has been made with not terminated cable. Few meters of cable has been WEM not terminated WEM not terminated with m shorter cable under the RGP measured E of WEMterminated with absorbing clamp in SAC calculated E fromwem in the free space, without GNP Fig. 4: Results of radiation emission from EDW, Fig. 5: Results of radiation emission from EDW, with not terminated cable to the ground plane. with cable placed in absorbing clamp according plane. CISPR :003 other hand, quarter of wavelength at MHz is 1.m. This explains the first resonance. Good measured results encouraged us to use such model on education seminars, where we are trying to explain, why some centimetres big ETs make such noise in FM radio band. Such experiment explains, that the smallest ET with cable is not only few centimetres long radiating object, but at least 80cm, because of connected cable. This fact has to be taken into account also, when we use alternative test methods for very small test samples (e.g. GTEM). We have to underline, that EDW results are comparable to real ET if real ET has cable connected to the ground at the same point than EDW. sually cables are not connected directly to the ground plane. Cables are drawn through the hole in the ground plane according to IEC CISPR (versions up to 1998). Terminations of the cables are not defined in all test procedures rolled up under the ground plane. The results on figure 4 show many resonance frequencies in lower frequency range (up to 1 MHz). The amplitude and frequency position of resonances depends on the termination of the cable to the ground plane. The length of the cable can be varied also under the ground plane. We have extended the cable under the ground plane for two meters and repeated the measurement. Doted curve in figure 4 represents new results. As it can be seen, the minimum of amplitude at first measurement can be at same frequency than maximum of amplitude of second measurement. Both measurements have been performed with the same ET (EDW) and at the same positions of cables above the ground plane. Experiment shows that repeatability of the results according to IEC CISPR :1998 is very poor. The first step be able to get more comparable results is proposed in new standard IEC CISPR

4 WEM terminated with abs clamp in SAC with add. floor abs. h=1-180cm WEM terminated with abs. clamp in SAC H=100-0cm WEM terminated with abs. Clamp in SAC with add. floor. abs. H=1cm Fig. 6: Results of radiation emission from EDW, with a) cable placed in absorbing clamp according to IEC CISPR :003, b) cable placed in absorbing clamp with additional floor absorbers and limited high scan and c) cable placed in absorbing clamp with additional floor absorbers with fixed antenna height :003. The standard requires the use of absorbing clamp on the ground plane for the cables of tabletop equipment. Third experiment has been made with EDW, which had cable terminated with absorbing clamp, as shown on figure 4. The calculated radiated emission of EDW in the free space has its first resonance at 10 MHz. In this case, EDW is 1. m long eccentrically driven dipole antenna. On the other hand, half of wavelength at 10 MHz is 1.m. There are some diferences between measured results of EDW terminated with absorbing clamp in SAC have and calculated radiated emission of 1.m long EDW in free space. Two obvious differences can be observed on figure 5. The measured radiated field values are higher at low frequencies and resonance occurs earlier than in the case of theoretical EDW in free space. Both phenomena have two main reasons: Absorbing clamp does not represent total reflectance and absorbing material on the floor is placed only partly. The measured results at low frequencies can be slightly lower in fully anechoic chamber with absorber placed on whole floor surface. Even in the FAC we can expect reflections from floor absorbers. Reflections from the floor and absorbing material are higher if we are close to the floor with ET s parts. The ends of vertical cables in our case are only some millimetres under the ground plane. Next experiment shows difference in radiation emission from EDW in semi anechoic chamber, with and without additional floor absorbers. With fixed antenna height the results are quite different at higher frequencies. The results would be much more comparable if the antenna has been scanned in limited height range (from 1 cm to 1). Lower emissions in frequency range between 00 MHz and 0 MHz resulted from destructive interference of direct radiated contribution and reflected radiated contribution from partly absorbing floor. The influence of fully absorbing floor and raised floor under the ET require further investigation According to previous investigations [3][4] the height scan in FAC is recommended if ET has a large vertical size (including cables). Measurements on figure 6 show, that even small objects with minimum vertical cables length require antenna scan, when SAC with additional floor absorbers is used. III. FTRE WORK There are many chalanges for future work on investigation of cable influence. Numerical models for absorbing clamp and absorbing material can be used for studing the influence of both critical elements in FAC. Influence of longer cables laying out a certain length partly in horizontal direction and partly in vertical direction can be addressed for gaining better agreement between laboratory tests and a real life. Debates of longer cables might be finished after use of IEC CISPR :003 as European harmonized standard under EMC directive. Many questions remain unanswered for floorstanding equipment. We have to point out that configuration for floor-standing equipment in IEC CISPR : 003 remains unchanged with regard to the current used standard. IV. CONCLSION End-driven wire model (EDW) of small ET has been introduced as comparative model between different measuring methods of radiofrequency emission. EDW of simple ET can be used for investigation of cable influence when using different measurement methods. Results of radiated emission from an ET on OATS are comparable to results in fully anechoic room, if the vertical cables are terminated with absorbing clamps. SAC with additional floor absorbers can be used as FAC. Limited scan of antenna is recommened for

5 SAC with additional floor absorbers (converted SAC- CSAC). After some additional investigation of EDW terminated with absorbing clamp as simulating tool, EDW can be used for comparation of different FACs and CSACs. V. REFERENCES [1] Albert A. Smith jr, Robert F German, James B. Pate, Calculation of attenuation from antenna factors, IEEE Transaction on electromagnetic compatibility, vol. EMC-4, No.3, August 198 [] Todd H. Hubing, J.Frank Kaufman, Modeling the electromagnetic radiation from electrically small table-top products, IEEE Transaction on electromagnetic compatibility, voč. EMC-31, No.1, February 1989 [3] J Welinder, W. Müllner, D. Festa, P. Philips, P. A. Beeckman, H. Klamm, E. Rodriguez, Devolepement of new measurement methods of the EMC characteristics in smaller relatively inexpensive fully anechoic rooms, FAR Project SMT4-CT Third edition:001 [4] J. Welinder, J. Carlsson, Lennart Hasselgren, Emilio Rodriguez, Simulated behaviour of fully anechoic room for measurement of radiated emission, International symposium on EMC, Zurich 1999 [5] J. Welinder, J Carlsson, L Hasselgren, Emilio Rodriguez, Simulating behaviour of fully anechoic rooms for measurement of radiated emissions, International symposium on EMC, Zurich 1999 [6] D. Hansen, D. Ristau, Improving OATS antenna factors for application in small fully anechoic chambers, International symposium on EMC, Bruge [7] M. A. K. Wiles, W. Müllner, Conversion of Semi to fully anechoic Rooms per CENELEC pren147-3, Proceeding of IEEE International Symposium on Electromagnetic Compatibility (Montreal QC, Canada: IEEE EMC Society, 001)