Small Antenna Measurements Using a GTEM Cell

Ping Hui

NOKlA MOBILE PHONES

Introduction

The cellular phone market has been growing rapidly across the world with over 200
million subscribers in China as in November 2002, 160 million in Europe, and 123
million in the USA [I]. One clear trend in the cellular phone technology over the last
decade has been the drastic reduction of the phone size and weight [2]. As a result,
the antenna size and volume have also been reduced significantly. A constant design
challenge is to design mobile handset antennas meeting electrical specifications such
as gain, bandwidth and SAR values while making antennas small size, low profile,
light weight, and better reliability.

In the process of antenna design, anechoic rooms covered with absorbing materials
and calibrated reference antennas are often used to measure antenna gain and
efficiency. The high installation cost and the space requirement have prevented the
anechoic rooms from being used widely in antenna measurements. As a less
expensive alternative, GTEM cells have been proposed for small antenna
measurements [3,4]. Since the field strength inside the GTEM cell is well correlated
to the input power, the antenna gain can be measured accurately provided that the
antenna size is not too big compared with the height of the GTEM cell at the
measurement point. In addition, .the GTEM cell has been accepted widely as a
sufficient tool in electromagnetic compatibility for both radiated susceptibility tests
and radiated emission tests. It is therefore feasible and desirable to measure antenna
gains, radiated sensitivity, radiated emission, and radiated susceptibility using a
GTEM cell.

In this work, two balanced dipole antennas in both the Cellular band and the PCS
band have been measured using a GTEM cell and the results have been compared
with those obtained using an anechoic room. It is pointed out that correct rotation
plane of the antenna under test (AUT) inside the GTEM cell is essential to produce
right antenna patterns. The effectiveness and limitation of antenna measurements
using a GTEM cell and associated measurement uncertainties will be discussed.

Antenna Gain Measurement Using a GTEM Cell

Since the electromagnetic wave propagates in the TEM wave inside the GTEM cell,
the electrical field strength is given by

where V,,, is the input voltage,e,,is the input power into the GTEM cell, Z, is the
characteristic impedance of the GTEM cell, and h is the distance between the septum
plate and the ground at the measurement point.

The effective aperture of an antenna can be expressed in terms of gain and

0-7803-7846-6/03/$l7.00 02003 IEEE 715

Authorized licensed use limited to: Manoel Barbin. Downloaded on October 13, 2009 at 13:02 from IEEE Xplore. Restrictions apply.
 2'
A, =-G
47c
Since the dominant mode for a GTEM is TEM, by invoking Eq. ( I ) , the output power
by the antenna is

For 50 B system and h = 1 m, we have

and the path loss of the GTEM from the input apex to the measurement point is
2
PL = -= -19.8
9 . 6 ~ ~
+ 2010gd(m) (dB) (5)
Therefore, the antenna gain using a GTEM cell can be calculated as
G =lOlog<,v, -lOlog<n -PL (dBi). (6)

Measurement Results
To characterize the reflection properties of the GTEM cell (ETS-541 I), we measured
the return loss of the empty GTEM cell over frequencies from 30 MHz to 4 GHz, as
shown in Figure 1. It can be seen that the retum loss is well below -20 d B over the
frequency spectrum. To characterize local characteristic impedance over the length of
the GTEM cell, we measured the step pulse responses of the GTEM cell using Time-
Domain Reflectometry (Service Option 001 of HP 8753ES Vector Network
Analyzer), as shown in Figure 2. Around the measurement point (h=l m) as marked
by cursor 4, which is about 4 meters away from the apex, the local impedance is 64 .2
Ohms. The deviation of the local impedance above 50 Ohms makes the apparent path
loss larger and the measured antenna gain lower. To correct for this impedance
deviation, the height of the septum should be increased a bit.

We measured antenna gains and patterns using a GTEM cell for two balanced dipole
antennas at 824 MHz (Cellular band) and 1.85 GHz (PCS band). The return losses of
the GTEM cell are -30 dB at 824 MHz and -27 dB at 1.85 GHz from Figure 1. The
input power to the AUT was first measured and the output power at the GTEM apex
was measured for different dipole rotation angles with respect to the vertical. A
coaxial rotary joint was used to facilitate the rotation of the AUT and to minimize the
movement of the feeding cable. The antenna gain can be determined from Eq. (6) and
the relevant path loss values. It is noted that the dipole antennas should be rotated in
the cross sectional plane of the GTEM, instead of in the horizontal plane as suggested
in Ref. [4]. When the arms of the dipole antenna are in the vertical position, the
electric fields of the GTEM cell and the dipole antenna are aligned, resulting in the
maximum output power at the apex. This is the position where the antenna gain can
be calibrated by a GTEM cell.

The antenna gain patterns obtained using the GTEM cell and the anechoic room are
shown in Figure 3 for 824 MHz and Figure 4 for 1.85 GHz. It is observed that at
frequency 824 MHz the ratio of the antenna gain maximum to the minimum is about
15 dB using the GTEM cell and about 25 dB using the anechoic room. This can be
attributed to the relative large size of the Cellular band dipole antenna (about 35 cm
long) as compared to the uniform field volume of the GTEM cell. At the PCS band
frequency of 1.85 GHz, the ratio of the antenna gain maximum to the minimum
obtained using the GTEM cell is about -25 dB, even 7 dB deeper than one measured
from the anechoic room. The antenna main lobes measured using both the GTEM cell
and the anechoic room agree well in the PCS band; whereas in the Cellular band, the
GTEM cell gives broader antenna main lobe than the anechoic room.
716

Authorized licensed use limited to: Manoel Barbin. Downloaded on October 13, 2009 at 13:02 from IEEE Xplore. Restrictions apply.
 There are three major uncertainties in antenna pattern measurements using a GTEM
cell: reflection from the matching termination, higher order GTEM waveguide
modes, and the leakage current on the feeding cable. A well-tuned GTEM cell should
be able to give around -25 dB return loss across the spectrum as shown in Figure 1.
For the GTEM cell used, the return loss could be further improved if the local
characteristic impedance is fine adjusted to SO Ohms. In the measurements, the AUT
is positioned about I meter away from the tips of the absorption materials and the
return loss from the matching termination would he expected worse than -25 dB due
to the near field effects. Nevertheless, since the AUT was rotated in the cross session
of the GTEM cell, the reflection from the matching termination should be remained
in the same level for all the angles.

The higher order waveguide modes excited by the AUT can propagate towards the
matching section, and will be cut off towards the input port of the GTEM cell [SI.
Due to the tapering structure, the GTEM cell behaves as a low-pass filter. The higher
order waveguide modes are responsible for the non-uniform fields in the test volume
inside the GTEM cell. As pointed out in Ref.[S], the higher order waveguide modes
propagating towards the apex have smaller amplitudes than the ones propagating
towards the matching section, and furthermore, convert to the fundamental TEM
mode nicely and completely. The higher order waveguide modes traveling towards
the matching section will not be absorbed as well as the TEM mode and thus are
reflected toward the apex of the GTEM cell, limiting the dynamic range of the
measurements.

The leakage current along the feeding cables gives. problems to antenna
measurements in both the GTEM cell and the anechoic room. The practical and
effective solutions to. reduce the leakage current are ferrite beads, baluns, and
reducing cable movements. .

Conclusions
It is feasible and economic to use GTEM cells to measure small antenna gain pattems
as well as to perform radiated sensitivity measurements, radiated emission tests, and
radiated susceptibility tests. A 6-meter long GTEM cell seems to be sufficient for
mobile phone applications. The main source of measurement uncertainties is the
reflection from the matching section for the TEM and the higher order waveguide
modes. One possible way to improve the dynamic range of the GTEM cell is to use
the time gating to remove the reflection from the matching section.

References
1. Various source of the Internet.
2. H. Morishita, Y . Kim, and K. Fujimoto, “Design concept of antennas for
small mobile terminals and the future perspective;” IEEE Antennas &
Propagation Magazine, vol. 44, Oct. 2002, pp. 3043.
3. C. Icheln, P. Vainikainen, P. Haapala, ‘:Application’of a GTEM cell to small
antenna measurements,” Proceedings of the 1997 IEEE AP-S International
Symposiunt and URSI North American Radio Science Meeting, Montreal,
Canada, July. 1997, pp. 546-549.
4. C. Icheln, “Method for measuring RF radiation properties of small antennas,”
Ph.D. dissertation, Helsinki University of Technology, Espoo, Finland, Oct.,
2001.
5 . R.D. Leo, L. Pierantmi, T. Rozzi, and L. Zappelli. “Dipole excitation and
scattering by spherical objects in GTEM cell,” IEEE Trans. Microwave
T h e o y Tech.. vol. MTT42, Sept. 1994, pp. 1700-1708.

717

Authorized licensed use limited to: Manoel Barbin. Downloaded on October 13, 2009 at 13:02 from IEEE Xplore. Restrictions apply.
 -.--_I_ l.._l_._._
............. ..LI ........

ne --* '3- I...II...*. .,....S! ,. IIV .... -

-
5
4 -20
-15

0 90 180 270 360

Figure 3 Antenna pattem of a dipole at 824 MHz

-5

-10
Io
c -15
1

c -20
0 90 180 270 360

Figure 4 Antenna pattem of a dipole at I .85 GHz

718

Authorized licensed use limited to: Manoel Barbin. Downloaded on October 13, 2009 at 13:02 from IEEE Xplore. Restrictions apply.