High Frequency Design

THROUGH-WALL COUPLER

Figure 2 · A wall coupler with WLAN

antenna, installed on a brick wall.

Figure 3 · Electromagnetic field of

Figure 4 · Electromagnetic field of

a vertical cut for 5.3 GHz couplers

a vertical cut for 2.4 GHz couplers

number of time steps.

on a 25 cm concrete wall.

on a 25 cm concrete wall.

Experimenting with FDTD is not

unlike testing a two-port device. S-

parameters can be measured and

input and output impedances are

coupler is more efficient than the 2.4

what can be expected as coupling loss

directly shown. An external shunt

GHz coupler. Material losses are

from input port to output port. The

component may be added to the 50-

higher at 5.3 GHz, so the 5.3 GHz

actual numerical result for the mag-

nitude of S21 shows a loss of 23 dB.

ohm input and output terminals for

coupler losses are not equal to the 2.4

In Figure 4, the couplers are oper-

the device under test. Here, the

GHz couplers when the material

ating at 2.4 GHz, also through a con-

devices are a wall coupler pair, with

independent losses due to the higher

crete wall of 25 cm thickness.

patch antennas inside of metal hous-

frequency are deducted.

Noticeable by a blue and purple col-

ings, placed on opposite sides of a

The coupler system gain is com-

ored field surrounding the couplers is

building wall [3].

puted from the field data by the

the remaining stray radiation, which

It took numerous configurations

FDTD software into normalized

is 48 dB down from the transmission

to create a radiation focus into solid

power gain. Figure 5 shows this gain

power at the lower port. In these cou-

walls of typical construction thick-

as 0.03, equal to a loss of 15 dB. For a

plers, the metal plating of the sub-

nesses. We targeted concrete as one of

secure wired LAN environment, such

strate has been enlarged to resonate

the more lossy materials which also

power losses are unimportant and

at 2.4 GHz. All electric parts are

has a high dielectric constant (due to

may actually need to be further

removed from the display, including

its high sand and gravel content).

increased by an attenuator in order

the brown color of the substrates in

Figure 3 shows an xz-cut through

to make unwanted transmissions

order to show the fields inside. Here,

near-center of both couplers with

undetectable to the outside [3].

the numerical readout for magnitude

patches in each housing and a 25 cm

For an application like that shown

of S21 shows a signal loss of 15 dB

(~10 inches) concrete wall between

in Figure 2, the losses are overcome

from port to port.

them. The bottom coupler is trans-

by using a panel antenna with a gain

The 8 dB higher loss at 5.3 GHz is

mitting 5.3 GHz (0 dB, relative) to the

of 18 dBi, therefore, retaining a 3 dB

not a surprise. The higher frequency

upper coupler. The concrete wall loca-

signal gain for the inside access

causes a loss penalty of 20 log

tion is easily recognized by the high

point. Operating an inside antenna is

(5.3/2.4) = 6.9 dB. This relationship

wave periodicity of its field in con-

often not possible because the outside

can be found by using equation (3-54)

trast to the free space pattern above

signal's angle of incidence for con-

of Kraus [4], solved for equal aper-

the signal receiving coupler. A color

crete and brick reaches additional

tures (coupler openings). The result

scale represents the field magnitudes

losses of 6 dB at an angle of 40º to 50º

points to the fact that the 5.3 GHz

and provides a rough estimate of

from the vertical (normal) of the out-

60

High Frequency Electronics