Microwave and RF Design
Microwave and RF Design
Microwave and RF Design
V (z ) I(z ) IL
+
Z 0 ,β VL
_ ZL
z
0
and referring to Equation (4.80) it is seen that the second square root is just
1, so r
L
Z0 = , (4.84)
C
which is frequency independent. So the important characteristics describing
signal propagation are independent of frequency and so the transmission
line is dispersionless.
The reflected wave has a similar relationship (but watch the sign change):
VL V (z = 0)
= = ZL . (4.87)
IL I(z = 0)
TRANSMISSION LINES 195
and the total current, I(z), is related to the traveling current waves by
V (0) V + + V0−
= ZL = Z0 0+ .
I(0) V0 − V0−
This can be rearranged as the ratio of the reflected voltage to the incident
voltage:
V0− ZL − Z0
= .
V0+ ZL + Z0
This ratio is defined as the voltage reflection coefficient
V0− ZL − Z0
ΓV = + = . (4.90)
V0 ZL + Z0
This relationship can be rewritten so that the input load impedance can be
obtained from the reflection coefficient:
1 + ΓV
ZL = Z0 . (4.91)
1 − ΓV
Similarly the current reflection coefficient can be written as
I0− −ZL + Z0
ΓI = + = = −ΓV . (4.92)
I0 ZL + Z0
The voltage reflection coefficient is used most of the time, so the reflection
coefficient, Γ, on its own refers to the voltage reflection coefficient, ΓV = Γ.
The total voltage and current waves on the line can be written as
Solution:
(a) C = 10 · 10−12 F ; R = 60 Ω ; f = 5.109 ; ω = 2πf ; Z0 = 50 Ω
2 2
2
− 1 V0− |Γ| V0+
P = = ,
2 Z0 2 Z0
TRANSMISSION LINES 197
where Equations (4.93) and (4.94) are used. Considering the conservation of
power, the power delivered to the load is
1
2
PL = ℜ{VL IL∗ } = P + − P − = P + 1 − |Γ| . (4.96)
2
Thus the total power flow on the line is equal to the incident power minus
the reflected power. The other noteworthy cases are when there is an open
circuit, a short circuit, or a purely reactive load at the end of a transmission
line. These have |Γ| = 1. Thus all power is reflected back to the source and
PL = 0.
The power that is absorbed by the load appears as a loss as far as the
incident and reflected waves are concerned. To describe this, the concept of
return loss (RL) is introduced and defined as
Solution:
The equivalent circuit will have the form
Z TH
E Z L = 100 Ω
TH
where ETH is the Thevenin equivalent generator and ZTH is the Thevenin equivalent
generator impedance. The amplitude of the forward-traveling voltage wave is obtained by
calculating the power in the forward-traveling wave:
1 + 2
P+ = (V ) /Z0
2
= (V + )2 /150
= 1 µW = 10−6 .
198 MICROWAVE AND RF DESIGN: A SYSTEMS APPROACH
REFERENCE REFERENCE
PLANE 1 PLANE 2
Z1
Z 01 Z 02
PI1 PI2
PT1 PT2
PR1 PR2
√
So V + = 150.10−6 = 12.25 mV. Note that V + is not ETH . To calculate ETH , consider the
following circuit which will result in maximum power transfer:
75 Ω
+
E V 75 Ω
TH = 12.25 mV
75 Ω
24.5 mV 100 Ω
The transmission line shown in Figure 4-16 consists of a source with Thevenin impedance
Z1 = 40 Ω and source E = 5 V (peak), connected to a quarter-wavelength long line of
characteristic impedance Z01 = 40 Ω, which in turn is connected to an infinitely long
line of characteristic impedance Z02 = 100 Ω. The transmission lines are lossless. Two
reference planes are shown in Figure 4-16. At reference plane 1 the incident power is PI1 (the
maximum available power from the source), the reflected power is PR1 , and the transmitted
power is PT 1 . PI2 (the maximum available power from Z01 ), PR2 , and PT 2 are similar
quantities at reference plane 2. PI1 , PR1 , PT 1 , PI2 , PR2 , and PT 2 are steady-state quantities.
TRANSMISSION LINES 199
Solution:
(a) PI1 is the available power from the generator. Since the Thevenin impedance of the
generator is 40 Ω, PI1 is the power that would be delivered to a matched load (the
maximum available power). An equivalent problem is
Z 1 = 40 Ω
E V Z L = 40 Ω
PI1 = Power in ZL
1 1 1 1
= (V)2 = (2.5)2
2 ZL 2 40
= 0.07813 W = 78.13 mW . (4.98)
1
Note that the 2
is used because peak voltage is used.
(b) Now the problem becomes interesting and there are many ways to solve it. One of
the key observations is that the first transmission line has the same characteristic
impedance as the Thevenin equivalent impedance of the generator, Z01 = Z1 , and
so can be ignored where appropriate. This observation will be used in this example.
One way to proceed is to directly calculate PT 2 , and a second approach is to calculate
the incident and reflected powers at reference plane 2 and then to determine PT 2 .
(i) First approach: Looking to the left from Port 2 the circuit can be modeled as an
equivalent circuit having a Thevenin equivalent resistance of 40 Ω and a Thevenin
equivalent voltage that produces the right amount of power. So consider the
following circuit:
40 Ω
+
E2 V_ ZL = 100 Ω
Here, E2 is the voltage required to deliver the maximum possible power to the load
and we know that this power is PI1 (= 78 mW), as calculated above. The load is 100 Ω
as the second transmission line is infinitely long. So based on the previous problem we
know that E2 is just E or 5 V. A reasonable question would be why is it not 2.5 V, as
this would be the voltage across ZL = 40 Ω in Part (a). However, 2.5 V is the voltage
of the forward-traveling voltage wave on the first transmission line with characteristic
impedance Z01 = 40 Ω. So the voltage across the load is
100 100
V =E =E = 3.57 Vpeak . (4.99)
40 + 100 140
200 MICROWAVE AND RF DESIGN: A SYSTEMS APPROACH
So
1 1
PT 2 = PL = (V )2
2 ZL
1 1
= (3.57)2
2 100
= 0.0638 W = 63.8 mW. (4.100)
ZL − Z01 100 − 40
Γ2 = = = 0.429
ZL + Z01 100 + 40
PR2 = Γ22 PI2 = 0.4292 × 78 mW = 14.36 mW .
This example is similar to Example 4.9. Again, the transmission line network of Figure 4-16
is considered, but now the characteristic impedance of the first transmission line is not the
same as the generator impedance so the simplification used in the previous example can
no longer be used. Now the generator has a Thevenin impedance Z1 = 40 Ω and source
E = 5 Vpeak , connected to a one-quarter wavelength long line of characteristic impedance
Z01 = 30 Ω which in turn is connected to an infinitely long line of characteristic impedance
Z02 = 100 Ω. The transmission lines are lossless. Two reference planes are shown in Figure
4-16. At reference plane 1 the incident power is PI1 (the maximum available power from
the source), the reflected power is PR1 and the transmitted power is PT 1 . PI2 (the maximum
available power from Z01 ), PR2 , and PT 2 are similar quantities at reference plane 2. PI1 , PR1 ,
PT 1 , PI2 , PR2 , and PT 2 are steady-state quantities.
Solution:
One of the first things to note is that the infinitely long 100 Ω transmission line is
indistinguishable from a 100 Ω resistor, so the reduced form of the problem is as shown
below:
TRANSMISSION LINES 201
REFERENCE REFERENCE
PLANE 1 PLANE 2
40 Ω
30 Ω
E 100 Ω
PI1 PI2
PT1 PT2
PR1 PR2
(b) The problem here is finding PT 2 . Recall that the powers here are steady state quantities
so that we are not considering multiple reflections of, say, a pulse. Since the system
is lossless, the power delivered by the generator must be the power delivered to the
infinitely long transmission line Z02 (i.e., PT 2 ). The telegraphist’s equation can be used
to calculate the input impedance, Zin , of the two transmission line system; that is, the
input impedance of Z01 from the generator end. However, a simpler way to find this
impedance is to realize that the Z01 line is a one-quarter wave transformer so that
√ √
Z01 = 30 = Zin Z02 = 100Zin , (4.102)
and so
Zin = 9 Ω . (4.103)
The equivalent circuit is thus
40 Ω
+
E V_ ZL = 9 Ω
(c) The power transmitted into the system at reference plane 1, PT 1 , is the same as the
power transmitted to the 100 Ω load, as the first transmission line is lossless; that is,
PT 1 = PT 2 = 46.86 mW . (4.106)
202 MICROWAVE AND RF DESIGN: A SYSTEMS APPROACH
Also
PR1 = PI1 − PT 2 = (78.13 − 46.86) mW = 31.27 mW. (4.107)
The two remaining quantities to determine are PI2 and PR2 . There can be a number of
interpretations of what these should be, but one thing that is certain is that
One interpretation that will be followed here is based on the equivalent circuit at
reference plane 2. Let Zout be the impedance looking to the left from reference plane 2.
Again, using the property of a one-quarter wavelength long transmission line,
√
Z01 = 30 = Zout × 40 (4.109)
and so
2
Zout = Z01 /40 = 302 /40 = 22.5 Ω . (4.110)
The equivalent circuit at reference plane 2 is then
Z 3 = 22.5 Ω
+
E3 Vx
_ Z L = 100 Ω
Now determine Vx that results in a power PT 2 being delivered to the 100 Ω load, so
1 2
PT 2 = V /100 = 0.04686 W, (4.111)
2 x
and
Vx = 3.06 V . (4.112)
From the circuit above,
100
Vx = E3 , (4.113)
100 + 22.5
and so
E3 = 3.75 V. (4.114)
The available power from this source is obtained by considering
Z 3 = 22.5 Ω
3.75 V 22.5 Ω
.
The available power at reference plane 2 is
V( z)
Vmax = 1.5
Vmin = 0.5
z
−3λ/4 −λ/2 −λ/4
l 0
reason, such a line is said to be “flat.” If the load is mismatched, the presence
of reflected waves leads to standing waves where the magnitude of the
voltage on the line is not a constant (see Figure 4-17). Thus from Equations
(4.93) and (4.94):
Similarly the minimum value of the voltage magnitude occurs when the
phase term e(Θ−2βl) = −1, and is given by
As |Γ| increases, the ratio of Vmax to Vmin increases and the mismatch defined
by the Voltage Standing Wave Ratio (VSWR) is given by
Hence the magnitude of the current has its maxima where the voltage is
minimum and has its magnitude minima where the voltage magnitude is
maximum.
TRANSMISSION LINES 205
Solution:
(a) From Example 4.7 on Page 196, ΓL = 0.9936 187.3◦ and so
1 + |ΓL | 1 + 0.993
VSWR = = = 285 .
1 − |ΓL | 1 − 0.993
A load has an impedance ZL = 45 + 75 and the system reference impedance, Z0 , is 100 Ω.
Solution:
R TH =100 Ω
E ZL
(f) It is tempting to think that the power dissipated in RTH is just PR . However, this is not
correct. Instead, the current in RTH must be determined and then the power dissipated
in RTH found. Let the current through RTH be I, and this is composed on forward- and
backward-traveling components:
I = I + + I − = (1 + ΓI )I + ,
Γin ΓL
ZG
Z in Z ,β ZL Z in Z ,β ZL
0 0
(a) (b)
EG VC EL V 50 Ω
A shorted coaxial line is used as a resonator. The first resonance is determined to be a parallel
(or shunt) resonance and is at 1 GHz.
(a)
R C L
TRANSMISSION LINES 209
λg /4
This example builds on Example 4.6. Communication filters are often constructed using
several shorted transmission line resonators that are coupled to each other. Consider a
coaxial line that is short-circuited at one end. The dielectric filling the coaxial line has
a relative permittivity of 20 and the resonator is to be designed to resonate at a center
frequency, f0 , of 1850 MHz.
(b)
Y L C
Y = 0 at resonance
1 ω −1
Y = YL + YC = + ωC = + ωC.
ωL L
ZL + Z0 tan βℓ
Zin = Z0 . (4.132)
Z0 + ZL tan βℓ
√
In this problem ZL = 0 Ω, ℓ = (0.0362 m)/4 = 9.05 mm, β = β0 εr , where β0 is the phase
constant of free space. From Equations (4.56) and (4.57),
√
β = 20.958 × f0 |( GHz) × 20 = 173.4 rad/m
(g) The equivalent circuit resonates at f0 = 1.85 GHz when 1/(ω0 L) = −ω0 C; ω0 = 2πf0 .
That is,
LC = ω0−2 = 0.7401 × 10−20 s−2 . (4.134)
Equating Equations (4.131) and (4.133),
1
+C = 1.101 × 10−11 F
ω02 L
1
C = 1.101 × 10−11 − F. (4.135)
ω02 L
SHORT−CIRCUITED STUB Z 01
Z 01
Z1 Z1 Z 01 Z 01
(a) (b) (c) (d) (e)
OPEN−CIRCUITED STUB Z 01
Z 01
Z1 Z1 Z 01 Z 01
(f) (g) (h) (i) (j)
Figure 4-19 Transmission line stubs: (a)–(e) short-circuited stubs; and (f)–(j) open-circuited stubs.
Y L C
, with L = 1.345 nH and C = 5.503 pF.
This equivalent circuit is valid for a narrow range (say 5%) of frequencies around 1.85 GHz.
π f
Z1 = Z01 tan . (4.139)
2 fr
212 MICROWAVE AND RF DESIGN: A SYSTEMS APPROACH
BOND
WIRE
Figure 4-20 Open-circuited stub with variable length realized using wire bonding
from the fixed stub to one of the bond pads.
REFERENCE REFERENCE
PLANE A PLANE B
o
45
Z 1 = 30 Ω
Z 01 = 60 Ω Z 02 = 40 Ω
Γin Z in
A B
A Τ21 B Τ21
Γ11 Γ11
A B
Τ12 A Τ12 B
Γ22 Γ22
Figure 4-21 Transmission line network with a finite length line of characteristic
impedance Z01 and an infinitely long transmission line of characteristic impedance
2 2
Z02 . Power conservation at an interface requires that A Γ11 + A T21 = 1, etc.
In the transmission line system in Figure 4-21 there are two planes at which reflection
and transmission (or scattering) occur. The electrical length of the Z01 line is 45◦ and
the Z02 line is infinitely long. The aim here is to find the input impedance, Zin . This
will be arrived at in two ways. The first technique uses a bounce diagram approach and
emphasizes reflection and transmission at the interface planes. The second approach uses
the telegraphist’s equation. It is more subtle to use the bounce diagram technique, but it
yields the same answer as using the telegraphist’s equation.
A Z01 − Z1 A
Γ11 = = 0.333; Γ22 = −A Γ11 = −0.333 (4.143)
Z01 + Z1
q q
A A
T21 = 1 − A Γ211 = 0.943; T12 = 1 − A Γ222 = 0.943 (4.144)
While A T12 is greater than one, this does not indicate power gain because of the
differences in impedance levels.
(b) What are the scattering parameters Γ2 and T2 at reference plane B? Now the reference
is Z0 , and so
B Z02 − Z0
Γ11 = . (4.145)
Z02 + Z0
The question now is what should the system reference impedance be—should it
be Z1 , Z01 , or even Z02 ? The problem could be solved using any of these, but the
simplest procedure is to use the same reference impedance throughout, and since the
eventual aim is to calculate the overall input reflection coefficient, the appropriate
choice is Z0 = Z1 . Note, however, that the actual voltage levels on the lines are not
being calculated (which would need to be referenced to the characteristic impedance
of the lines they are on), but instead to a traveling wave referenced to a universal
system impedance. If it were required that the actual voltages be calculated, then a
system reference impedance change would be necessary. So the scattering parameters
at reference plane B referenced to the impedance Z1 are
B Z02 − Z1 B
Γ11 = = 0.143; Γ22 = −B Γ11 = −0.143 (4.146)
Z02 + Z1
q q
B B
T21 = 1 − A Γ211 = 0.990; T12 = 1 − B Γ222 = 0.990 . (4.147)
The second transmission line is infinitely long and so no signal from the line will be
incident at reference plane B from the right.
(c) What is the transmission coefficient of the Z01 transmission line?
Using a reference impedance of Z01 , the transmission coefficient is one and rotates by
45◦ or 0.785 radians (it is the ratio of the forward traveling wave at the end of the line
and its value at the start of the line):
A
A T12 A T21 T01
2 B
Γ11
Γin = Γ11 + 2B A
1 − T01 Γ11 Γ22
0.943 × 0.943 × (0.707 − 0.707)2 × (−0.2)
= 0.333 +
1 − (0.707 − 0.707)2 × (−0.2) × (−0.333)
= 0.345 − 0.177 . (4.149)
(f) Use the lossless telegraphist’s equation, Equation (4.128), to find Zin .
The infinitely long line presents an impedance Z02 to the 60 Ω transmission
line. So the input impedance looking into the 60 Ω line at reference plane A is,
using the lossless telegraphist’s equation,
Z02 + Z01 tan βl
Zin = Z01 ,
Z01 + Z02 tan βl
which is the same result obtained using the bounce diagram method.
In the above, two techniques were used to evaluate the input impedance of a
transmission line system. The bounce diagram technique required a subtle choice
of several different reference impedances which were required to keep track of
real power flow. It is quite difficult to make these choices. It can be viewed
that the bounce diagram considers instantaneous reflections of pulses. These are
summed to yield the steady-state input reflection coefficient, and thus the input
impedance. The telegraphist’s equation directly captures the steady-state response.
So while the bounce diagram technique aids in physical understanding, using the
telegraphist’s equation is a much less error prone approach to solving transmission
line problems.