Title:J2016E-02-05-01.indd p71 2017/03/01/ 水 10:26:55
2 Research and Development of Calibration Technology
2-5 Antenna Calibration 2-5-1 Calibration of Loop Antennas for EMI Measurements in the Frequency Range Below 30 MHz
Katsumi FUJII, Kojiro SAKAI, Tsutomu SUGIYAMA, Kouichi SEBATA, and Iwao NISHIYAMA
This paper describes the calibration method of loop antennas for EMI measurements in the frequency range below 30 MHz. NICT developed the method by ourselves and started to provide the calibration service that is certified by ASNITE accreditation program. Because the ASNITE is compliant with the international standard ISO/IEC 17025, the loop antennas calibrated by NICT can be used for all tests such as validations of radio equipment and EMI measurements of electronic devices.
1 Introduction from a transmitting loop antenna that interlinks with an antenna under calibration (receiving antenna). The strength New ways of using 30 MHz and lower radio frequencies of the magnetic field emitted from a transmitting antenna have appeared in recent years: IH induction rice cookers, is determined by correctly measuring the RF current flow- contactless chargers, contactless IC cards, etc., that differ ing through the element of the transmitting antenna. The from conventional radio broadcasts and commercial wire- RF current flowing through the element is measured by less communications. Both new and conventional uses using the conversion component attached to the element, share the same frequency band, so actually one must and converting the RF current flowing through the element measure the electromagnetic fields, and check that there is into DC voltage. A resistor that converts RF current into no interference. Also, electric and electronic devices such heat, or a thermocouple in a vacuum tube that uses heat as AC power supply adapters and LED lights that contain to generate DC voltage, are used for conversion compo- switching regulators are often used in our daily lives, so nents. In this case, correctness and traceability to national there are increasing needs to take measurements to regulate standards of the strength of the magnetic field emitted from unintentionally emitted electromagnetic noise. In this en- the transmitting loop antenna can be maintained by cor- vironment, NICT obtained ASNITE certification to fulfill rectly calibrating the conversion coefficient of the conversion the ISO/IEC 17025 standard for loop antenna calibration, component. However, there were limits to improving its and began providing calibration services in July 2015. precision, partly because this calibration is complex [2][3]. NICT developed a system that can issue internationally To solve this problem, NICT developed a new calibra- recognized calibration certificates which can be used for all measurements, without distinguishing between use for wireless communications or for electromagnetic noise Plane wave Loop antenna measurements. E FaH Electromagnetic field measurements of 30 MHz or lower have conventionally used loop antennas that have sufficiently smaller dimensions than the wavelengths. NICT H Receiver has provided calibration services and performed research and development on calibration of loop antennas since the V time of its Radio Research Laboratory [1]. The “standard magnetic field method” is used to calibrate loop antennas; this method theoretically obtains a magnetic field emitted FFig. 1 Magnetic field measurement and definition of magnetic field antenna factor
71 Title:J2016E-02-05-01.indd p72 2017/03/01/ 水 10:26:55
2 Research and Development of Calibration Technology
tion method in recent years [4][5], and started providing 3 Calibration method calibration services. This method is classified the same as a conventional “standard magnetic field method”, but in- Several methods of calibrating loop antennas have al- stead of conventional calibration methods that calibrate ready been proposed [6]–[8]. Some of these methods have from different physical quantities, this method uses an been made international standards by the International already calibrated loop antenna to calibrate another loop Special Committee of Radio Interference (CISPR) [8]. antenna. The advantages of this calibration method are that There are also several classification methods; Table 1 clas- it measures RF power instead of RF current, and that it sifies them into absolute calibration methods which obtain uses a loop antenna to obtain traceability to the national the loop antenna’s magnetic field antenna factor from standard. This paper shows the result of calibrating an other physical amounts (RF current or RF power), and actual commercially sold loop antenna, and explains a relative calibration methods that use a loop antenna with method to evaluate uncertainty associated with the calibra- an already known magnetic field antenna factor to obtain tion result. the magnetic field antenna factor of the antenna under calibration (hereinafter written as “AUC”). Table 1 writes 2 Definition of magnetic field antenna the types of devices needed for calibration, number of factor and how to use it times to measure, features, etc. This calibration method developed at NICT [4][5] is A loop antenna operates by electricity generated by a classified as a “relative calibration” method. It requires a magnetic field that interlinks with the loop plane. Therefore, loop antenna with an already known magnetic field an- as shown in Fig. 1, the loop antenna’s characteristic is de- tenna factor, but this calibration method has the advan- fined by using the ratio of strength H of the magnetic field tages that it can ensure direct traceability to the national that interlinks with the loop plane, and voltage V gener- standard, and that it only needs one measurement. ated in the receiver connected to the loop antenna. Requiring few measurements is an essential condition for H reducing uncertainty of calibration. FaH [S/m] (1) V Here, we consider using a “loop antenna with already
This FaH is called the “magnetic field antenna factor”. known magnetic field antenna factor” (hereinafter, “Standard”) as the transmitting antenna. The magnetic field antenna fac- Usually, it is expressed using dB. tor, as defined in Equation (1), is a parameter that expresses
dB characteristics of the receiving antenna. If an antenna has FaH log20 FaH10 [dB(S/m)] (2) reciprocity, then it can be used as the transmitting antenna. The unit is dB(S/m). By calibrating the loop antenna, if the As shown in Fig. 2, a circular loop antenna with radius
magnetic field antenna factor was already obtained, then size rRx for which we want to obtain the magnetic field one can read the output voltage V of the receiver, and use antenna factor (AUC) is placed at distance d from the
the following equation to know the strength of the mag- Standard (radius rTx), with its loop plane parallel to the netic field that interlinks with the antenna. Standard, with their centers aligned. Then, if we use a
dB dB dB vector network analyzer (hereinafter, “VNA”) to measure H A/m)][dB( V V)][dB( FaH [dB(S/m)] S21 between the Standard and AUC, the magnetic field (3) antenna factor of the AUC can be determined from the If the transmitting source is sufficiently far, and plane waves following equation. are being received, then in addition to wave impedance in 2K 1 F (AUC) (5) aH )( SSTDFZ free space (0 0.377120 ) , one can use aH00 21
dB dB dB Here, E V/m)][dB( V V)][dB( FaH [dB(S/m)] 51.53 )][dB( FaH (STD) ：Magnetic field antenna factor of Standard (4) [S/m] and convert into electric field strength. ：Angular frequency (2πf [rad/s]) -7 （ 53.510.377log20 ） 0 ： 10 Permeability of vacuum (4π × 10 H/m) Z 0 ：Characteristic impedance of measurement system (50 Ω)
72 Journal of the National Institute of Information and Communications Technology Vol. 63 No. 1 (2016) Title:J2016E-02-05-01.indd p73 2017/03/01/ 水 10:26:55
Calib ration of Loop AnatononCas orf EMI tCassuremenaas lont a h Freqtonncy Ronng Below 30 MH 5-1 2-
Also, On the other hand, as shown in Fig. 3 (b), the Standard
2 2 4 at NICT is a shielded loop antenna with ϕSTD＝3.7 mm and 1 R 15 rr 315 rr 0 RxTx RxTx K 3 1 2 2 (6) radius rRx＝ 5.0 cm (REPIC, M201A-100R). It is calibrated 2R0 8 R0 64 R0 at the National Physical Laboratory (NPL) which maintains 22 f the UK’s national standards. Figure 4 shows the traceabil- (λ is wavelength, c is speed of light) c ity chart.
2 2 2 Table 2 shows the calibration points where NICT pro- 0 Tx rrdR Rx vides calibration values. NPL does not provide all the d Distance between planes of Standard and AUC [m] calibration points that NICT provides, so we use the inter- rTx Loop radius of Standard [m] polated values of NPL’s calibration certificate. When inter- rRx Loop radius of AUC [m] polating, two regression curves are prepared with an 8 MHz frequency boundary, and used as the calibration Refer to [9]. If dB is used to calculate, then it is values. The next chapter describes the size of uncertainty
dB dB dB due to interpolation. aH AUCF log209.45 10 MHz Sf 21 log20 10 aH STDFK As shown in Fig. 2, measurements used a VNA (Rohde [dB(S/m)] (7) & Schwarz ZNB4) near the center in a semi-anechoic (See its derivation in the Appendix). chamber, with antennas placed with distance between an- tennas d = 20 cm, height from ground plane until loop
4 Calibration results element’s center h = 1.6 m, and S21 was measured. The VNA was calibrated by a procedure called “Unknown Thru” This paper shows results of calibrating the loop an- before measurement. tenna for EMI measurements shown in Fig. 3 (a) (Teseq, Figure 5 (a) shows calibration results. In the figure, the
HLA-6210). This is a shielded loop antenna with ϕAUC＝2.0 solid line shows calibration results at NICT. The dashed
cm and radius rRx＝ 30 cm. It contains a preamplifier. line shows calibration results at NPL. The loop under
TTable 1 Loop antenna calibration method No. of Calibration Name times Note instruments measure TEM cell • Determine the intensity of the magnetic field propagating in the TEM cell Power meter TEM cell from the input power. 1 Method Signal generator • The size of the AUC is restricted by the size of the TEM cell. Receiver • 1 measurement is enough. • RF current through the element is measured, determining the Absolute Transmitting loop strength of the magnetic field generated from the loop. Traceable Current calibration (with thermo- with RF current. measurement 1 couple or current • Magnetic field strength depends on the input resistance of the Method probe) thermocouple or current probe. • 1 measurement is enough. Three-antenna 3 antennas • Traceable with RF attenuation. 3 method VNA • Must measure 3 times. • If Standard and AUC have different dimensions, then must Transmitting loop compensate. Substitution Standard 2 • Instead of the Standard, management of transmitting loop is method Relative VNA required. calibration • Must measure 2 times (compare measurements). Magnetic field • Antenna with already known magnetic field antenna factor is used Standard antenna factor 1 for transmitting antenna to generate magnetic field. VNA method • 1 measurement is enough. Note) VNA: Vector Network Analyzer. Can also calibrate a combination of a signal generator and receiver.
73 Title:J2016E-02-05-01.indd p74 2017/03/01/ 水 10:26:55
2 Research and Development of Calibration Technology
AUC
rRx
STD Hav r Tx Port #2
L Port #1 G d h
S21 Ground plane Vector network analyzer
FFig. 2 Configuration to calibrate a loop antenna
TESEQ REPIC HLA-6120 M201A-100R
(a) (b)
FFig. 3 Loop under calibration and standard (a) AUC, (b) Standard
calibration is designed so its internal preamp’s amplifica- It is the value of NICT’s calibration results minus NPL’s tion ratio differs greatly depending on frequency [7], and calibration results. Looking at the results, about 1.0 dB we see that the AUC has almost the same magnetic field differences occur in the low frequency band, and the dif- antenna factors for various frequencies. In NICT’s calibra- ference decreases near 10 MHz, but the differences tend to tion results, discontinuity occurs at 8 MHz frequency be- increase again in higher frequencies. As described in the cause calibration values of the Standard are interpolated, next chapter, this is caused by NPL’s calibration values for
and with 8 MHz as the boundary, the regression curves are the Standard, and measurements of the loop radii rTx, rRx used separately in interpolation. and distance d between antennas. Figure 5 (b) shows the differences between two results.
74 Journal of the National Institute of Information and Communications Technology Vol. 63 No. 1 (2016) Title:J2016E-02-05-01.indd p75 2017/03/01/ 水 10:26:55
Calib ration of Loop AnatononCas orf EMI tCassuremenaas lont a h Freqtonncy Ronng Below 30 MH 5-1 2-
UK National Institute of Advanced Industrial TTable 2 Conditions and calibration points of loop antenna National Physical Laboratory Science and Technology (NPL) (AIST) Loop under calibration Frequency Frequency Frequency range Japan Quality Assurance antenna interval points Organization (JQA) conditions NICT Radius 10 cm 9 kHz～19 kHz 1 kHz 11 points to 60 cm 20 kHz～150 kHz 5 kHz 26 points Reference Standard Secondary Standard Reference Standard Type-N50 150 kHz～1 MHz 50 kHz 17 points Loop antenna RF attenuator Steel tape measure connector 1 MHz～30 MHz 0.1 MHz 291 points
Reference Standard Reference Standard Vector network analyzer Tape measure all size 1, so it is omitted.
dB 2 dB 2 dB 2 dB 2 dB 2 2 aH 21 FuKuSuAUCFu aH STD av Tx sIuHu Customer Loop antenna [dB] (8)
FFig. 4 Traceability chart Here, Fu dB (AUC) aH : Uncertainty of AUC dB -25 Su 21 : Uncertainty of S21 measurements TESEQ HLA-6120 Calibrated by NICT dB (K dB log20 K ) -27 Calibrated by NPL Ku : Uncertainty of the calculated value 10
dB Fu aH (STD) : Uncertainty of Standard. Value written in -29 the calibration certificate issued by NPL
dB -31 Hu av : Uncertainty due to unevenness of incident mag- netic field of AUC
-33 dB Iu Tx : Uncertainty because the current distribution that conducts through the element of the Standard is Magnetic field antenna factor field antenna factor [dB(S/m)] Magnetic -35 0.001 0.01 0.1 1 10 100 uneven. Frequency [MHz] s : Variability of measurements
Values used when obtaining a constant number of terms in Equation (7): uncertainty of frequency can be ignored relative to other sources of uncertainty; permeability in a vacuum is a defined value so it has no uncertainty; uncer- tainty of characteristic impedance is handled as included
in uncertainty of S21 measurements. Values of Equation (8): value are called “combined standard uncertainty,” so fi- nally, multiply by coverage factor k=1.96, to obtain 95% level of confidence expanded uncertainty. However, we use coverage factor k=2 for simplicity, to obtain expanded FFig. 5 Calibration results example (Teseq, HLA-6120) uncertainty with an approximately 95% level of confidence (strictly speaking, 95.45%). Below, as a result of estimating 5 Uncertainty uncertainty for each item, uncertainty at 30 MHz fre- quency is estimated as shown in the Table 3 (a) budget Uncertainty associated with calibration results ex- table; expanded uncertainty with an approximately 95% presses expanded uncertainty estimated to have approxi- level of confidence is estimated at 1.2 dB. mately a 95% level of confidence. The size of its value can dB 21 be estimated by using the following equation to combine 5.1 Su 21 : Uncertainty of S measurements uncertainty that occurs due to the multiple sources de- Table 3 (b) shows a budget table used to study and
scribed below [10]. However, the sensitivity coefficients are compose uncertainty sources of S21 Measurements. For the
75 Title:J2016E-02-05-01.indd p76 2017/03/01/ 水 10:26:55
2 Research and Development of Calibration Technology
VNA used to measure S21, mainly the following 5 items surement results. The frequencies are low, so were evaluated as main sources. changes like a microwave band do not occur. (1) Linearity Therefore, we provided 0.1 dB as the upper limit for As shown in Fig. 4, linearity of the VNA’s re- uncertainty. Its probability distribution is uniform. ceiver part was validated using an RF attenuator (secondary standard) calibrated by the National We composed the five sources of uncertainty described Institute of Advanced Industrial Science and above as shown in Table 3 (b). This resulted in our esti-
Technology (NMIJ/AIST). The largest difference mated 0.18 dB uncertainty for S21 measurements. between calibration values by the National Institute of Advanced Industrial Science and Technology writ- 5.2 Ku dB : Uncertainty of calculated values of K ten in the calibration certificate, and values displayed Table 3 (c) shows the results from our estimation of on the VNA, was taken as the upper limit of uncer- uncertainty that occurs when calculating K. These values tainty. are determined from Equation (6) and results of calcula- (2) Incompleteness When VNA Was Calibrated tions using the NEC2 electromagnetic field numerical The VNA is used after calibration using standard simulation software based on the method of moments. For devices called a calibration kit, compensating for example, we calculated (1) The size of uncertainty of dis- characteristics of circuits in VNA and the connected tance was the difference between value of KdB calculated coaxial cable. This time, compensation is performed when we input d = 20 cm, versus the values of KdB when by the calibration method called “Unknown Thru we input d = 20 cm – 2 mm and d = 20 cm + 2 mm; (2) calibration”, but the amount which could not be Uncertainty of measurement of the transmitting loop ra- compensated for remains as uncertainty due to in- dius (5 cm ± 2 mm); (3) Uncertainty of measurement of completeness of calibration. The 0.1 dB residual the receiving loop radius (30 cm ± 3 mm); (4) Uncertainty source match written in the VNA’s data sheet was of positioning of the Standard’s loop plane parallel to the added to uncertainty. This is mismatch uncertainty, AUC’s loop plane (± 5° when 0° parallel); (5) Uncertainty so the probability distribution is assumed to be a of positioning of the AUC’s loop plane parallel to the U-shaped distribution. Standard’s loop plane (± 5° when 0° parallel); (6) Uncertainty (3) Signal Leaks between Transmitting and Receiving that the Standard and AUC are positioned on the same axis Cables (0 mm if positioned on the same axis, ± 10 mm); (7) Effects We measure extremely weak signals at low fre- of reflection of the ground plane (we used the difference quency, and due to the loop antenna’s structure, between value calculated if height is 1.6 m, versus value signal can leak to the cable’s outer layer, which can calculated if in free space); (8) Uncertainty from using the create unneeded couplings between coaxial cables, approximation equation (difference versus Equation (A.4) affecting measurement values. These effects were is used as uncertainty). We considered that values diverging estimated to have an upper limit of 0.1 dB uncer- further than these values would not be measured, provid- tainty. Its probability distribution was assumed to be ing upper limits and lower limits for all these values, and uniform. assumed a uniform probability distribution. (4) SN Ratio dB Especially when the frequency is low, there is a 5.3 Fu aH (STD) : Uncertainty of standard
weak coupling between loop antennas, and S21 As shown in the traceability system diagram in Fig. 4, reaches to around -100 dB. This time, the AUC the Standard that NICT maintains is calibrated by the NPL contains a preamplifier, so sufficient signal strength in the UK. NICT calibrates using as reference the calibration could be obtained, but we provided 0.1 dB as the values and uncertainty written in the calibration certificate upper limit for uncertainty. Its probability distribu- issued by NPL. The value of uncertainty written in the tion is uniform. calibration certificate is expanded uncertainty for coverage (5) Changes in Measurement Values due to Cable factor k= 1.96 with approximately a 95 % level of confidence, Bending and Stretching so the value written in the calibration certificate divided by We took measurements while bending and coverage factor k=2 is used for standard uncertainty. It is stretching the cable, and evaluated changes in mea- actually 1 dB, so we input 0.5 dB in Table 3 (a).
76 Journal of the National Institute of Information and Communications Technology Vol. 63 No. 1 (2016) Title:J2016E-02-05-01.indd p77 2017/03/01/ 水 10:26:55
Calib ration of Loop AnatononCas orf EMI tCassuremenaas lont a h Freqtonncy Ronng Below 30 MH 5-1 2-
5.4 Uncertainty of interpolation factor is defined when plane waves are incident. As shown There is uncertainty from using the calibration points in Fig. 7, the magnetic field strength is constant, but it has calibrated by NPL to interpolate magnetic field antenna phase delay vs. the loop lane, and that tendency becomes factors at other frequencies. Figure 6 (a) shows results of much stronger with larger radius of the receiving loop. On calibration by NPL’s Standard. The dB values of magnetic the other hand, when calibrating, we use the average value field antenna factors change linearly vs. values with loga- of the magnetic field generated from the Standard placed rithms removed for frequencies below 8 MHz, so we use a opposite to and very near the AUC, and phase delay is not dB double logarithmic equation ( a ln bfaF ) for the considered. Therefore, both differences are uncertain. Thus, regression curve below 8 MHz, and use a third-order we use the numerical integral to obtain the difference be- dB 23 polynomial ( a dcfbfafF ) for above 8 MHz. tween when there was phase delay vs. when there was no For each, the largest divergence from the regression curve phase delay, and use that as uncertainty. For the probabil- at a frequency is taken as the value of uncertainty. ity distribution, we consider the values obtained by calcu- Divergence is within the maximum, so we assume a uni- lation as worst case values providing the upper limit and form probability distribution. Figure 6 (b) is a graph that lower limit, with a uniform distribution. shows how the regression curve diverges from the calibra- tion values obtained by NPL, in units of dB. It exceeds 0.1 5.6 Uncertainty due to distribution of current dB at 9 kHz and 10 kHz frequencies, but it is within 0.1 conducting through the standard dB at greater than 40 kHz frequency, so we used 0.1 dB as K is calculated under the assumption that the distribu- the basis of uncertainty. tion of current conducting through the standard is constant, but actually, the larger the standard loop antenna is com- 5.5 Uncertainty due to unevenness of magnetic pared to the wavelength, that is, the higher the frequency, field that interlinks with the loop area this leads to greater current distribution occurring, so As shown in Equation (1), the magnetic field antenna electric field component occurs, and unneeded links to the AUC occur. Numerical simulation using NEC2 shows there is a desired current value between the maximum value and 70 the minimum value of the current distribution, and that 60 difference was estimated to be uncertainty. We consider the 50
[dB( S/m)] values obtained provide an upper limit and lower limit for 40
factor uncertainty, and with a uniform probability distribution.
30 antenna 20 5.7 Repeatability
10 If repeated measurements are taken, we consider the ic field agnet ic
M degree of variability; if measured again another day, we 0 consider whether the same measurement results were ob- -10 0.001 0.01 0.1 1 10 100 tained. In Table 3 (a), we substituted the variability (ex- Frequency [MHz] perimental standard deviation) when measured 3 times. (a) For the calibration result, we use the average value of results 0.4 from measuring 3 times, so standard uncertainty is the 0.3 0.2 experimental standard deviation divided by the square root [dB ] 0.1 √ 0 of the number of times measured ( 3).
ce feren ce -0.1
Di f -0.2 -0.3 Sources of uncertainty that occurs when calibrating a loop -0.4 antenna were described above. As you can see by looking 0.001 0.01 0.1 1 10 100 Frequency [MHz] at Table 3 (a), the main sources of greater uncertainty are (b) uncertainty associated with the NPL’s calibration values of FFig. 6 Magnetic field antenna factor the Standard, and uncertainty of calculated values of K. As (a) Magnetic field antenna factor by NPL, (b) Differences between calibration values by NPL and values obtained by interpolating calibration you can see by looking at Table 3 (c), uncertainty of cal- values culated values of K is dominated by uncertainty of mea-
77 Title:J2016E-02-05-01.indd p78 2017/03/01/ 水 10:26:55
2 Research and Development of Calibration Technology
TTable 3 Uncertainty budget
30 MHz, rTx=5 cm, rRx=30 cm, d=20 cm (a) Value Probability Divisor Standard Sensitivity Contribution Source distribution uncertainty coefficient [dB] Note [dB] u(xi) c(xi) |c(xi)|u(xi)
1 S21 0.18 － － 0.18 -1 0.18 See Table 3 (b) 2 K 0.14 － － 0.14 1 0.14 See Table 3 (c) Calibration certificate of Normal 3 FaH(STD) 1.00 2 0.50 -1 0.50 superior calibration (k=2) organization 4 Interpolation 0.10 Rectangle √3 0.06 -1 0.06 5 Incident magnetic field 0.04 Rectangle √3 0.02 1 0.02 6 Current distribution 0.03 Rectangle √3 0.02 1 0.02 7 Repeatability 0.10 Normal √3 0.06 1 0.06 Measure 3 times Combined standard uncertainty 0.56 Expanded Uncertainty (Approx. 95 % level of confidence) 1.2 Coverage factor k=2 (b)
Uncertainty of S21 measurement(30 MHz, rTx=5 cm, rRx=30 cm, d=20 cm) Standard Sensitivity Contribution Value Probability Source Divisor uncertainty coefficient [dB] Note [dB] Distribution u(xi) c(xi) |c(xi)|u(xi) Actually measured by RF 1 Linearity 0.20 Rectangle √ 0.12 -1 0.12 3 attenuator 2 Imperfection of calibration 0.10 U-shape √2 0.08 1 0.08 Spec. sheet of VNA 3 Signal leak between cables 0.10 Rectangle √3 0.06 -1 0.06 4 S/N ratio 0.10 Rectangle √3 0.06 -1 0.06 S/N > 38 dB 5 Cable bending, stretching 0.10 Rectangle √3 0.06 1 0.06 Combined standard uncertainty 0.18 (c)
Uncertainty of K(30 MHz, rTx=5 cm, rRx=30 cm, d=20 cm) Standard Sensitivity Contribution Value Probability Source Divisor uncertainty coefficient [dB] Note [dB] Distribution u(xi) c(xi) |c(xi)|u(xi) 1 d 0.17 Rectangle √3 0.19 1 0.099 d=20 cm±5 mm
2 rTx 0.01 Rectangle √3 0.39 1 0.006 rTx=5 cm±2 mm
3 rRx 0.15 Rectangle √3 0.04 1 0.087 rRx=30 cm±3 mm
4 θTx 0.04 Rectangle √3 0.01 1 0.024 θTx=0°±5°
5 θRx 0.03 Rectangle √3 0.01 1 0.018 θRx=0°±5° 6 l 0.01 Rectangle √3 0.00 1 0.006 l=0 mm±10 mm 7 Effect of ground plane 0.01 Rectangle √3 0.00 1 0.006 h=1.6 m Deviation from 8 Approx. of Eqn. (6) 0.01 Rectangle √ 0.00 1 0.006 3 Eqn. (A.4) Combined standard uncertainty 0.14 Antenna Positions Diagram Side view Top view z d l x y
h θTx θRx
Ground plane x
78 Journal of the National Institute of Information and Communications Technology Vol. 63 No. 1 (2016) Title:J2016E-02-05-01.indd p79 2017/03/01/ 水 10:26:55
Calib ration of Loop AnatononCas orf EMI tCassuremenaas lont a h Freqtonncy Ronng Below 30 MH 5-1 2-
Definition Calibration Plane wave rRx rRx E
H x
jkrRx jkrRx eH H 0 (Constant) 0eH H 0 0
r Rx 22 jkx SHjV j H 2 rS x e d x 0 jV 0 H S R x0 0 0 R x0 Rx rRx
FFig. 7 Unevenness of magnetic field that interlinks with the loop area
surement of distance between antennas, and measurement if one measures incident power to the transmitting an- of loop radii. Said another way, if we can accurately measure tenna, then one can determine the magnetic field strength, the loop radii and distance between antennas, then we can and there is no need to calibrate the thermocouple or reduce uncertainty. There remains room to improve mea- current probe used to measure RF current. This method surements of distance and radii, that is, “length”. also has the advantages that for characteristics of the transmitting antenna, we use the magnetic field antenna 6 Calibration and measurement capability factor when operating as a receiving antenna, making it easy to obtain traceability to the superior calibration orga- Calibration and measurement capability (hereinafter nization, and we can minimize equipment for maintaining written as “CMC”) is the smallest uncertainty associated the Standard. with calibration results provided by NICT. ASNITE certi- There are more diverse uses of radio waves below fication calibration is also for an AUC that does not contain 30MHz for more purposes in recent years. In response, in a preamplifier, so the values of CMC are larger than the addition to in tests of wireless equipment, to enable its use values of uncertainty estimated in Table 3. Table 4 shows in electromagnetic interference (EMI) measurements, CMC for calibration of a loop antenna for magnetic fields. NICT obtained the ASNITE certification that it satisfies the If the AUC has small dimensions, then the receiving level ISO/IEC 17025 standard, and began providing internation- (signal to noise ratio) decreases, and it becomes more dif- ally recognized calibration values. It is predicted that vari- ficult to measure, especially in the low frequency band ous forms of use will be developed, so it seems that demand under 40 kHz, so the frequency bands are divided for will increase for high precision loop antenna calibration determining this. To improve the receiving level, inserting services. an external preamplifier is a possible method, but if the output port of the AUC has a large reflection coefficient, Appendix. Derivation of Equation (5)
then one must sufficiently reduce effects of reflection of the When incident power Pin (incident power, not power preamplifier’s input port. consumed in the antenna) is incident into the connector
part of the Standard shown in Fig. 2, the current ITx con- 7 Conclusion ducting through power feed gap part of the loop element of the Standard can be obtained by using the magnetic field
We explained a loop antenna calibration method devel- antenna factor FaH (STD) of the Standard oped by NICT, using EMI measurements at 30 MHz or 2 lower frequencies. Instead of conventional methods which I Tx Pin (A.1) aHTx00 STDFSZ )( measure RF current conducted through an element of the
transmitting loop antenna, in the method NICT developed, Here, ω is angular frequency (= 2πf), μ0 is permeability in
79 Title:J2016E-02-05-01.indd p80 2017/03/01/ 水 10:26:55
2 Research and Development of Calibration Technology
−7 a vacuum (= 4π⋅10 H/m), Z0 is input impedance of the 2 2 2 0 Tx rrdR Rx signal generator and receiver (= 50 Ω), and STx is area of the loop antenna of the Standard. 22 f (λ is wavelength, c is speed of light) c Now, as shown in Fig. 2, at a place distance d from the Standard, the loop antenna for which we want to obtain and Equation (6) in the main text is obtained [9]. If the
the magnetic field antenna factor (AUC) is positioned so Hav obtained and the received voltage V measured at the its loop plane is parallel to the Standard, with their centers receiver connected to the AUC are substituted into Equation
aligned. Then, assuming the current TxI conducting through (1), then the equation that obtains the magnetic field an- the transmitting loop element is constant, the average tenna factor of the AUC is expressed in the following
value Hav of the strength of the magnetic field that interlinks equation. with the loop lane of the AUC can be obtained using the 2K Pin following Equation [1]–[6][9]. FaH (AUC) (A.6) aH00 STDFZ )( V av TxTx KSIH (A.2) This equation is for when the signal generator and re- Therefore, if we substitute Equation (A.1), we obtain ceiver are each prepared separately, but as shown in Fig. 2, 2K if a VNA is used to calibrate, then the reflection coefficient H P av in (A.3) ΓG of the signal generation side’s port connected to the aH00 STDFZ )( Standard, and the reflection ΓL of the receiving side’s port Here, K is the value that expresses coupling due to the connected to the AUC, are both considered to be 0, so this magnetic component between the two loop antennas, becomes given by 1 Pa in (A.7) 1 e jkR K dd ll TxRx (A.4) 4 SS CCTx Rx R V RxTx b V (A.8) 2 1 L 0 Here, STx is the loop area of the Standard, SRx is the loop L area of the receiving loop antenna, and R is the distance and
between line element vectors dlTx and dlRx on the loop ele- b2 V S 21 (A.9) ments. This integral can be applied to rectangular and a1 Pin triangular loop antennas, not only to circular loops. so Equation (A.6) becomes Numerical calculations can be done easily using a com- 2K 1 FaH (AUC) (A.10) puter. aH00 )( SSTDFZ 21 If the Standard and AUC are circular loop antennas and we obtain Equation (5) in the main text. Using dB to kR 0.1 with radius rTx and rRx, and the conditions 0 and obtain this, we take the log of both sides and multiply by 2 10 Rrr 0RxTx 161 are satisfied, then 20 (20 log ), so it becomes 2 2 4 dB dB dB aH AUCF log209.45 10 MHz Sf 21 log20 10 aH STDFK 1 R0 15 rr RxTx 315 rr RxTx K 1 2R 3 8 R 2 64 R 2 (A.11) 0 0 0 (A.5) Here, the constant term of Equation (A.11) is the value Here, it can be approximated by obtained from
2 2 log20 10 log20 10 6 7 log20 fMHz10 TTable 4 Maximum measurement capability Z00 1 02 50104 Frequency range log209.45 fMHz10 DUT loop radius, rRx 9 kHz ≤ Freq. < 40 kHz 40 kHz ≤ Freq. ≤ 30 MHz (A.12)
5 cm ≤ rRx < 10 cm 1.6 dB 1.4 dB and fMHz is the frequency when MHz is the unit. 10 cm ≤ rRx < 20 cm 1.4 dB 1.4 dB
20 cm ≤ rRx ≤ 30 cm 1.4 dB 1.4 dB
80 Journal of the National Institute of Information and Communications Technology Vol. 63 No. 1 (2016) Title:J2016E-02-05-01.indd p81 2017/03/01/ 水 10:26:55
Calib ration of Loop AnatononCas orf EMI tCassuremenaas lont a h Freqtonncy Ronng Below 30 MH 5-1 2-
RRen fere ce Kouichi SEBATA 1 K. Nakamura, Y. Nanai, and Y. Sudo, “CALIBRATION OF VHF FIELD Senior Researcher, Electromagnetic STRENGTH METER USING LOOP ANTENNA,” Journal of RRL (Kiho), vol.5, Compatibillity Laboratory, Applied no.18, pp.52–59, Jan. 1959. Electromagnetic Research Institute 2 A. Suzuki, M. Sakasai, K. Koike, H. Masuzawa, “Uncertainty Estimation of Loop Calibration of Measuring Instruments Antenna Calibration System in MF/HF Band,” Journal of NICT, vol.53, no.1, and Antennas for Radio Equipment, pp.19–28, March 2006. geodesy 3 K. Fujii, K. Koike, and Y. Matsumoto, “Calibration of Standard Magnetic Field Generators below 30 MHz,” IEICE Trans. B, vol.J96-B, no.4, pp.446–457, April 2013. (in Japanese) 4 K. Fujii and M. Ishii, “Calibration Method of Loop Antenna for EMI Iwao NISHIYAMA Measurements below 30 MHz Using Reference Antenna,” IEICE Trans. B, vol. Electromagnetic Compatibillity Laboratory, J96-B, no.4, pp.437–445, April 2013. (in Japanese) Applied Electromagnetic Research Institute 5 K. Fujii, K. Wake, and M. Ishii, “Development of a Standard Magnetic Field Calibration of Measuring Instruments Generator in the Frequency Range below 30 MHz,” IEICE Trans. B, vol.J99-B, and Antennas for Radio Equipment no.3, pp.124–134, March 2016. (in Japanese) 6 M. Kanda, E. B. Larsen, M. Borsero, P. G. Galliano. , I. Yokoshima, and N. S. Nahman, “Standard for Electromagnetic Field Measurements,” Proc. of IEEE, vol.74, no.1, pp.120–128, Jan. 1986. 7 M. J. Alexander, M. J. Salter D. A. Knight, B. G. Loader, and K. P. Holland, “Calibration and use of antennas, focusing on EMC applications,” A National Measurement Good Practice Guide, no. 73, Dec. 2004, available from http:// www.npl.co.uk/publications/good_practice/ 8 Specification for radio disturbance and immunity measuring apparatus and method – Part 1-6: Radio disturbance and immunity measuring apparatus – EMC antenna calibration, CISPR 16-1-6, Edition 1.0, 2014-12. 9 F. M. Greene, “The Near-Zone Magnetic Field of a Small Circular-Loop Antenna,” J. Res. NBS, vol.71C, no.4, pp.319–326, Dec. 1967. 10 ISO, Guide to the Expression of Uncertainty in Measurement, 1st edition, 1995.
Katsumi FUJII, Dr. Eng. Research Manager, Electromagnetic Compatibility Laboratory, Applied Electromagnetic Research Institute Calibration of Measuring Instruments and Antennas for Radio Equipment, Electromagnetic Compatibility
Kojiro SAKAI Technical Expert, Electromagnetic Compatibility Laboratory, Applied Electromagnetic Research Institute Calibration of Measuring Instruments and Antennas for Radio Equipment
Tsutomu SUGIYAMA Senior Researcher, Electromagnetic Compatibillity Laboratory, Applied Electromagnetic Research Institute Calibration of Measuring Instruments and Antennas for Radio Equipment
81