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
 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
characteristics of the receiving antenna. If an antenna has
FaHdB  20 log10 FaH  [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
vector network analyzer (hereinafter, “VNA”) to measure
H dB [dB( A/m)]  V dB [dB( V)]  FaHdB [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
FaH (AUC)  (5)
free space ( 0  120  377.0  ), one can use  0 Z 0 FaH ( STD) S 21
Here,
E dB [dB( V/m)]  V dB [dB( V)]  FaHdB [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])
（20 log10 377.0  51.53）  0 ：Permeability of vacuum (4π × 10-7 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)
  2-5-1 Calibration of Loop Antennas for EMI Measurements in the Frequency Range Below 30 MH

Also, On the other hand, as shown in Fig. 3 (b), the Standard

at NICT is a shielded loop antenna with ϕSTD＝3.7 mm and
1  R0   15  rTx rRx 
2 2 4
 315  rTx rRx 
K 1   2      (6) radius rRx＝ 5.0 cm (REPIC, M201A-100R). It is calibrated
2R03
 8  R0  64  R02  
at the National Physical Laboratory (NPL) which maintains
2 2f the UK’s national standards. Figure 4 shows the traceabil-
  (λ is wavelength, c is speed of light)
 c ity chart.
Table 2 shows the calibration points where NICT pro-
R0  d 2  rTx2  rRx2
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
due to interpolation.
FaHdB  AUC   45.9  20 log10 f MHz  S 21
dB
 20 log10 K  FaHdB STD 
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 •D  etermine 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.
•R  F 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.
• I f 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 •A ntenna 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
2 Research and Development of Calibration Technology

AUC
rRx

STD
H av
rTx
Port #2

L
G Port #1

d
h

S 21
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)
  2-5-1 Calibration of Loop Antennas for EMI Measurements in the Frequency Range Below 30 MH

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
Japan Quality Assurance
Frequency range
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.

 
uFaH  AUC   u S 21
dB
       
 u K dB  u FaHdB STD  u H avdB  u I Tx
2 2 2 2 dB2
s2
Customer
Loop antenna  [dB] (8)

FFig. 4　Traceability chart Here,

u FaHdB (AUC) : Uncertainty of AUC

-25  dB
u S 21 : Uncertainty of S21 measurements
TESEQ HLA-6120 Calibrated by NICT
 
Magnetic field antenna factor [dB(S/m)]

-27 Calibrated by NPL

u K dB : Uncertainty of the calculated value ( K dB  20 log10 K )

u F (STD)  : Uncertainty of Standard. Value written in

dB
aH
-29 the calibration certificate issued by NPL

-31
 
u H avdB : Uncertainty due to unevenness of incident mag-
netic field of AUC
-33
 dB
u I Tx : Uncertainty because the current distribution that
conducts through the element of the Standard is
-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
be estimated by using the following equation to combine u S 21
5.1 　　  dB
 : Uncertainty of S21 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
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- K dB  : Uncertainty of calculated values of K
5.2 u　　 
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
Especially when the frequency is low, there is a FaHdB (STD) : Uncertainty of standard
5.3 u　　　　 
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)
  2-5-1 Calibration of Loop Antennas for EMI Measurements in the Frequency Range Below 30 MH

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
double logarithmic equation ( FadB  a ln f  b ) 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 3 2
polynomial ( Fa  af  bf  cf  d ) 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
Magnetic field antenna factor [dB(S/m)]

difference was estimated to be uncertainty. We consider the

50 values obtained provide an upper limit and lower limit for
40 uncertainty, and with a uniform probability distribution.
30

20 5.7 Repeatability
10
If repeated measurements are taken, we consider the
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
Difference [dB]

0.1
0
of the number of times measured (√3).
-0.1
-0.2
-0.3
Sources of uncertainty that occurs when calibrating a loop
-0.4
0.001 0.01 0.1 1 10 100
antenna were described above. As you can see by looking
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
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 uncertainty coefficient [dB]
Source Divisor Note
[dB] Distribution
u(xi) c(xi) |c(xi)|u(xi)
Actually measured by RF
1 Linearity 0.20 Rectangle √3 0.12 -1 0.12
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 uncertainty coefficient [dB]
Source Divisor 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 √3 0.00 1 0.006
Eqn. (A.4)
Combined standard uncertainty 0.14
Antenna Positions Diagram
Side view Top view
z
d
l
x y

h θ Tx
θ Rx
Ground plane x

78　　　Journal of the National Institute of Information and Communications Technology Vol. 63 No. 1 (2016)
  2-5-1 Calibration of Loop Antennas for EMI Measurements in the Frequency Range Below 30 MH

Definition Calibration
Plane wave rRx rRx
E

H
x
H 0 e  jkrRx H0 H 0 e  jkrRx H 0 (Constant)
 rRx
V
0  j 0 H 0 S Rx  2 rRx2  x 2 e  jkx dx V0   j 0 H 0 S 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)
 0 Z 0 S Tx FaH ( STD )
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
2 Research and Development of Calibration Technology

a vacuum (= 4π⋅10−7 H/m), Z0 is input impedance of the

R0  d 2  rTx2  rRx2
signal generator and receiver (= 50 Ω), and STx is area of
the loop antenna of the Standard.
2 2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 ITx 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)
 0 Z 0 FaH ( STD ) V
H av  I Tx S Tx K (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 av  Pin (A.3) ΓG of the signal generation side’s port connected to the
 0 Z 0 FaH ( STD )
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
a1  Pin (A.7)
 jkR
1 e
K
4 S Tx S Rx  
CTx CRx R
dl Rx  dl Tx (A.4)
V
b2  V (A.8)
1  L
Here, STx is the loop area of the Standard, SRx is the loop L  0

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.  0 Z 0 FaH ( STD) S 21
If the Standard and AUC are circular loop antennas and we obtain Equation (5) in the main text. Using dB to
with radius rTx and rRx, and the conditions kR0  1.0 and obtain this, we take the log of both sides and multiply by
rTx rRx R02  1 16 are satisfied, then 20 (20 log10), so it becomes

1  R0   15  rTx rRx  FaHdB  AUC   45.9  20 log10 f MHz  S 21  20 log10 K  FaHdB STD 
2 2 4 dB
 315  rTx rRx 
K 1       
2R03 8  R02  64  R2  (A.11)
   0  
(A.5) Here, the constant term of Equation (A.11) is the value
Here, it can be approximated by obtained from

2 2
20 log10  20 log10 f MHz
TTable 4　Maximum measurement capability 0 Z 0 2 10 4 10 50  20 log
6 7 10

Frequency range  45.9  20 log10 f MHz

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)
  2-5-1 Calibration of Loop Antennas for EMI Measurements in the Frequency Range Below 30 MH

RReference 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)
Iwao NISHIYAMA
4 K. Fujii and M. Ishii, “Calibration Method of Loop Antenna for EMI
Measurements below 30 MHz Using Reference Antenna,” IEICE Trans. B, vol.
Electromagnetic Compatibillity Laboratory,
Applied Electromagnetic Research Institute
J96-B, no.4, pp.437–445, April 2013. (in Japanese)
Calibration of Measuring Instruments
5 K. Fujii, K. Wake, and M. Ishii, “Development of a Standard Magnetic Field
and Antennas for Radio Equipment
Generator in the Frequency Range below 30 MHz,” IEICE Trans. B, vol.J99-B,
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