electronics

Article
Design of a Low-Cost Microstrip Directional Coupler
with High Coupling for a Motion Detection Sensor
Giovanni Sanna 1 , Giorgio Montisci 2, * ID
, Zushen Jin 3 , Alessandro Fanti 2 ID
and
Giovanni Andrea Casula 2
1 R&D Labs, CIAS Elettronica, 20158 Milano, Italy; sanna.giovanni@cias.it
2 Dipartimento di Ingegneria Elettrica ed Elettronica, Università degli Studi di Cagliari, 09123 Cagliari, Italy;
alessandro.fanti@diee.unica.it (A.F.); a.casula@diee.unica.it (G.A.C.)
3 EMC Research and Measurement Center of Navy, Shanghai 200235, China; jin_hexi@126.com
* Correspondence: giorgio.montisci@unica.it; Tel.: +39-070-675-5780

Received: 2 January 2018; Accepted: 22 February 2018; Published: 24 February 2018

Abstract: A coupled-line coupler based on the asymmetric cascade connection of two coupled line
sections is used to achieve a high coupling factor using low-cost material and technology in the
X frequency band, and its performance is compared with a standard quarter-wave, coupled-line
coupler, showing an increase in the coupling factor of about 3 dB. The proposed coupler can be
effectively used in a Doppler motion detection sensor, due to its strong coupling and relatively
high isolation. The coupler is designed through a simple, yet rigorous, equivalent circuital model.
Then, an optimization procedure was performed using the commercial software Ansys HFSS in order
to compensate for losses and second order effects. A prototype of the designed coupler was realized,
and the measured data show very good agreement with simulations.

Keywords: coupled lines; directional coupler; motion detection systems

1. Introduction
The aim of motion detection systems (MDS) is the detection of the movement of scattering
objects. MDS are normally used for security purpose, such as to protect indoor areas, and locations
where important assets are located (museums, libraries), or in automotives to avoid car collisions.
The detection of a movement usually triggers an alarm, or, more generally, activates other
electronic subsystems.
MDS use different methods to detect movement, and can be divided into two main categories:
passive detectors and active detectors. Passive detectors (for example, most infrared detectors)
do not send out signals but merely receive signals, such as changes in brightness temperature,
light intensity and so on. On the other hand, active detectors (such as microwave detectors) act as
a radar, i.e., they transmit energy that propagates in the protected area, it is scattered by the existing
bodies, and then it is received back by the MDS. In this case, for example, an alarm can be triggered by
any perturbation of the reflected waves caused by a moving object or by any movement of a body with
a Radar Cross Section (RCS) larger than a prescribed threshold (e.g., to detect humans but prevent the
detection of mice). A block diagram of a microwave MDS is shown in Figure 1. The microwave part of
the MDS is made by a voltage-controlled oscillator, a low pass filter, a transmitting and a receiving
antenna, a mixer and a directional coupler. The directional coupler is required to perform coherent
detection of the received signal, thus allowing a larger receiver sensitivity and then a larger detection
range. Thus, this coupler requires both strong coupling and large isolation. On the other hand, it should
be compact and low-cost. As such, the directional coupler is one of the most critical components of
a microwave MDS.

Electronics 2018, 7, 25; doi:10.3390/electronics7020025 www.mdpi.com/journal/electronics

Electronics 2018, 7, 25 2 of 7
Electronics 2018, 7, 25 2 of 7

Figure
Figure 1. Block diagram
1. Block diagramof
ofthe
thewhole
wholeMURENA
MURENAsensor
sensor(the
(themicrowave
microwavesubsystem
subsystemisisincluded
includedinina
adashed
dashedbox).
box).

In this paper we focus on the design of the directional coupler employed to feed the mixer for
the “MURENA” Doppler motion detection sensor (DMDS) [1]. MURENA MURENA is is a volumetric X band
microwave sensor,sensor,using usingthe theDoppler
Doppler effect
effectand anddigital signal
digital processing,
signal processing, for external
for externalprotection. Like
protection.
all sensors
Like usingusing
all sensors microwavemicrowavetechnology
technologyto detect motion,
to detect MURENA
motion, MURENA exploits the principle
exploits the principleof a
Doppler
of a Dopplerradar, where
radar, a continuous
where a continuous wavewave of microwave
of microwave radiation is emitted
radiation by a by
is emitted RF asource. The
RF source.
frequency shift in the reflected waves, due to the motion of an
The frequency shift in the reflected waves, due to the motion of an object toward (or away from) object toward (or away from) the
receiver,
the results
receiver, in ain signal
results a signal at atlow
lowfrequency.
frequency.The TheMURENA
MURENAinterface interface between
between the the microwave
transmitter–receiver is is designed
designedusing usinga aspecial
specialcircuit,
circuit,which
which generates
generates a modulation
a modulation signal on the
signal on
transmitter,
the transmitter,demodulating
demodulating and and
pre-amplifying
pre-amplifying the received
the received signal usingusing
signal two separate
two separate channels. This
channels.
performance
This performance is thenisenhanced by subsequent
then enhanced signal processing
by subsequent which, after
signal processing amplification,
which, is digitized
after amplification,
using
is an analogue/digital
digitized convertor. This
using an analogue/digital digital This
convertor. signal is handled
digital signal is in handled
real timeinby realthetimeon-board
by the
microprocessor.
on-board Using appropriate
microprocessor. algorithms
Using appropriate it is possible
algorithms it is to measure
possible the distance
to measure and the mass
the distance and theof
the object
mass of themoving within the
object moving detection
within area, analyzing
the detection all the received
area, analyzing data using
all the received data ausing
“fuzzy” logic.
a “fuzzy”
FigureFigure
logic. 1 reports a block
1 reports diagram
a block diagram of the whole
of the whole sensor.
sensor.TheThe mixer
mixer usedusedininthis
thissystem
systemisis aa single
balanced 180° ◦
180 “rat-race”
“rat-race” type,
type, which,
which, in in order
order to work properly, requires a DC-isolated DC-isolated directional
coupler with a large coupling, typically typically larger
larger thanthan − −77dB.
dB.TheTheoperating
operatingfrequency
frequency range,
range, 10–11
10–11 GHz,
GHz,
was chosen to comply with different different national
national regulations.
regulations.
The most popular configuration of a directional coupler is the standard coupled-line directional
coupler (SC) [2–4], which allows the realization of quite a compact coupler coupler withwith good
good bandwidth.
bandwidth.
Moreover, the coupled port is DC isolated isolated from the the through
through line. line. On the other hand, the strongest
coupling factor of a standard coupler realization is limited to about 9–10 dB [2]. Such values are too
small for our application (and for many similar applications). Increasing the coupling requires very
close coupling lines, which can be realized only using high-cost high-cost processing
processing over over expensive
expensive PCB PCB [5].
[5].
As an analternative,
alternative, stronger
stronger coupling
coupling factors
factors can be can be obtained
obtained with more with more complex
complex (and bigger) (and bigger)
structures,
structures,
such as the such
Langeascoupler
the Lange coupler
[2], or other [2], or other
special special[6,7],
techniques techniques [6,7], not
not compatible compatible
with the system with the
target
systemTherefore,
price. target price. Therefore,
a compact a compact
coupler coupler able
configuration, configuration,
to realize aable to realize
relatively stronga relatively
coupling strong
factor,
coupling factor, with low-cost material and technology (and therefore
with low-cost material and technology (and therefore using large inter-line spacing), such as those using large inter-line spacing),
such as by
needed those needed
a MDS, by a MDS,
requires requires
a different a different configuration.
configuration.
Simple and andcompact
compactcoupler couplerconfigurations,
configurations, based
based onon thetheasymmetric
asymmetric cascade
cascade connection
connectionof twoof
coupled line sections, are proposed in [8] in the lower UHF frequency
two coupled line sections, are proposed in [8] in the lower UHF frequency band (600 MHz) in order band (600 MHz) in order to
improve
to improve flexibility
flexibilityin the design
in the of couplers
design of couplers for complex
for complexcircuit environments,
circuit environments, within a limited
within area.
a limited
In thisIn
area. paper, we usewe
this paper, essentially the same
use essentially theconfiguration
same configurationpresented in [8], but
presented infor
[8],a but
different
for a purpose,
different
i.e., in order
purpose, i.e.,toinachieve
order to tight coupling
achieve tightand high directivity,
coupling compared to
and high directivity, the SC [2],tointhe
compared theSCX frequency
[2], in the
band,
X using low-cost
frequency band, using materials and materials
low-cost technology. andHowever,
technology. the improvement
However, the is improvement
not straightforward, is not
since increasing the coupling factor using low-cost material in a far higher frequency range, where
dielectric and ohmic losses could be critical, can be challenging. We will demonstrate that the selected
Electronics 2018, 7, 25 3 of 7

straightforward, since increasing the coupling factor using low-cost material in a far higher frequency
range, where dielectric and ohmic losses could be critical, can be challenging. We will demonstrate
that the selected configuration is able to strengthen the coupling factor over that of a SC with the
sameElectronics
inter-line spacing. This improvement is obtained at the expense of the total coupler
2018, 7, 25 3 of 7
size.
However, as shown in Section 3, this size increase does not significantly affect the coupler loss
efficiency. In fact, is
configuration inable
our tocase, the actual
strengthen the coupling regionover
coupling factor increases
that of only
a SC by 10%
with thecompared to the SC
same inter-line
90◦ to 100
(fromspacing. This◦ ).improvement is obtained at the expense of the total coupler size. However, as shown
in Section
The design3,of this
thesize increase does
directional not was
coupler significantly
performed affect the coupler
through loss efficiency.
a simplified circuitalInmodel,
fact, in and
our then
case, the actual coupling region increases only by 10% compared to the SC (from
optimized by using the commercial software HFSS by Ansys. A prototype has been realized by R&D 90° to 100°).
The design of the directional coupler was performed through a simplified circuital model, and
Labs of CIAS Elettronica, using a low-cost laminate. Measured results on the realized prototype are in
then optimized by using the commercial software HFSS by Ansys. A prototype has been realized by
very good agreement with simulations, allowing the achievement of the coupling factor required by
R&D Labs of CIAS Elettronica, using a low-cost laminate. Measured results on the realized prototype
the design
are in specifications.
very good agreement The proposed directional
with simulations, coupler
allowing thewas also compared
achievement with a SC
of the coupling with the
factor
samerequired
inter-line spacing, showing its ability to increase the coupling factor.
by the design specifications. The proposed directional coupler was also compared with a SC
with the same inter-line spacing, showing its ability to increase the coupling factor.
2. Analysis of the Proposed Configuration
2. Analysis of the Proposed Configuration
The scheme of the proposed coupler is shown in Figure 2a. Networks A and D are two equal
The sections
coupled line scheme of(CLSs)
the proposed coupler
acting as is shown
directional in Figure
couplers 2a.port
with Networks A and DZ0are
impedances twoΩ,
= 50 equal
whereas
coupled line sections (CLSs) acting as directional couplers with port impedances Z =
B and C are microstrip transmission lines with characteristic impedances Z0 . Let ϑB and ϑC denote
0 50 Ω, whereas
B and C are
the electrical microstrip
lengths transmission
of the lineslines
transmission with characteristic
B and C; ϑB impedances Z0. Let to
should be equal ϑB the
and required
ϑC denote the
spacing
electrical lengths of the transmission lines B and C; ϑB should be equal to the required spacing
between the two CLSs, whereas ϑC is usually chosen to be longer than ϑB , as line C can be realized as
between the two CLSs, whereas ϑC is usually chosen to be longer than ϑB, as line C can be realized as
a meander line. The microstrip realization of the proposed coupler is reported in Figure 2b.
a meander line. The microstrip realization of the proposed coupler is reported in Figure 2b.

(a) (b)

Figure 2. Structure of the proposed coupler: (a) Circuital model; (b) Microstrip realization, including
Figure 2. Structure of the proposed coupler: (a) Circuital model; (b) Microstrip realization, including test
test ports (dashed box shows the actual coupler size).
ports (dashed box shows the actual coupler size).
Since we are interested in achieving the strongest coupling factor available, the inter-line spacing
S is the
Since wesmallest spacingincompatible
are interested achieving with the chosen
the strongest technology.
coupling Then,
factor the matching
available, condition
the inter-line spacing
(ZeZsmallest
S is the o)/(Zo2) = 1spacing
[2] must compatible
be enforced, wherein
with theZechosen
, Zo are the even and odd
technology. mode
Then, theimpedances
matching ofcondition
the
(Ze Zocoupled
)/(Zo 2 ) =line section.
1 [2] mustAsbea consequence, the width
enforced, wherein Ze W
, Z(see Figure 2b) is also completely determined [2].
o are the even and odd mode impedances of the
We further assume, for the moment, that the CLSs are ideal, i.e., that the propagation constants
coupled line section. As a consequence, the width W (see Figure 2b) is also completely determined [2].
of both the even and the odd mode are equal, and we denote this constant by βC. In this case, we have
We further assume, for the moment, that the CLSs are ideal, i.e., that the propagation constants of
all ports matched and an infinite isolation over the whole frequency range [2].
both the even
Letting andϑAthe
= βCodd
LA bemode are equal,
the electrical andofwe
length thedenote thisitsconstant
first CLS, scatteringbymatrix
βC . Incan
this
becase, we have
obtained by all
ports[2]
matched and an infinite isolation over the whole frequency range [2].
Letting ϑA = βC LA be the electrical length of the first CLS, its scattering matrix can be obtained
by [2] 1 k 2
S21  T0 (A )  √ (1)
1  k 2 cos(1A− ) k2j sin(A )
S21 = T0 (ϑ A ) = √ (1)
1 − k2 cos(ϑ A ) + j sin(ϑ A )
Electronics 2018, 7, 25 4 of 7
Electronics 2018, 7, 25 4 of 7

jk sin( )
K0K(0ϑ(A)A )= √
31 
jk sin(ϑA )
S31S= A (2) (2)
cos(ϑAA))+j sin(
11 −kk2 cos( j sin(ϑA )A )
2

wherewhere
k = (Z −e Z
k =e (Z − oZ)/(Z
o)/(Z +Z
e e+ Zoo))isisthe
themaximum
maximum of the
of coupling coefficient
the coupling S31 achieved
coefficient for ϑA = π/2.
S31 achieved forThen,
ϑA = π/2.
Then,the
thetotal coupling
total coupling factor of the
factor ofcoupler in Figure
the coupler 2 is derived
in Figure in ([8], Equation
2 is derived in ([8], (4)), and can(4)),
Equation be written
and can be
as as
written  
 1 +  T 2 − K2  exp[− j(ϑ + ϑ )] 
C = K0 1  To 2 K o  exp[ j (B  C )] 
o 2 o
2 B C 
(3)

C  K 0 1 − Ko2exp[− j(ϑB + ϑC )] (3)
1  K o exp[ j (B  C )]


Now, k is determined by the CLS geometry, so C is only a function of ϑA and of ϑB + ϑC .

Therefore,Now,the kindividual
is determined by the CLS
lengths geometry,
ϑB and so Cbe
ϑC can is only a function
chosen of ϑAwith
to cope and of
theϑB +size
ϑC. Therefore,
requirements.
the individual lengths ϑ B and ϑC can be chosen to cope with the size requirements. In Figure 3 we
In Figure 3 we show the behavior of C as a function of ϑB + ϑC (for different ϑA values) for k = −9.6 dB.
Fromshow the behavior
the results of C3,as
in Figure it afollows
function of the
that ϑB +optimal
ϑC (for different
value ofϑϑABvalues)
+ ϑC isfor k = −9.6
around 280dB. From
◦ for theCLSs,
short
results in Figure 3, it follows that the optimal value of ϑB + ϑC is around 280° for short CLSs, and
and becomes smaller as the length of the CLS increases. Moreover, the longer the ϑA , the stronger the
becomes smaller as the length of the CLS increases. Moreover, the longer the ϑA, the stronger the
optimal coupling factor.
optimal coupling factor.

Figure 3. Coupling C of the directional coupler in Figure 2.

Figure 3. Coupling C of the directional coupler in Figure 2.
3. Coupler Design and Testing
3. Coupler Design and Testing
The proposed coupler can be used in many different microwave systems. However, in order to
experimentally
The assess thecan
proposed coupler proposal,
be used weindecided to realize microwave
many different and test a coupler
systems.to be used in X-band
However, in order to
experimentally assess the proposal, we decided to realize and test a coupler to be used inand
DMDS. The requirements are therefore a strong coupling (C ≥ −7 dB) between the through line X-band
DMDS. theThe
coupling port overare
requirements thetherefore
bandwidth 10–11 GHz,
a strong couplingusing(Ca low-cost
≥ −7 dB) PCB and process.
between Therefore,
the through line we
and the
selected an organic–ceramic laminate, with εr = 3.5 and tan(δ) = 0.003. The dielectric slab thickness
coupling port over the bandwidth 10–11 GHz, using a low-cost PCB and process. Therefore, we selected
was 0.762 mm and the copper thickness was 35 μm. The smallest spacing available on this PCB with
an organic–ceramic laminate, with εr = 3.5 and tan(δ) = 0.003. The dielectric slab thickness was 0.762 mm
a standard (cheap) process was 0.15 mm, and this value was used for the gap S between the coupled
and the copper thickness was 35 µm. The smallest spacing available on this PCB with a standard
lines.
(cheap) process wasthe
Using [9], 0.15 mm, and
matching this value
condition waso2)used
(ZeZo)/(Z for the
= 1 results W =S
ingap 1.4between
mm. Then, thethe
coupled
even andlines.
odd
2 = 1 results in W = 1.4 mm. Then, the even and odd
Using [9], the
impedances matchingusing
(computed condition
HFSS) (Zaree Z
Zoe )/(Z o )Ω,
= 69.8 Zo = 35.2 Ω, leading to k = −9.6 dB.
impedances The(computed using
main available HFSS)
degree are Ze =is69.8
of freedom Ω, Zo =ϑA35.2
the length Ω,CLSs.
of the leading
It isto k = −9.6
apparent dB.Figure 3
from
that the longer the ϑ , the larger both C and the increase in coupling, C/k.
The main available degree of freedom is the length ϑA of the CLSs. It is apparent from Figure
A However, increasing ϑA3 that
produces larger losses in the CLSs and a trade-off is required between C/k on the
the longer the ϑA , the larger both C and the increase in coupling, C/k. However, increasing ϑA produces one hand, and total
largercoupler
losses size
in theand losses
CLSs and onathe other. We
trade-off selected ϑbetween
is required A = 50° as a value, giving C/k = 3.3 dB, aiming at C
C/k on the one hand, and total coupler size
≥ −7 dB over the operating bandwidth. From ◦ ϑ A = 50° and choosing ϑB = 90°, we get, from Figure 3,
and losses on the other. We selected ϑA = 50 as a value, giving C/k = 3.3 dB, aiming at C ≥ −7 dB over
the starting value of ϑC = 160°.
the operating bandwidth. From ϑA = 50◦ and choosing ϑB = 90◦ , we get, from Figure 3, the starting
value of ϑC = 160◦ .
An optimization is then needed for ϑC (i.e., on LC2 in Figure 2, since LC1 is fixed by LB ) to
compensate for losses and second-order effects. As a matter of fact, the bend angles of the meander
Electronics 2018, 7, 25 5 of 7

line (αC2 ) and the bends towards the coupler ports (αF1 ) also need to be optimized (see Figure 2).
Moreover, in order to avoid the tapering of the microstrip lines B and C in connection with the CLSs
(W = 1.4 mm, whereas a 50 Ω microstrip line has width W0 = 1.7 mm), in the optimized configuration
these lines were selected with the same width W = 1.4 mm as the CLSs. Since this value corresponds to
a characteristic impedance of about 56 Ω, we can assume that this approximation does 5not
Electronics 2018, 7, 25 of 7
affect the
required coupler behavior. is then needed for ϑC (i.e., on LC2 in Figure 2, since LC1 is fixed by LB) to
An optimization
To perform
compensatethe optimization, the coupler
for losses and second-order of Figure
effects. 2a was
As a matter simulated
of fact, the bendusing
anglesHFSS. The optimization
of the meander
line (α C2) and the bends towards the coupler ports (αF1) also need to be optimized (see Figure 2).
parameters are: LC2 , the angles αC2 (determining the electrical length ϑC ), and the angles αF1 of the
Moreover, in order to avoid the tapering of the microstrip lines B and C in connection with the CLSs
bends in the feeding lines. The geometrical parameters of the optimized geometry, obtained after
(W = 1.4 mm, whereas a 50 Ω microstrip line has width W0 = 1.7 mm), in the optimized configuration
a short trialthese
andlines
errorwereprocedure
selected with LC2same
on the , αC2width
, andW αF1 , are
= 1.4 mmsummarized in Table
as the CLSs. Since 1 (optimized
this value corresponds parameters
are in bold).
to a The last bend
characteristic in theoffeeding
impedance about 56lines, before
Ω, we can each
assume thatfeeding port, must
this approximation doesleave room for the
not affect
soldering the required
of SMA PCBcoupler behavior. whereas the length L (= 3 mm) was to be as short as possible,
connectors, F2
To perform the optimization, the coupler of Figure 2a was simulated using HFSS. The
provided that it did not affect the coupler behavior.
optimization parameters are: LC2, the angles αC2 (determining the electrical length ϑC), and the angles
αF1 of the bends in the feeding lines. The geometrical parameters of the optimized geometry, obtained
1. Geometrical
Table after a short trial parameters of the optimized
and error procedure on LC2, αC2,coupler
and αF1,(see Figure 2). The
are summarized optimized
in Table parameters
1 (optimized
parameters
are reported are in bold). The last bend in the feeding lines, before each feeding port, must leave room
in bold.
for the soldering of SMA PCB connectors, whereas the length LF2 (= 3 mm) was to be as short as
possible, provided that it did not affect the coupler behavior.
Parameter Value
Table 1. Geometrical parameters of W
the0optimized coupler1.7
(seemm
Figure 2). The optimized parameters
are reported in bold. W 1.4 mm
S 0.15 mm
Parameter Value
LA 2.6 mm
W0 1.7 mm
LB 4.3 mm
W 1.4 mm
LC1 1.5 mm
S 0.15 mm
LC2 1.6 mm
LA 2.6 mm
αC1 45 degrees
LB 4.3 mm
αC2 20 degrees
LC1 1.5 mm
αF1 43 degrees
LC2 1.6 mm
LF1 7.45 mm
αC1 45 degrees
αF2 αC2 45 degrees
20 degrees
αF1 43 degrees
LF1 7.45 mm
A prototype of the designed coupler was realized by
the R&D Labs of CIAS Elettronica
αF2 45 degrees
(see Figure 4, right). For comparison, a SC with the same inter-line spacing S was also realized
(Figure 4, left).AIn Figureof5the
prototype thedesigned
simulated (HFSS)
coupler and by
was realized measured frequency
the R&D Labs response(see
of CIAS Elettronica of Figure
the couplers in
4, right).
Figure 4 are For comparison,
reported. a SC withbetween
The agreement the same inter-line spacing Sand
the simulated was experimental
also realized (Figure
data4, is
left).
veryIn good for
Figure 5 the simulated (HFSS) and measured frequency response of the couplers in Figure 4 are reported.
both prototypes.
The agreement between the simulated and experimental data is very good for both prototypes.

Figure 4. Photo of the realized prototypes. Right: directional coupler of Figure 2 with the parameters
Figure 4. Photo
of Tableof the Standard
1; left: realizedquarter-wave,
prototypes.coupled-line-directional
Right: directional coupler of with
coupler (SC) Figure 2 with
the same the parameters
inter-line
of Table 1;spacing
left: Standard
S. quarter-wave, coupled-line-directional coupler (SC) with the same inter-line
spacing S.
Electronics 2018, 7, 25 6 of 7
Electronics 2018, 7, 25 6 of 7

Electronics 2018, 7, 25 6 of 7

(a) (b)

Figure5.5.Simulated
Figure Simulatedandand measured
measured frequency
frequency response:
response: (a) standard
(a) standard quarter-wave,
quarter-wave, coupled-line-
coupled-line-directional
directional coupler of Figure 4, left; the simulated S41 is below −40 dB (b) proposed directional
coupler of Figure 4, left; the simulated S41 is below −40 dB (b) proposed directional coupler coupler
of
of Figure 4, right.
Figure 4, right.

From the results in Figure 5b (a) it appears that the response of (b)the directional coupler in Figure 4
From the results in Figure 5b it appears that the response of the directional coupler in Figure 4
(right) is coherent
Figure 5. with the simplified
Simulated and measured analysis based on
frequency response: Equation
(a) standard (3) and coupled-line-
quarter-wave, fully complies with the
(right) is coherent with the simplified analysis based on Equation (3) and fully complies with the design
design specifications, since ofCFigure
directional coupler ≥ −74,dB left;in
thethe operating
simulated bandwidth.
S41 is below Moreover,
−40 dB (b) proposed thecoupler
directional resulting coupling
specifications, of since C ≥ −
Figure 4, right. 7 dB in the operating bandwidth. Moreover, the resulting coupling factor at
factor at the center frequency (6.1 dB—measured data) is about 2.9 dB stronger than the standard
the center frequency (6.1 dB—measured data) is about 2.9 dB stronger than the standard quarter-wave
quarter-waveFrom coupler on the
the results same5bPCB
in Figure and that
it appears withthethe sameofinter-line
response the directionalspacing
coupler(see Figure
in Figure 4 5). From
coupler on the same PCB and with the same inter-line spacing (see Figure 5). From Figure 5, it also
Figure 5, it alsoisfollows
(right) coherentthat withthe the directivity
simplified analysis(measured
based on data) of the
Equation (3)proposed couplerwith
and fully complies (28–30
the dB in the
follows that thespecifications,
directivitysince (measured data) of the proposed coupler (28–30 dB coupling
in the operating
operatingdesign
bandwidth) is aboutC 7≥ −7 dBdBlarger
in the operating
than the bandwidth. Moreover,
directivity of the the SC.
resulting
Therefore, the initial
bandwidth) is about
factor 7 dB frequency
at the center larger than the directivity
(6.1 dB—measured data)ofisthe SC.2.9Therefore,
about dB stronger thethaninitial hypothesis on
the standard
hypothesis on the equality of the phase velocities can be accepted, and no compensation for the
the equality quarter-wave
of the phase coupler on the same
velocities can PCB and with the
be accepted, and samenointer-line
compensationspacing (see
for Figure 5). From in those
the difference
differenceFigure
in those speeds
5, it also followsis that
needed [10]. (measured data) of the proposed coupler (28–30 dB in the
the directivity
speeds is needed [10].
Finally, we wanted
operating bandwidth) to evaluate
is about 7the dB loss efficiency
larger of the modified
than the directivity of the SC. coupler
Therefore, in Figure
the initial2 compared
Finally, we wanted toequality
evaluate thephase
loss efficiency of be
theaccepted,
modified coupler in Figurefor2 compared
the columnto
to a SC. Inhypothesis
order toon dothe this, of the
we observe that,velocities can
for a lossless network,and thenodot
compensation
product of any of
a SC. In order to dointhis,
difference thosewe speedsobserve
is needed that, for a lossless network, the dot product of any column of its
[10].
its scattering matrix with the conjugate of that column is equal to one [2]. Therefore, the loss efficiency
scattering matrix Finally,
withwethe wanted to evaluate
conjugate of thatthe loss efficiency
column of theto
is equal modified
one [2]. coupler in Figure
Therefore, 2 compared
2 the loss 2 efficiency can2
 2 S11of any  Scolumn
221  of S312  S41 2 ,
2
can be computed evaluating
to a SC. In order to do this, how much that,
we observe the parameter η, defined
for a lossless network, the dotas product
be computed evaluating how much the parameter η, defined as η = |S11 | + |S21 | + |S31 | + |S41 | ,
its scattering matrix with the conjugate of that column is equal to one [2]. Therefore, the loss efficiency
deviates from
deviates fromunity.
unity. In In Figure
Figure 6,6, we weplot plotthe theparameter
parameter ηηfor forboth
boththe the2modified
modified andthe
and the standard
standard
can be computed evaluating how much the parameter η, defined as   S11  S21  S31  S41 ,
2 2 2

coupler.The
coupler. Theaverage
average ηover overthe thebandwidth
bandwidth10–11 10–11GHz GHzfor forthe
theSCSCisisabout
about88%,88%,whereas
whereasthe theaverage
average
deviates from ηunity. In Figure 6, we plot the parameter η for both the modified and the standard
 of the modified
η of the modified
coupler. The
coupler
coupler
average η
is about
isover
about 84%. This
84%. This
the bandwidth
difference
difference
10–11
is mainly
GHz for theisSC mainly due
is aboutdue
to the
88%, to
overall
the overall
whereas
longer
the average longer (100°
CLS CLS
compared
(100 ◦  to
compared 90°),
of the 90since
◦
tomodified the the
coupler
), since losses
is losses
aboutare84%.concentrated
are This differenceatisat
concentrated andandaround
mainly due to the
around the gaps
overall
the of
of the
gapslonger CLS
the CLSs, where the
(100°where
CLSs, the
surface current
compared density
to 90°),issince
surface current density is stronger. stronger.
the losses are concentrated at and around the gaps of the CLSs, where the
surface current density is stronger.

Figure 6. Comparison (loss efficiency η) between the proposed coupler, and the standard coupler with
Figure 6. Comparison (loss
the same inter-line efficiency η) between the proposed coupler, and the standard coupler with
spacing.
the same inter-line spacing.
Figure 6. Comparison (loss efficiency η) between the proposed coupler, and the standard coupler with
the same inter-line spacing.
Electronics 2018, 7, 25 7 of 7

4. Conclusions
We presented a coupled-line microstrip coupler for high coupling over a relatively large
bandwidth (10%). This coupler was designed and realized using a low cost PCB in the X frequency
band, with standard technology, and allows us to achieve stronger coupling over a standard
coupled-line directional coupler with the same inter-line spacing. It presents a simple and compact
structure compared to other configurations used to increase the coupling factor, and with its strong
coupling and relatively high isolation, can be effectively used in a Doppler motion detection sensor.

Author Contributions: All the authors contributed to the design of the directional couplers, Giovanni Sanna and
Giorgio Montisci performed the experimental characterization of the realized prototypes; Giorgio Montisci wrote
the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Cias Elettronica Srl. Available online: http://www.cias.it/en/products/sensors/murena/ (accessed on
2 January 2018).
2. Collin, R.E. Foundations for Microwave Engineering, 2nd ed.; McGraw–Hill: New York, USA, 1992; Section 6.4;
ISBN 0-7803-6031-1.
3. Valente, G.; Montisci, G.; Mariotti, S. High–performance microstrip directional coupler for radioastronomical
receivers at cryogenic temperature. Electron. Lett. 2014, 50, 449–451. [CrossRef]
4. Ladu, A.; Montisci, G.; Valente, G.; Navarrini, A.; Marongiu, P.; Pisanu, T.; Mazzarella, G. High performance
cryogenic fractal 180◦ hybrid power divider with integrated directional coupler. Radio Sci. 2017, 52, 757–766.
[CrossRef]
5. Chen, C.C.; Kuo, J.T.; Jiang, M.; Chin, A. A Fully Planar Microstrip Coupled–line Coupler with a High
Coupling Level. Microw. Opt. Technol. Lett. 2005, 46, 170–172. [CrossRef]
6. Kim, C.S.; Kim, Y.T.; Song, S.H.; Jung, W.S.; Kank, K.Y.; Park, J.S.; Ahn, D. A Design of Microstrip Directional
Coupler for High Directivity and Tight Coupling. In Proceedings of the 31st European Microwave Conference,
London, UK, 24–26 September 2001; pp. 1–4. [CrossRef]
7. Wu, Y.; Sun, W.; Leung, S.W.; Diao, Y.; Chan, K.H.; Siu, Y.M. Single–Layer Microstrip High-Directivity
Coupled–Line Coupler with Tight Coupling. IEEE Trans. Microw. Theory Tech. 2013, 61, 746–753. [CrossRef]
8. Park, M.J.; Lee, B. Compact foldable coupled line cascade couplers. IEE Proc. Microw. Antennas Propag.
2006, 153, 237–240. [CrossRef]
9. Kirschning, M.; Jansen, R.H. Accurate Wide–Range Design Equations for the Frequency–Dependent
Characteristics of Parallel Coupled Microstrip Lines. IEEE Trans. Microw. Theory Tech. 1984, 32, 83–90.
[CrossRef]
10. Mongia, R.K.; Bahl, I.J.; Bhartia, P.; Hong, J. RF and Microwave Coupled–Line Circuits, 2nd ed.; Artech House:
Boston, FL, USA, 2007; Section 6.4.1; ISBN 978-1-59693-156-5.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).