Traceable loop antenna calibration methods using a VNA

Document Type: Activity Report

Authors: Osman Sen, Soydan Cakır, Tomáš Pavlíček, Martin Hudlička,
Urban Lundgren
Activity Partners: TÜBİTAK UME, CMI, RISE

56
 1. Summary

Loop antennas are widely used in various military and civil test applications in order to
measure the magnetic fields especially below 30 MHz. They are commonly calibrated as per
the standards IEEE 291 and SAE-ARP 958. Although there are a good number of
researches in literature about the comparisons of loop antenna calibration methods, we
brought the comparison measurements one more step further in this research and
investigated the existing 3 antenna loop calibration method that requires a Vector Network
Analyzer (VNA) and included it into our loop antenna calibration comparison. Ultimately, we
present the determined advantages and disadvantages of the three antenna method with
respect to the other standard methods. The measurements were performed by TUBITAK
UME and CMI the use of different calibration methods. The literature search was performed
by RISE.

2. Introduction

Loop antennas are extremely important especially in Electromagnetic Compatibility (EMC)

tests and measurements such as electromagnetic pollution measurements. They are widely
calibrated according to the standards IEEE 291 [1], SAE-ARP 958 [2] and other methods
[3,4,5]. The MIL-STD 461E/F standard [6,7] is one of the famous military EMC standards
which require the use of loop antennas. One of the important tests, RE101, requires a
shielded loop antenna with 36 turns and a diameter of 13.3 cm. The aim of this test is to
measure magnetic radiated emissions from Equipment Under Test (EUT) and subsystem
enclosures including electrical cable interfaces in the frequency range 30 Hz – 100 kHz. The
calibration of the RE101 loop antenna is specially defined in SAE-ARP 958 at a calibration
distance of 12 cm by using a reduced calculation equation. The factors of the RE101
antenna are often presented in unit of dBpT/µV since the limits in the standard are in the unit
of dBpT. The other test in MIL-STD 461E/F which requires the use of loop antennas is the
RS101 test. For this test, the standard defines two loop antennas. Unlike the RE101 test, this
test is an immunity test and its aim is to expose the EUT to magnetic radiated emissions in
the frequency range 30 Hz - 100 kHz. One of the required antennas for the RS101 test is a
transmitting loop antenna with 20 turns and a diameter of 12 cm for exposing EUTs to
magnetic fields. This transmitting loop antenna is also chosen as a transmitting antenna by
SAE-ARP958 standard for the calibrations of the RE101 receiving loop antenna. The other
antenna required by the RS101 test is a receiver loop with 51 turns and a diameter of 4 cm
for verifying the test setup before the actual test. In the RS101 test, the current that flows on
the transmitting loop is measured by means of a current monitor probe, and the produced
magnetic field on the EUT is just calculated by using this measured current, therefore the
precise measurement of the current becomes essential for the reliability of the test. In

57
addition to military applications, loop antennas are commonly utilized in screening
effectiveness measurements below 30 MHz. In screening effectiveness measurements,
calibration factors are not needed. The comparison of two measurement cases with and
without the target shield, yields the desired screening effectiveness value [8]. Moreover, loop
antennas are also used in special measurements such as the planar scanning measurement
technique of electromagnetic (EM) near-field (NF) emitted by electronic printed circuit boards
(PCBs) fully in the time-domain (TD)[9]. In [9], the probes having two different orientations of
metallic loops are used to measure magnetic NF horizontal and vertical components.

The loop antennas calibrations are very well studied in literature in [10] by means of the
comparison of the calibration methods. In this work, we also performed a comparison
between the loop antenna calibration methods but, unlike the researches in literature, we
also included the three antenna method introduced in [11-12] into the comparison and
showed its advantages and disadvantages over the other standard methods such as the
standard field method of IEEE 291.

3. Theory

The AFs for loop antennas are commonly expressed as AF=H(µA/m)/V(µV) in dBS/m or as
AF=B(pT)/V(µV) in dBpT/µV. One needs to subtract 1.98 dB from dBpT/µV and obtains
dBS/m. Sometimes, the AF is expressed in terms of equivalent electric field intensity
AF=E(µV/m)/V(µV). In this case, the unit of the AF is dBm-1[10].

The three antenna loop calibration method introduced in [11,12] requires three loop antennas
taken in pairs and yields the factors of each antenna separately. Unlike all the other standard
methods, this method requires a VNA and the measurement of each loop antenna pair in
turn as seen in Fig.1.

Figure 1. Three-antenna loop calibration setup [11,12]

58
The method necessitates the use of complex calculations given in (6).

'*+, .-+ .-,

!"#$ (%) = # &* (S/m) (6)
-+ *-, .+,

(/0 10 2) = (30405)0 (40305)0 (50304)

Where;

&3 8 2 9 :7"
9
3C(D7 D" )9 53C (D7 D" )F
67" = B3 8 8 J
1%;< >?@ :7"
A
E:7"F GH :7" I

:7" = &K7"
9
8 D79 8 D"9

(L0 M) = (304)0 (305)0 (405)

AFm i is the #i loop antennas’s magnetic antenna factor,

k is the wave number (=2π/N),
A12, A13, A23 are the transmission S-parameters between the antennas,
ω is the angular frequency,
µ< is the permeability of free space,
ZL is the load impedance matched to the transmission line,
r1, r2, r3 are the radii of the loop antennas,
d12, d13, d23 are the distances between the two loop antennas.

4. Research results (TÜBİTAK UME)

a. Other standard methods and experimental setup

The standard field method as per IEEE 291 is performed by using two antennas at
frequencies up to 100 MHz. One of them is the transmitting antenna with one turn and
produces a known free-space magnetic induction field at the location of a relatively close
receiving antenna. Since the antennas are very close to each other, the effect of reflection
from the ground and nearby objects is minimized. In this method, the near-zone field is
primarily magnetic and the value of magnetic field strength is expressed in terms of the
equivalent free-space electric component E that would exist in a far-zone radiation field. The
relation between E and H is given in (1) and the resulting value of the equivalent field for a
single-turn transmitting loop averaged over the area of the receiving loop is given in (2). If the
number of turns is not one, it is taken into account by multiplying the calculated E or H field
with the number of turns.

E = ho H (1)

59
 1/2
60 AI é æ 2p Ro ö2 ù
E@ ê1 + ç 2 ÷ ú
Ro3 êë è l ø úû
(2)

Where;
R0 = d 2 + r12 + r22 in m,
E is the magnitude of the equivalent free-space RMS electric field strength in V/m,
A is the area of the transmitting loop antenna in m2,
I is the uniform RMS current flowing on the transmitting loop in A,
l is the free space wavelength in m,
d is the distance between the centers of the transmitting and receiving loop antennas in m,
r1 is the mean radius of the transmitting loop antenna in m,
r2 is the mean radius of the receiving antenna in m.

This standard recommends the measurement of the current I flowing on the transmitting loop
antenna via a thermocouple ammeter therefore a vacuum thermocouple should be built into
the transmitting loop. The involved lead should be short and should not alter the mean radius
of the loop. Because of the difficulties of the construction of the transmitting loop together
with a vacuum thermocouple on it, the resistor method on the input of the loop or the current
clamp method stated in SAE-ARP 958 may be preferred to measure the flowing current on
the transmitting loop [10].

SAE-ARP 958 includes calibration of loop antennas required in both of military and civil EMC
standards. Unlike the method in IEEE 291, the SAE-ARP 958 standard method states a
calibration exactly at 1m from the transmitting antenna and a resistor along with current
sensor measurement methods in order to measure the current on the transmitting loop. In
addition, a transmitting loop with multi-turns in the calibration can be used while a
transmitting antenna with one turn is stated in IEEE 291. The magnetic field averaged over
the area of the receiving loop is calculated as follows [10]:

æ 2p f ö é 2 æ d xmt ö æ d rcv ö ù
2 2 2

1+ ç ÷ êL + ç ÷ +ç ÷ ú
2
1 d xmt è c ø êë è 2 ø è 2 ø úû (3)
H= I nxmt
2 3/2
2 4 é 2 æd ö æd ö2
ù
ê L . + ç xmt ÷ + ç rcv ÷ ú
êë è 2 ø è 2 ø úû

Where;
H is the magnetic field strength (A/m),
d xmt is the diameter of the transmit loop (m),
d rcv is the diameter of the receiving loop (m),
L is the distance between the loops, center-to-center (m),

60
I is the injected current into the transmitting loop (A),
n xmt is the number of turns in the transmitting loop,
f is the calibration frequency (Hz),
c is the speed of light (m/s).

In addition, SAE-ARP 958 also includes a special calibration, alias SAE-ARP 958 RE101
antenna calibration, for RE101 receiving loop antennas by stipulating the use of a reduced
frequency-independent equation given in (4) in order to calculate the magnetic field at a
calibration distance of 12 cm from the transmitting antenna. Moreover, in this special
calibration, SAE-ARP 958 requires the use of the special RS101 transmitting antenna whose
details are stated in the standard MIL-STD 461E/F.

m 0 INR 2
B = mH = (4)
2( R 2 + Z 2 ) 3 / 2
Where;
H is the magnetic field strength (A/m),
I is the injected current into the transmitting loop (A),
N is the number of turns in the transmitting loop
R is the radius of the transmitting loop (m),
Z is the distance between the loops, center-to-center (m),
µ is the permeability of free space
The antenna factor (AF) in units of dBS/m can be calculated from the H field, as:

æHö
AF = 20 logç ÷ (5)
èV ø
Where;
H is the magnetic field strength (A/m),
V is the voltage induced in the receiving loop (V) [10].

For comparison, as a first step, we used three large commercial antennas. Two of them are
ETS 6512 antennas (diameter: 56 cm) and the third antenna is a shielded home-made
antenna (diameter: 31 cm). Firstly, the three antenna calibration method was applied to the
antennas at a distance of 25 cm and the antenna factors of each antenna were determined
by using the VNA as seen in Fig.2.

61
 Figure 2. Three-antenna loop calibration setup of large commercial antenna
(ETS Model: 6512)

Thereafter, one of the commercial loop antennas (ETS 6512) was chosen as a target
antenna and the IEEE 291 standard field calibration method at the same measurement
distance was applied to it to be compared with the results of the three antenna method
between 10 kHz and 30 MHz (see Fig. 3). In the IEEE291 standard field method, as
transmitting antennas on which the flowing current was measured to calculate the magnetic
field just on the antenna under calibration, we employed two reference transmitting antennas;
Schwarzbeck Model: HFRA 5152-017 (diameter: 25 cm, DC - 3 MHz) and Schwarzbeck
Model: 5154-012 (diameter: 10 cm, 0.1 MHz - 30 MHz).

Figure 3. IEEE 291 calibration setup of large commercial antenna (ETS Model: 6512)

As a next step, we applied the same calibration methods to a set of RE101 receiving
antennas, which are normally designed to be used in RE101 tests of the standard MIL-STD
461E/F, between 30 Hz – 10 MHz at a distance of 12 cm that complies with the test distance
stated in MIL-STD 461E/F as seen in Fig.4 and Fig.5.

62
 Figure 4. Three-antenna calibration setup of RE101 antenna (R&S Model: HZ-10)

In the IEEE 291 standard method, as transmitting antennas, we used the same transmitting
antennas used in the previous step along with the RS101 transmitting antenna (Solar Model:
9230-1, d: 12cm, 30 Hz-100 kHz).

Figure 5. IEEE 291 calibration method for RE101 antenna (R&S Model: HZ-10)

In this second step, the loop antenna (R&S Model: HZ-10, diameter: 13.3 cm) that has the
operating frequency range between 30 Hz and 10 MHz was chosen as the target antenna
under calibration. As a final step, to be able to include the special RE101 antenna calibration
of the standard SAE-ARP 958, alias SAE-ARP 958 RE101 antenna calibration, into the
comparison, the same RE101 loop antenna (R&S Model:HZ-10) was measured between 30
Hz – 100 kHz by firstly using the IEEE 291 method at 12 cm and 33 cm in turn by means of
the RS101 transmitting antenna.

63
 Figure 6. Special SAE-ARP RE101 calibration method for RE101 antenna
(R&S Model: HZ-10)

Thereafter, also, as seen in Fig. 6, it was calibrated as per the SAE-ARP 958 military RE101
calibration method by using the same RS101 transmitting antenna at a distance of 12 cm
and 33 cm in turn and using the reduced equation given in (4). Ultimately, the results of the
SAE-ARP 950 RE101 calibration method were compared to the results of the IEEE 291
method in this final step. Finally, based on the performance and the ease of use of each
method, a conclusion is presented in this work.

b. Experimental results and discussions

The results of the commercial loop antenna (ETS Model: 6512) are presented in Fig.7 along
with the calibration data obtained from the calibration report of National Physical Laboratory
(NPL), UK which used the TEM calibration method for the calibration of this loop antenna. At
first glance, the NPL and three antenna method results look very compatible and the
maximum deviation is 0.4 dB between them. On the other hand, when we compare the
results of the standard field method with the results of the three antenna method, we see that
the results of IEEE 291 deviate from the results of the other methods by 2 dB maximum
especially in higher frequencies and by around 1 dB in the rest of the frequency range. This
clearly arises from the higher uncertainty of the IEEE 291 calibration method due to
uncertainty contributions coming from the measurement requirement of the current flowing
on the transmitting antenna and from the inclusion of the cable losses.

64
 Figure 7. Antenna factor results of large commercial antenna (ETS Model:6512)
at 25 cm

Figure 8. Antenna factor results of RE101 antenna (R&S Model: HZ-10) at 12 cm

The results of the RE101 antenna (R&S Model: HZ-10) are presented in Fig.8 along with an
example supplementary comparison average result of the same model antenna, which was
acquired from [13] just for information. The results are in very good agreement and look
satisfactory.

Ultimately, the results of the same RE101 antenna (R&S Model: HZ-10) calibrated as per the
IEEE 291 and special SAE-ARP 958 RE101 calibration methods are given in Fig. 9. The

65
graph clearly reveals that at first glance the special SAE-ARP 958 RE101 calibration method
results obtained at 12 cm deviate by 2 dB from the results of the same method performed at
33 cm and also from the results of the IEEE 291 method in the entire frequency range.
Moreover, the curves in Fig.9 also say that the antenna factors obtained in the IEEE 291
method are slightly dependent on the measurement distance while the antenna factors
obtained in the special SAE-ARP 958 RE101 calibration method for RE101 antennas are
significantly dependent on the measurement distance, most probably due to the reduced
equation given in (4). Despite this difference of 2 dB, the special SAE-ARP 958 RE101
calibration method stipulates the calibration of RE101 antennas at 12 cm from the RS101
transmitting antennas during the calibration. Moreover, the test standard MIL-STD 461E/F
requires the calibration of RE101 antennas as per this special SAE-ARP 958 RE101
calibration. This means that calibrations of RE101 antennas in any calibration method other
than the special SAE-ARP 958 RE101 calibration method at 12 cm may lead to test result
differences of 2 dB during a RE101 test due to the antenna factor.

Figure 9. Special SAE-ARP RE101 calibration method of RE101 antenna (R&S

Model: HZ-10) at 12 cm and 33 cm and its comparison with IEEE291 method

All the results and experiences acquired so far in this work show that all the investigated
methods yield consistent results in fact. On the other hand, the three antenna method seems
to be a very promising method and become prominent during the research in terms of ease
of use and lower uncertainty. While the measurement uncertainty of the three antenna
method was calculated approximately as 0.8 dB in our research as per the uncertainty
budget guidelines given in [12], the uncertainty of the standard field method (IEEE 291) was

66
around 1.8 dB when the uncertainty budget given in [13] was adapted for our measurements
of this research. The standard field method necessitates the measurement of the current on
the transmitting antenna and the measurement of cable losses, which requires much more
time and effort in comparison to the three antenna method. The necessity of measuring the
flowing current and taking the cable losses into account is one of the main contributions to
the measurement uncertainty in the standard field method. The three antenna method saves
us from the burden of measuring the current on the transmitting antenna and also from the
bother of the inclusion of cable losses. However, the three antenna method requires an
expensive network analyser and another two supplementary antennas to obtain the antenna
factors of the target antenna.

c. Uucertainty budget and calculation

The model function is determined according to [14], [15] and the uncertainty value AFMF for
this calibration method is calculated as follows;

AFMF = UVNA + UAlign + UDistan + UDiam + UDistr + URefl + URept

Where,

AFMF is the uncertainty of the Antenna Under Calibration (AUC)

UVNA is the uncertainty of S21 measurement of Vector Network Analyser (VNA)
UAlign is the uncertainty of the alignment loop antennas
UDistanLA1-LA2 is the uncertainty due to the distance between Loop Antenna (LA)1 – LA2
UDistanLA1-LA3 is the uncertainty due to the distance between LA1 – LA3
UDistanLA2-LA3 is the uncertainty due to the distance between LA2 – LA3
UDiamLA1 is the uncertainty due to the diameter of LA1
UDiamLA2 is the uncertainty due to the diameter of LA2
UDiamLA3 is the uncertainty due to the diameter of LA3
UDistr is the uncertainty due to the distribution of magnetic field inside loop antenna
URefl is the uncertainty due to the reflection from nearby object and the ground plane
URepeat is the uncertainty of the repeatability measurement system

67
Frequency Range: 30 Hz – 30 MHz

Component Value Probability

Input Quantity Units Divisor Ci Ui (dB)
symbol +/- Distribution
Vector Network
UVNA 0,35 dB Rectangular 1,732 1,00 0,20
Analyzer
Alignment loop
UAlign 0,18 dB Rectangular 1,732 1,00 0,10
antennas
Distance between
UDistanLA1-LA2 0,08 dB Rectangular 1,732 1,00 0,05
LA1-LA2
Distance between
UDistanLA1-LA3 0,08 dB Rectangular 1,732 1,00 0,05
LA1-LA3
Distance between
UDistanLA2-LA3 0,08 dB Rectangular 1,732 1,00 0,05
LA2-LA3
UDiamLA1 Diameter for LA1 0,05 dB Rectangular 1,732 1,00 0,03
UDiamLA2 Diameter for LA2 0,05 dB Rectangular 1,732 1,00 0,03
UDiamLA3 Diameter for LA3 0,05 dB Rectangular 1,732 1,00 0,03
Distribution of
UDistr 0,25 dB Rectangular 1,732 1,00 0,14
magnetic field
URefl Reflection 0,20 dB Rectangular 1,732 1,00 0,12
URepeat Repeatability 0,28 dB Normal 1 1,00 0,28
Combined
Uc Normal 0,42
uncertainty

= 2! = 0,84!#$
Taking a confidence interval of 95 % (k=2) for the expanded uncertainty:
"

68
 5. Results (CMI)

a. Other standard methods and experimental setup

CMI compared three antenna method using VNA to following standard methods:
- Calibration in Helmholtz coil
- Calibration using reference transmitting loop
- Calibration in TEM cell
For the measurement, we used three self-constructed single turn loop antennas:
Antenna #1 – radius 50 mm, unshielded, unbalanced (Fig. 10a) with built-in 50 Ω resistor
Antenna #2 – radius 40 mm, shielded, balanced, (Fig. 10b) with built-in 50 Ω resistor
Antenna #3 – radius 56 mm, shielded, balanced, (Fig. 10b) with built-in 50 Ω resistor

a) b)
Figure 10. Structures of used antennas
The transmission Aij between antenna pairs (1 – 2), (1 – 3), (2 – 3) was measured using the
borrowed vector network analyser R&S ZVA8 calibrated by OSL method. Distance between
coaxially placed antennas was d = 100 mm for each pair. The principal measurement setup
is shown in Fig. 10.

The measurement on VNA was performed in wide frequency range from 9 kHz to 30 MHz,
however, due to high noise level at lower frequencies, only values of AF for frequencies of 20
kHz and higher are shown. The dynamic range of obtained magnetic antenna factors is very
high (~ 60 dB), and the differences (< 1 dB) to other measurement method would not be
visible in the wide frequency range. Hence, we applied a normalization to AF using a
theoretic value of magnetic antenna factor for each frequency:

69
%&' = %& ( %&)*+- , .dB; dB.S/m1, dB.S/m11 (7)

356 7 9:> ? 7 52ACD:> ?

@ @
9< 9 <
%&)*+- = 20log
C
E0A F G@ >H
Where:
AF, AFteor are magnetic antenna factors (dB(S/m)),
AFN is normalized magnetic antenna factor (dB),
RS is serial resistance (50 Ω) ,
LS is loop inductance (H),
R0 is load resistance (50 Ω),
D is loop diameter (m).
The loop inductance was determined from measured input impedance and is 283 nH, 203 nH
and 340 nH for antenna #1, #2 and #3, respectively.
Experimental setup for other standard methods used for comparison with VNA method:

I. Helmholtz coil (‘H.C.’ label in figure legend)

Reference magnetic field was generated using Helmholtz coil – 2x 1 turn, 0.4m diameter.
Antenna factor was calculated as the ratio between known generated H-field and the voltage
at antenna output terminated by 50 Ω.

Measurement was carried out in the frequency range from 9 kHz to 1.3 MHz.

II. Reference transmitting loop (‘Ref. ant.’ label in figure legend)

In our laboratory, the single turn unshielded unbalanced antenna #1 (Ø 100 mm) is used as
reference transmitting loop with calculable AF for frequencies up to 30 MHz. CMCs are (0.53
– 1.0) dB and up to 10 MHz are confirmed by EURAMET.EM.RF-S27 comparison [16].

Measurement distance between coaxially placed antennas was chosen as 0.25 m (meets the
condition that the distance is greater than four times the radius of the larger antenna).
Magnetic antenna factor of measured loop (AF2) is obtained using following formula:

%
# (')*%,-.3
!" = 10 log #$ & !' + 20 log .
/
9 ?@ (7)
% "4567 89 :; % )<$% )<%% >%
/

Where
AF1 is magnetic antenna factor of reference transmitting loop,
P1 is transmitting antenna incident power,
P2 is power delivered by receiving loop to the load (R0 = 50 Ω),
d is distance between antennas,
r1 , r2 is loops’ radii,
f is frequency,
c is light velocity.

70
Measurement was carried out in the frequency range from 100 kHz to 30 MHz.

Crawford TEM cell (‘TEM’ label in figure legend)

Calibration was performed in Crawford TEM cell with septum height 0.25 m, antenna was
placed in the center of the cell, halfway between the bottom wall and the septum. Antenna
factor was calculated as the ratio between known generated H-field and the voltage at
antenna output terminated by 50 Ω.

Measurement was carried out in the frequency range from 100 kHz to 30 MHz.

Measurement of unshielded loop in the TEM cell in frequency band 10 MHz – 30 MHz is
stated only for illustration. Uncertainty caused mainly due to sensitivity to E-field component
is close to 3 dB.

b. Experimental results and discussions

Graphical outputs of comparison VNA 3-antenna method to other standard methods are
shown in figures 11 - 16. VNA calibration method is always marked by red colour, each trace
includes own uncertainty ‘band’.

Differences between bold traces in normalized plot directly corresponds to original AFdB(S/m)
error.

Antenna factors of antenna #2 and #3 are stated first, because there is a comparison with
reference antenna method performed by antenna #1.

71
Figure 11. Normalized magnetic antenna factor of antenna #2

Figure 12. Normalized magnetic antenna factor of antenna #2

72
Figure 13. Normalized magnetic antenna factor of antenna #3

Figure 14. Normalized magnetic antenna factor of antenna #3

73
 Figure 15. Normalized magnetic antenna factor of antenna #1

Figure 16. Normalized magnetic antenna factor of antenna #1 – y-axis scale changed

74
Following uncertainty sources were assumed:

label source of uncertainty for (p - q) antenna configuration

uVNA … uncertainty of Spq measurement using vector network analyzer
uREPEAT … repeatability of measurement
uE-FIELD/NOISE … antennas' sensitivity to E-field component / noise floor*
uREFLECTIONS … reflections from nearby object
uFIELD_DIST … magnetic field distribution inside the loop
uAXIS_ANGLE … angle between antenna axes (assumed ± 5 deg. deviation)
uΔRp (1 mm) … receiving antenna radius measurement error (assumed ± 1 mm)
uΔRq (1 mm) … transmitting antenna radius measurement error (assumed ± 1 mm)
measurement error of distance between antenna (assumed ± 2
uΔDIST (2 mm) … mm)

* The receiving antenna is rotated by 90 deg. relative to its original position - in this case, for
Spq measurement, transmission due to E-field antennas' sensitivity or Spq noise floor is
dominating.

75
Uncertainty budged for magnetic antenna factor AFM, example for 1 MHz

standard uncertainty
antenna value probability sensitivity
label divisor uncert. contribution
configuration (dB) distribution coef. ci
(dB) (dB)
(1 - 2) u(1-2)VNA 0.141 norm. 1 0.141 0.5 0.070
u(1-2)REPEAT 0.021 norm. 1 0.021 0.5 0.010
u(1-2)E-
FIELD/NOISE 0.098 norm. 1 0.098 0.5 0.049
u(1-2)REFLECTIONS 0.100 rect. 1.732 0.058 0.5 0.029
u(1-2)FIELD_DIST 0.100 rect. 1.732 0.058 0.5 0.029
u(1-2)AXIS_ANGLE 0.033 rect. 1.732 0.019 0.5 0.010
u(1-2)ΔR1 (1 mm) 0.084 rect. 1.732 0.049 0.5 0.024
u(1-2)ΔR2 (1 mm) 0.061 rect. 1.732 0.035 0.5 0.018
u(1-2)ΔDIST (2 mm) 0.387 rect. 1.732 0.224 0.5 0.112
(2 - 3) u(2-3)VNA 0.138 norm. 1 0.138 0.5 0.069
u(2-3)REPEAT 0.097 norm. 1 0.097 0.5 0.049
u(2-3)E-
FIELD/NOISE 0.205 norm. 1 0.205 0.5 0.103
u(2-3)REFLECTIONS 0.100 rect. 1.732 0.058 0.5 0.029
u(2-3)FIELD_DIST 0.100 rect. 1.732 0.058 0.5 0.029
u(2-3)AXIS_ANGLE 0.033 rect. 1.732 0.019 0.5 0.010
u(2-3)ΔR2 (1 mm) 0.056 rect. 1.732 0.032 0.5 0.016
u(2-3)ΔR3 (1 mm) 0.091 rect. 1.732 0.053 0.5 0.026
u(2-3)ΔDIST (2 mm) 0.374 rect. 1.732 0.216 0.5 0.108
(3 - 1) u(3-1)VNA 0.132 norm. 1 0.132 0.5 0.066
u(3-1)REPEAT 0.067 norm. 1 0.067 0.5 0.033
u(3-1)E-
FIELD/NOISE 0.018 norm. 1 0.018 0.5 0.009
u(3-1)REFLECTIONS 0.100 rect. 1.732 0.058 0.5 0.029
u(3-1)FIELD_DIST 0.100 rect. 1.732 0.058 0.5 0.029
u(3-1)AXIS_ANGLE 0.033 rect. 1.732 0.019 0.5 0.010
u(3-1)ΔR3 (1 mm) 0.068 rect. 1.732 0.039 0.5 0.020
u(3-1)ΔR1 (1 mm) 0.082 rect. 1.732 0.047 0.5 0.024
u(3-1)ΔDIST (2 mm) 0.361 rect. 1.732 0.208 0.5 0.104
Combined Uncertainty (k = 1) = 0.27
Expanded Uncertainty for confidence interval 95 % (k = 2) = 0.54

76
Measured AF for individual antennas

Frequency AF1 AF2 AF3 Uncertainty

(kHz) (dB(S/m)) (dB(S/m)) (dB(S/m)) (dB)
20 64.23 67.84 62.23 0.79
30 60.73 64.55 58.71 0.64
40 58.28 61.76 56.20 0.63
50 56.34 59.99 54.24 0.59
60 54.74 58.37 52.66 0.59
70 53.40 57.03 51.33 0.59
80 52.24 55.86 50.16 0.58
90 51.22 54.84 49.13 0.58
100 50.15 53.99 48.20 0.70
200 44.33 47.87 42.24 0.63
300 40.74 44.37 38.69 0.58
400 38.27 41.89 36.21 0.57
500 36.33 39.97 34.25 0.57
600 34.75 38.38 32.64 0.56
700 33.41 37.04 31.33 0.55
800 32.26 35.89 30.18 0.56
900 31.23 34.85 29.15 0.56
1 000 30.31 33.95 28.25 0.54
1 100 29.48 33.14 27.41 0.55
1 200 28.75 32.36 26.65 0.54
1 300 28.04 31.67 25.97 0.54
2 000 24.31 27.93 22.24 0.53
3 000 20.81 24.43 18.75 0.53
4 000 18.33 21.95 16.28 0.53
5 000 16.41 20.01 14.37 0.53
6 000 14.84 18.45 12.79 0.53
7 000 13.48 17.09 11.47 0.52
8 000 12.33 15.90 10.35 0.53
9 000 11.32 14.91 9.33 0.52
10 000 10.40 14.05 8.46 0.54
11 000 9.61 13.19 7.64 0.53
12 000 8.89 12.48 6.94 0.53
13 000 8.22 11.79 6.30 0.53
14 000 7.62 11.16 5.69 0.53
15 000 7.08 10.60 5.11 0.52
16 000 6.53 10.05 4.63 0.52
17 000 6.03 9.54 4.16 0.52
18 000 5.56 9.05 3.70 0.52
19 000 5.11 8.62 3.29 0.52
20 000 4.70 8.20 2.89 0.52
21 000 4.34 7.83 2.51 0.52
22 000 3.96 7.42 2.19 0.52
23 000 3.62 7.06 1.87 0.52
24 000 3.30 6.72 1.56 0.53
25 000 2.95 6.41 1.27 0.52
26 000 2.68 6.12 0.99 0.53
27 000 2.40 5.82 0.73 0.53
28 000 2.14 5.52 0.47 0.53
29 000 1.89 5.25 0.24 0.53
30 000 1.64 5.00 0.01 0.53

77
Sensitivity of AF to source uncertainties was determined from change of AF depending on
the change of equation’s input quantity. Analysis of sensitivity coefficients using partial
derivations of complex formula with respect to each input variable was not performed.

The relatively new loop antenna calibration method, the three antenna method using vector
network analyser, was compared to other standard calibration method. For the comparison,
three small circular loop antennas was used. Neglecting the calibration of unshielded loop in
TEM cell at higher frequencies, the differences between methods are very small (max. 0.7
dB, usually about 0.3 dB – 0.4 dB) in wide frequency range from 20 kHz to 30 MHz and
mostly are overlapped by uncertainty of measurement. The new calibration method is fast
and produces results comparable to presented standard calibration methods.

6. Conclusion

In this work, firstly, the three-antenna loop calibration method using a VNA was investigated
in detail. Thereafter, two loop antennas were calibrated as per both the standard field method
and the three-antenna loop calibration method. These methods were introduced and the
calibration results were compared. Finally, the advantages and disadvantages of both these
methods were presented. As a final result, the three-antenna loop calibration method looks
significantly more promising in terms of low uncertainty and ease of use in comparison to the
other time-consuming and exhaustive standard methods. The three antenna method also
saves us from the burden of measuring the current on the transmitting antenna and also from
the bother of the inclusion of cable losses.

7. References

[1] IEEE Std 291-1991, IEEE Standard Methods for Measuring Electromagnetic Field
Strength of Sinusoidal Continuous Waves, 30 Hz to 30 GHz.

[2] SAE-ARP 958 Rev. D, Electromagnetic Interference Measurement Antennas; Standard

Calibration Method.

[3] M. Ishii and K. Komiyama, “Impedance Method for a Shielded Standard Loop Antenna,”
IEEE Transactions on Instrumentation and Measurement, vol.56, no.2, pp. 422–425, April
2007.

[4] D.C. Arthur, Y. Ji and M.P.J. Daly, “A simplified method for the measurement of magnetic
loop antenna factor,” Precision Electromagnetic Measurements (CPEM) Digest, June 2008.

78
[5] T.H. Jang, J.H. Lim and B.W. Lee, “An Alternative Method for Calibration of monopole
and loop antenna in GTEM Cell,” 7th Asia Pacific International Symposium on
Electromagnetic Compatibility, pp. 873–875, May 2016.

[6] MIL-STD-461E-1999, Interference Standards Requirements for the Control of

Electromagnetic Interference Characteristics of Subsystems and Equipment, Department of
Defence, USA.

[7] MIL-STD-461F-2007, Interference Standards Requirements for the Control of

Electromagnetic Interference Characteristics of Subsystems and Equipment, Department of
Defence, USA.

[8] CENELEC EN50147-1: 1996, Anechoic Chambers Part 1: Shield Attenuation

Measurement.

[9] Y. Liu and B. Ravelo, “Fully time-domain scanning of EM near-field radiated by RF

circuits”, Progress In Electromagnetics Research (PIER) B, Vol. 57, pp. 21-46, 2014.

[10] S. Cakir, R. Hamid, L. Sevgi, “Loop-Antenna Calibration” IEEE Antennas and

Propagation Magazine, 53, 5, pp. 243-253, October 2011.

[11] Ishii, M and Komiyama, K., "A Measurement Method for Magnetic Antenna Factor of
Small Circular Loop Antenna by 3-Antenna Method", URSI. North American Radio Science
Meeting (Columbus), pp.458, July 2003.

[12] Ishii, M and Komiyama, K. “Estimation of uncertainty of calibration for loop antennas by
three-antenna method using automatic network analyzer”, 67th ARFTG Conference 2006,
June 2006.

[13] Frédéric Pythoud, Michele Borsero, Daniel Bownds, Soydan Çakır, Mustafa Çetintaş,
Karel Dražil, David Gentle, Domenico Giordano, Stuart Harmon, Josef Kupec, Yannick Le
Sage, Beat Mühlemann, Michal Ulvr, Giuseppe Vizio, Dongsheng Zhao, “Final report
EURAMET Supplementary Comparison EURAMET.EM.RF-S27”

Online:http://kcdb.bipm.org/appendixB/KCDB_ApB_info.asp?cmp_idy=1090&cmp_cod=EUR
AMET.EM.RF-S27&page

[14] Masanori Ishii ; Koji Komiyama, “Estimation of uncertainty of calibration for loop
antennas by three-antenna method using automatic network analyzer”, 2006 67th ARFTG
Conference, pp:156 – 163, June 2006

[15] Katsumi FUJII, Kojiro SAKAI, Tsutomu SUGIYAMA, Kouichi SEBATA, Iwao
NISHIYAMAA, “Calibration of Loop Antennas for EMI Measurements in the Frequency

79
Range Below 30 MHz,” Research and Development of Calibration Technology, vol. 63, no. 1,
pp. 71–81, 2016.

[16] Dražil, K: EURAMET Supplementary Comparison: EURAMET.EM.RF-S27, Antenna

factor for Loop Antennas, Czech Metrology Institute, Department of primary metrology of RF
electrical quantities, 2012.

80