82
IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 46, NO. 1, FEBRUARY 2004
Fig. 12.  SARs and radiation efficiencies as a function of chassis width with two chassis lengths (
101 mm and
61 mm) at (a) 900 and (b) 1800 MHz.
7 mm.
1 W.
Distance from head to phone
thus equal distances between the head and front side of the
duced compared to the values without hand. However, the SARs
in hands increase at the same time. At 900 MHz, when the dis-
chassis, are roughly equal regardless of the chassis thickness.
tance increases from 2 to 14.5 mm, the SARs in the head with
C. Distance Between the Head and Phone
hand1 are approximately 70%­60% and with hand2 90%­80%
of those without hand. Correspondingly, at 1800 MHz the SARs
The effect of the distance between the head and phone chassis
in the head with hand1 are approximately 95%­50%, and with
­14.5 mm on the SARs and radiation efficiency was
hand2 90%­70% of the values without hand. Hand1 reduces the
61 mm and
studied with two different chassis lengths,
radiation efficiency the most with the largest distances; the de-
101 mm, which were selected based on the chassis resonances
crease is roughly 4­6 dB at 900 MHz and 2­3 dB at 1800 MHz
(
41 mm,
3.6 mm).
­14.5 mm. The decrease due to hand2 is roughly
when
Predictably [10], the SARs decrease and the radiation
4 dB at 900 MHz and 2 dB at 1800 MHz.
efficiencies increase as the distance from the head to chassis
increases (Fig. 13). However, the results of Fig. 13 give novel
D. Hand Position
information, as it is noted that when the chassis is close to
resonant, the SARs decrease slower as a function of distance
The effect of the hand position relative to the longitudinal axis
than when the chassis is nonresonant. An exception to the
of the chassis on the SARs and radiation efficiency was studied
trend is the situation when the metal chassis touches the head,
for both hand models with
121 mm
41 mm
in which case the SARs are roughly equal with both chassis
3.6 mm . With that length, the contribution of the chassis wave-
lengths. The radiation efficiencies are roughly 1.5 dB lower at
mode is at its maximum at 900 and at its minimum at 1800 MHz
900 MHz and 0.5 dB lower at 1800 MHz with chassis close
(see Fig. 3). Only the results with hand2 are illustrated with fig-
to resonant lengths than with nonresonant chassis lengths,
ures. In Fig. 14, "0 mm" denotes the location of the upper edge
excluding the case when the chassis touches the head.
of the hand in the plane of the upper edge of the phone chassis,
When the effect of distance was studied with hand models,
and "120 mm" denotes the location of the upper edge of the hand
the location of the hand was fixed with respect to the chassis
in the plane of the lower edge of the chassis, i.e., the hand has
(see Fig. 2), i.e., the hand was also moved as a function of dis-
slipped off the chassis.
tance between the head and phone. Generally, with either one of
With both hand models at 900 MHz, the hand position has
the hand models, the behavior of the SARs in the head is similar
only a minor effect on the SARs in the head. Generally, the
to the case without hand (Fig. 13). The farther the chassis and
SARs in both hands decrease and the radiation efficiency in-
creases slowly as the hand moves farther away from the antenna
hand are from the head, the more the SARs in the head are re-