Evaluation Technique for EMI Anechoic Chamber

over 1 GHz by Plane-wave Spectral Decomposition

and Fish-eye Lens Camera

Michitaka Ameya, Satoru Kurokawa
The Metrology Institute of Japan
National Institute of Advanced Industrial Science and
Technology
Tsukuba, Ibaraki, Japan
m.ameya@aist.go.jp
Ikuo Watanabe, Mikito Yamaguchi, and Ryoichi
Hasumi
DEVICE Co. LTD.,
Kohnosu, Saitama, Japan
sales@deviceco.co.jp


AbstractWe propose a new site evaluation technique using
plane-wave spectral decomposition and fish-eye lens camera. The
proposed method enables to obtain the intensity and the angle of
arrival of reflection waves from walls, ground and ceiling of the
anechoic and facilitates finding out the cause of the reflection.
Keywords-component; anechoic chamber evaluation, plane-
wave spectral decomposition, fish-eye lens camera; DOA estimation

I. INTRODUCTION
Electromagnetic Interference (EMI) test over 1 GHz has
attracted much attention because EU and VCCI in Japan have
started the regulation for EMI above 1 GHz from October 2010.
Above 1 GHz, the EMI test is operated in the fully anechoic
room (FAR) or semi-anechoic chamber (SAC) with enough
absorbers on the ground to emulate the free-space environment.
In the anechoic chamber, the standing wave is generated by the
interference of the direct wave emitted from the EUT with the
reflection wave from the walls, ground and ceiling of the
chamber. The standing wave causes a dependency on the EUT
positions of the EMI test results and the dependency on the
positions increases the measurement uncertainty of the EMI
test. Therefore, the detailed reflection analysis of the EMI
anechoic chamber is most important issue to realize the
accurate EMI test.
Many evaluation methods for anechoic chamber are
proposed before now [1], [2]. The SVSWR method is one of
the evaluation methods of anechoic chambers for EMI over 1
GHz. This method is prescribed as the standard method in
CISPR 16-1-4 [3] to evaluate the performance of the anechoic
chamber. SVSWR is derived from the voltage standing wave
ratio (VSWR) in circuit theory, and defined as the ratio of the
maximum electric field strength to the minimum electric field
strength of the area. In the CISPR standard, SVSWR is
measured at the only four positions (front, right, left, and
center) at the height of middle of the test volume or 1 m above
the bottom of the test volume. Only these positions may not be
enough for the detailed analysis of anechoic chambers for EMI
over 1GHz because there is a possibility of existing higher
standing waves at the other positions. Moreover, SVSWR
cannot identify the cause of the reflection of the anechoic
chamber. Based on these backgrounds, we propose a new site
evaluation technique using plane-wave spectral decomposition
and fish-eye lens camera. The proposed method enables to
obtain the intensity and the angle of arrival of reflection waves
and facilitates finding out the cause of the reflection.
In the next section, we explain the procedures of the
proposed evaluation method. In Section III, the measurement
setup and measured results are shown. Finally, Section IV
presents the conclusions of this paper.
II. EVALUATION METHOD
A. Plane-wave Spectral Decomposition of Electric-field
Distribution on EMI Turn-table
If the characteristics of FAR is perfect in EMI test, the
electric fields only consists of direct wave emitted from the
EUT because there are no reflection waves. In such a no
reflection environment, the variation of the electric field
distribution is minimized because standing waves never
happen. Therefore, we can evaluate the performance of
anechoic chambers by checking the uniformity of electric field
distribution.
The SVSWR method [3] is one of the evaluation methods
of anechoic chambers by checking the uniformity of the fields.
In this method, the small biconical antenna and the double
ridged-guide horn antenna are generally used as a transmitting
antenna and a receiving antenna. In reciprocal media, the
transmission characteristics between transmitting and
receiving antennas and the antenna interchanged transmission
characteristics are exactly the same. Therefore, the measured
SVSWR should be the same even if the small biconical
antenna is used for receiving and the double ridged-guide horn
antenna for transmitting. In other words, the variation in EMI
results caused by the standing waves is equivalent to the
nonuniformity of the electric fields transmitted from the
receiving horn antenna.
Based on these facts, we use the double ridge-guide horn
antenna as a transmitting antenna and focused on the
uniformity evaluation of the electric field distribution over the
turn-table by scanning the receiving small biconical antenna
along x-axis and y-axis over the turn table.
According to the electromagnetic wave theory [4],
arbitrary electric field
) , , ( z y x E
can be expressed by
superposition of plane-waves of the form
) (
) , (
z k y k x k j
y x
z y x
e k k
+ +
X
, which have the same frequency,
different amplitudes, and are travelling in different directions.
Thus the field ) , , ( z y x E can be written as

} }
+

+

+ +
=
y x
z k y k x k j
y x
dk dk e k k z y x
z y x
) (
) , ( ) , , ( X E

(1)

while
) , (
y x
k k X
is the vector plane-wave spectrum (PWS),
z y x , , is the Cartesian coordinates of the electric field, and
z y x
k k k , ,
is the wave number for each axis. These wave
numbers are related to the traveling direction of the plane-
waves, the relationship between wave numbers and direction
angle is

| u cos sin k k
x
=
| u sin sin k k
y
=
u cos k k
z
=
(2)

where k is the magnitude of wave vector k ,
t / 2 = = k k
,
and | u, indicate travelling directions of a plane-wave in
spherical coordinates. By calculating two-dimensional Fourier
transformation to the x and y , the PWS can be written as

} }
+

+

+ +
= dxdy e z y x e k k
y k x k j z jk
y x
y x z
) (
) , , ( ) , ( E X

(3)

A field distribution created by only the reflected waves in
the anechoic chamber is obtained from subtraction of the
reference field
direct
E , which is an ideal no-reflection field
(direct wave only), from the measured target field
meas
E
including direct wave and reflected waves. Therefore, the
electric field
reflect
E , which is generated by only the reflection
waves at
0
z z = , and its PWS is expressed by

) , , ( ) , , ( ) , , (
0 direct 0 meas 0 reflect
z y x z y x z y x E E E =
(4)
| |
} }
+

+

+ +

= dxdy e z y x z y x
e k k
y k x k j
z jk
y x
y x
z
) (
0 direct 0 meas
reflect
) , , ( ) , , (
) , (
0
E E
X

(5)

The PWS of reflection waves,
) , (
reflect y x
k k X
, indicates the
angular distribution of reflection waves. Therefore, we can
estimate the intensity and the direction of arrival (DOA) of
each reflection waves and facilitate improving performance of
anechoic chambers.
Due to the difficulty of measuring electric fields directly,
we use a small biconical antenna, which is used in SVSWR
method, as an electric field sensor. To simplify the problems,
we only consider the vertical polarization field on this paper.
The z component of the electric field,
) , , (
0 reflect ,
z y x E
z
, and its
PWS is expressed by

| | ) , , ( ) , , (
) , , ( ) , , ( ) , , (
0 direct 0 meas
0 direct , 0 meas. , 0 reflect ,
z y x V z y x V AF
z y x E z y x E z y x E
z z z
~
=

(6)
| |
} }
+

+

+ +

= dxdy e z y x V z y x V
e
AF
k k X
y k x k j
z jk y x z
y x
z
) (
0 direct 0 meas
reflect ,
) , , ( ) , , (
) , (
) , (
0
| u

(7)

Where AF is the antenna factor of the receiving small
biconical antenna and V
meas
and V
direct
are the receiving voltage
of target fields in the anechoic chamber and reference fields.
In practice, we cannot measure the electric field of infinite
area ( + = to , y x ). Hence, two-dimensional Blackman
window [5]

> <
s s
(

+
(

> <
s s
(

+
(

=
=
) , ( 0
) (
) ( 4
cos 08 . 0
) ( 2
cos 5 . 0 42 . 0
) (
) , ( 0
) (
) ( 4
cos 08 . 0
) ( 2
cos 5 . 0 42 . 0
) (
) ( ) ( ) , (
max min
max min
min max
min
min max
min
1
max min
max min
min max
min
min max
min
1
1 1
y y y y
y y y
y y
y y
y y
y y
y w
x x x x
x x x
x x
x x
x x
x x
x w
y w x w y x w
d
d
d d
t t
t t

(8)
is multiplied to the differential field
reflect , z
E . The example of
Fourier transformation pair of two-dimensional Blackman
window is shown in Fig. 1. The interval of measurement point
should be less than the half of the wavelength to fulfill the
sampling theorem [5]. In contrast to the SVSWR method, the
maximum and minimum amplitude of electric field
distribution are not required to be measured as long as
fulfilling the sampling theorem. A vector network analyzer
can measure the ratio of the input signal to the output signal
easily, therefore (7) is rewritten by S-parameter as

| |
| |
} }
} }
+ +
+ +

=
=
= '
max
min
max
min
max
min
max
min
0 0
) (
0 direct 21, 0 meas 21,
) (
0 direct 0 meas
reflect ,
reflect ,
) , , ( ) , , ( ) , (
/ )) , , ( ) , , ( ( ) , (
) , (
) , (
) , (
) , (
y
y
x
x
y k x k j
y
y
x
x
y k x k j
in
y x
in
y x z z jk z jk
y x z
dxdy e z y x S z y x S y x w
dxdy e V z y x V z y x V y x w
k k W
AF V
k k X
e e k k X
y x
y x
z z
| u
(9)
where XY means the convolution of X and Y. V
in
is the
input voltage of the transmitting horn antenna. We
defined
) , (
reflect , y x z
k k X ' as pseudo plane-wave spectrum
(PPWS) for the reflection waves. Although the PPWS is
influenced by the antenna factor of the receiving small
biconical antenna, the radiation pattern of the transmitting
double ridged-guide horn antenna and the window
function ) , ( y x w , the DOA information of reflection wave is
still included in the PPWS. In a similar way, the PPWS of the
target field and the reference field is calculated by (10) and
(11).

} }
+

+ +
=
'
max
min
max
min
0
) (
0 meas 21,
meas ,
) , , ( ) , (
) , (
y
y
x
x
y k x k j z jk
y x z
dxdy e z y x S y x w e
k k X
y x z
(10)
} }
+

+ +
=
'
max
min
max
min
0
) (
0 direct 21,
direct ,
) , , ( ) , (
) , (
y
y
x
x
y k x k j z jk
y x z
dxdy e z y x S y x w e
k k X
y x z
(11)
) , ( y x w
(b)
) , (
y x
k k W
(a)
x (mm)
y
(
m
m
)
3000 3500 4000
-600
-400
-200
0
200
400
600
N
o
r
m
a
l
i
z
e
d

A
m
p
l
i
t
u
d
e

(
d
B
)
-100
-80
-60
-40
-20
0
k
x
/k
k
y
/
k
-1 -0.5 0 0.5 1
-1
-0.5
0
0.5
1
N
o
r
m
a
l
i
z
e
d

A
m
p
l
i
t
u
d
e

(
d
B
)
-100
-80
-60
-40
-20
0

Figure 1. Fourier transformation pair of two-dimensional Blackman window
where y
min
= 700 mm, y
max
= +700 mm, x
min
= +3000 mm, and x
max
= 4400 mm.

B. Fish-eye Lens Camera
Fish-eye lens camera [6] can take a photo of 180
hemisphere by some projection schemes. There are many
projection methods, such as orthographic projection,
stereographic projection, equal area projection and so on. The
image height of the orthographic projection is described as

u sin | | f = h
y y x x
h h e e h + =
| u
| u
sin sin
cos sin
f h
f h
y
x
=
=

(12)
where h is two-dimensional vector of the image height, h
x
and
h
y
are the x and y component of h, f is the focal length of fish-
eye lens and u, | are the incident angles of light in spherical
coordinates. By comparing with (2), the image height and the
wave number have the scaling relationship. Therefore, PPWS
2D-image calculated by (9) can be related to the photo taken
by orthographic projection fish-eye lens camera with one-to-
one relationship.

u
u sin | | f = h
IncidentAngleofLight
f
FocalLength
h
ImageHeight
ImagePlane
|
y y x x
h h e e h + =
| u cos sin f h
x
=
| u sin sin f h
y
=

Figure 2. Orthogonal projection

III. MEASUREMENT SETUP AND RESULTS
Measurement setup is shown in Fig. 3. We assume a turn-
table with 1.4 m diameter arranged on the center of 25 m
length by 16 m width by 9 m height semi-anechoic chamber.
The S
21
distribution is measured along the measurement grid
using XY field scanner shown in Fig. 4. The grid interval is 20
mm for both x and y directions. This interval fulfills the
sampling theorem with the frequency up to 6 GHz. The
measurement area is from x = 3000 mm to 4400, and from y =
700 mm to +700 mm. The origin of this coordinate is set to
the center of the aperture plane of the transmitting ridged-
guide horn antenna. The number of the total measurement
point is 5041 (71 points by 71 points). This XY scanner can
move the receiving small biconical antenna into two directions
(x and y-directions). By entering the trigger signal from the x-
axis controller to the VNA, the field distribution of x-direction
at each y-coordinate can be measured at one sweep. For
connecting the transmitting port and the receiving port of the
VNA to the antennas, a 20 m coaxial cable and a 30 m coaxial
cable are used. We use a double ridged waveguide horn
antenna (ETS-Lindgren Model 3115) as a transmitting antenna.
The biconical broadband antenna (Schwartzbeck SBA9119) is
employed as a receiving antenna. The radiation pattern of the
receiving biconical antenna is compliant with CISPR-16-1-4
specification according to the manufactures datasheet [6]. To
generate the reflection waves obviously, a whiteboard is
located at the 2 m aside from the absorbers on the ground. The
whiteboard is backed with the 2 m by 1 m metal plate;
therefore we assume the whiteboard as an electric reflector.
Only the vertical polarization component is measured in this
trial.
Fig. 5 shows the SVSWR measured in the target chamber
with whiteboard and without whiteboard. In Fig. 5(a), the
SVSWR is more than 6 dB in the frequency range below 4
GHz. Although we can judge the chamber with whiteboard is
unsuitable for EMI measurement according to the CISPR
standard, we cannot find the cause of the high SVSWR only
from these results. In the case of no whiteboard, The SVSWR
is less than 5 dB in the frequency range from 1 GHz to 6 GHz
and less than 2 dB especially above 3.5 GHz.
Fig. 6 shows the distribution of the field with whiteboard,
the field without whiteboard and the difference between two
fields. The measurement frequency is 4 GHz. From Fig. 6(a),
an interference pattern is observed due to the reflecting wave
from the reflector. The reference field is measured after
removing the whiteboard; hence, there is a still interference
pattern due to the small reflecting waves from the XY field
scanner and the anechoic chamber walls.
Fig. 7 shows the PPWS of the field with whiteboard, the
PPWS of the field without whiteboard and the differential
PPWS. From Fig. 7(c), the 1
st
maximum peak is observed at
) / , / ( k k k k
y x
= ) 63 . 0 , 69 . 0 ( and its intensity is 6.5 dB
compared to the direct wave intensity. The 2
nd
maximum peak
is probably the diffraction wave from the edge of the
whiteboard. The 3
nd
maximum peak observed at
) / , / ( k k k k
y x
= ) 0 . 0 , 0 . 1 ( is a residual of the direct wave
PPWS. DOA of the reflecting wave is calculated by

|
|
.
|

\
|
=
x
y
k
k
arctan |

(13)

which is derived from (2). The estimated DOA is = | 133.
This value agrees with the direction of angle of image source
caused by specular reflection of whiteboard.
The photo taken by fish-eye lens camera is shown in Fig. 8.
By comparing with the Fig. 7(c), the whiteboard position and
the 1
st
maximum angle are coincide with each other. Therefore,
the reflection object causing the reflection wave is easily
found.


VNA
Coax.cable(30m)
3.0m
1.4m
Coax.cable(20m)
whiteboard
Smallbiconical
Antenna
(Receiving)
DoubleRidgedGuide
HornAntenna
(Transmitting)
2.0m
2.0m
x
y z
PC
Yaxis
Controller
Xaxis
Controller
GP-IB
Control
Trigger
signal
Controlcable
Control
cable
XYfieldscanner
Port2 Port1

Figure 3. Measurement setup for PPWS

XYfieldscanner
Smallbiconical
Antenna
Double RidgedGuide
HornAntenna
z
y x


Figure 4. XY field scanner. (This photo was taken at the different site from the
measurement site in this paper)

1 2 3 4 5 6
0
5
10
15
Front
Left
Right
Center
SVSWR Limit
Frequency [GHz]
S
V
S
W
R

[
d
B
]
1 2 3 4 5 6
0
5
10
15
Front
Left
Right
Center
SVSWR Limit
Frequency [GHz]
S
V
S
W
R

[
d
B
]
(a) Withwhiteboard
(b) Nowhiteboard(reference)

Figure 5. SVSWR of site (a) with whiteboard and (b) without whiteboard

x [mm]
y

[
m
m
]


3000 3500 4000
-600
-400
-200
0
200
400
600
S
2
1

[
d
B
]
-70
-65
-60
-55
-50
-45
x [mm]
y

[
m
m
]


3000 3500 4000
-600
-400
-200
0
200
400
600
S
2
1

[
d
B
]
-70
-65
-60
-55
-50
-45
(c)Difference
x [mm]
y

[
m
m
]


3000 3500 4000
-600
-400
-200
0
200
400
600
D
i
f
f
e
r
e
n
c
e

[
d
B
]
-66
-64
-62
-60
-58
-56
-54
(b) Nowhiteboard(reference)
(a) Withwhiteboard

Figure 6. Measured target S
21
field, reference S
21
field and the differential S
21

field at 4 GHz. The black dotted line is the test circle.
kx/k
k
y
/
k


-1 -0.5 0 0.5 1
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
N
o
r
m
a
l
i
z
e
d

S
p
e
c
t
r
u
m

[
d
B
]
-110
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
kx/k
k
y
/
k


-1 -0.5 0 0.5 1
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
N
o
r
m
a
l
i
z
e
d

S
p
e
c
t
r
u
m

[
d
B
]
-110
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
kx/k
k
y
/
k


-1 -0.5 0 0.5 1
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
D
i
f
f
e
r
e
n
c
e

[
d
B
]
-120
-100
-80
-60
-40
-20
1
st
peak(6.5dB)
2
nd
peak(21.0dB)
3
rd
peak(28.5dB)
(a) Withwhiteboard
(b) Nowhiteboard(reference)
(c) Difference

Figure 7. PPWS of measured target S
21
field, reference S
21
field and the
differential S
21
field at 4 GHz. The black dotted line is the circle where the
square root of (k
x
/k)
2
+(k
y
/k)
2
equals one.

Whiteboard
Horn
Antenna


Figure 8. Photo taken by fish-eye lens camera located at center position of the
test circle.

IV. CONCLUSIONS
In this paper, we proposed a new evaluation technique for
anechoic chamber using plane-wave spectral decomposition
and a fish-eye lens camera. From the wave number of the peak
in the PPWS distribution of differential field, the DOA of the
reflecting wave can be estimated. In addition, by comparing the
photo taken by fish-eye lens camera and the differential PPWS
2D-image, the cause of the reflection is easily found out.
REFERENCES

[1] J. Appel-Hansen, Reflectivity Level of Radio Anechoic Chambers,
IEEE Trans. on Ant. and Prop., vol.AP-21, no.4, July 1973.
[2] B. B. Tian, Free Space VSWR Method for Anechoic Chamber
Electromagnetic Performance Evaluation, MI Technologies, Technical
Papers, Nov.2008.
[3] CISPR 16-1-4, Specification for radio disturbance and immunity
measuring apparatus and methods Part 1-4: Radio disturbance and
immunity measuring apparatus Ancillary equipment Radiated
disturbances, Edition 2.1, 2008-01.
[4] C. A. Balanis, Antenna Theory: Analysis and Design 3
rd
ed., John Wiley
& Sons, New York, 2005.
[5] Oppenheim, A.V., and R.W. Schafer, Discrete-Time Signal Processing,
Upper Saddle River, NJ: Prentice-Hall, 1999, pp. 468-471.
[6] K. Miyamoto, Fish Eye Lens, Journal of the Optical Society of
America, Vol.54, No.8, pp.1060-1061, Aug. 1964.
[7] Schwarzbeck Mess Elektronik, Microwave Biconical Broadband
Antenna SBA 9119, http://www.schwarzbeck.de/Datenblatt/K9119.pdf,
accessed on July 2011