Lecture4 New

E4318 - Microwave Circuit Design
Columbia University
Spring 2005
Yves Baeyens
Outline of Lecture 4
Short recap of lecture 3 (TL’s), power capacity

HW1
Network analysis
Impedance and admittance matrix
Scattering matrix
Calculating S-parameters, signal flow graphs
Impedance matching and tuning
Matching with lumped elements
Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 2/46

Announcements
CVN has no room available for an extended lecture, so we

will have the rescheduled class Next Tuesday (2/22)at
4:10-6:40PM also in Mudd 1024. Please let me know if this
is a problem for any of you (will be taped)
Midterm exam: Thursday 3/3, will provide formula-sheet by
next week,

Calendar
Course: Th 4:10-6:40 PM, 1024 Mudd

03/10
01/20
03/17 Spring Holidays
01/27
03/24
02/03 rescheduled
03/31
02/10
04/07
02/17
04/14
02/22 lecture 5
04/21
02/24
Final
03/03 Midterm

Recap practical transmission lines
TEM: β = ω µε = k kc = 0 vp = c εr
the transverse fields of TEM wave are same as static fields,
2 or more conductors needed, no TEM in closed conductor

voltage, current and impedance well-defined
TE or TM:
closed conductor or higher order modes TEM
propagation constant β dependent frequency & geometry
β = k 2 − k c2 k =ω µε = 2 π λ
quasi-TEM (different εr under and above line)

leads to concept of effective dielectric constant
ω
β = =ω µ 0ε 0ε e = ε e k0 vp = c εe
vp

Recap practical transmission lines (2)
Attenuation in transmission lines
α = αd + αc αc ~ Rs
Rs =
1
=
ωµ
σδ s 2σ
k 2 tan δ
αd ≅ Np/m (TE or TM waves)
2β
k tan δ
≅ Np/m (TEM waves)
2
Coaxial line:
TEM mode, use from DC-to mm-waves
EM-fields from static fields using cylindrical coordinates
First higher order mode is TE11, cutoff frequency approx.
2 ck c
kc ≅ fc =
a+b 2π εr

Recap (3): Waveguides, (micro)stripline, CPW
Rectangular waveguide has limited bandwidth

1
dominant mode is TE10: f c 10 = β = k 2 − (π a )2
2 a µε
For f<fc, β is imaginary, all field components will decay
exponentially: cut-off or evanescent modes
Higher order modes: TE10 , for normal case a>2b waveguide
BW typically factor of two
Strip-line: integrated, TEM, low dispersion and loss

Microstrip: integrated, quasi-TEM, most used, requires via
Coplanar Waveguide: integrated, quasi-TEM, more recent, no
via-holes

Power capacity of TL’s
Power in TL’s is limited by voltage breakdown, for air
occurring at breakdown electric field Ed=3x106 V/m
Calculation of capacity requires knowledge of E-field
For air-filled coax: E ρ = V0 (ρ ln b a ) this is max. for ρ=a
Vmax = Ed a ln b a
and maximum power capacity becomes:
2
Vmax πa 2 Ed2 b
Pmax = = ln
2 Z0 η0 a
As expected power capacity increases for larger diameter
cable, limit is cut-off frequency of higher order mode TE11
2 2
0.025 ⎛ cEd ⎞ ⎛ E ⎞
Pmax = ⎜⎜ ⎟⎟ = 5.8 × 1012 ⎜⎜ d ⎟⎟ @10GHz = 520kW
η0 ⎝ fmax ⎠ ⎝ fmax ⎠

Power capacity of waveguides
For air-filled rectangular waveguide: E y = Eo sin(πx a ) this is
max. Eo at x=a/2 and maximum power capacity becomes:
abEo2 abEd2
Pmax = =
4Zw 4Zw
As expected, power capacity increases for guide size, for

most waveguides b ≅ a/2, to avoid TE20 mode, a<c/fmax, with
fmax the maximum operating frequency. Maximum power
capacity of guide can be shown:
2 2
0.11 ⎛ cEd ⎞ ⎛ E ⎞
Pmax = ⎜⎜ ⎟⎟ = 2.6 × 1013 ⎜⎜ d ⎟⎟ @10GHz = 2300kW
η0 ⎝ fmax ⎠ ⎝ fmax ⎠
In practice, safety factor of two + some care for reflections

(for |Γ|=1, max. voltage can double)
Higher breakdown using inert gas or dielectric
Microwave Circuit Analysis
Circuit dimensions << wavelength

Lumped passive and active components.
Negligible phase change throughout the circuit.
Circuit theory — Kirchhoff’s laws and Ohm’s law.
Circuit dimensions ≈ wavelength

Distributed passive and active components.
Phase depends on position. Components are characterized

by their dimensions, propagation constants and
characteristic impedances.
Microwave network theory.

Impedance, voltage and current
The voltage, current and characteristic impedance of

transmission lines are defined as:
−
V = Φ+ − Φ− = ∫ E ⋅dl
+
I = ∫ H ⋅dl
C+
V
Z0 =
I
TEM-type TL have unique V, I and Z0 because:

The lines have well defined terminal pairs.
The above integrations are independent of path.

Characteristics & calculations non-TEM lines
Non TEM-type transmission lines such as rectangular
waveguide do not have unique V, I and Z0 values because:
The lines DO NOT have well defined terminal pairs.
The above integrations are path dependent.
For the dominant TE10 mode in rectangular waveguide,

voltage from the transverse fields can be written as:
− j ωµ a π x − jβ z
V = A ⋅ sin e ∫ dy
π a y
The above voltage depends on the position, x, as well as the

length of the integration contour along the y-direction.
For non-TEM: equivalent I, V & Z used (not discussed here)

An arbitrary N-port Microwave Network

Impedance Matrix
Two-terminal pair ⇒ Port.

V and I ⇒ Equivalent V and I.
Reference planes are defined to provide a phase reference for the
(equivalent) V and I phasors.
At the nth reference plane, the total voltage and current are:
V n = V n+ + V n− I n = I n+ − I n−
The impedance matrix relates these voltages and currents:
[V ] = [Z ][I ], Z ij =
Vi
Ij
I k = 0 for k ≠ j
Zii: input impedance

Zij: transfer impedance between ports i and j, i≠j

Admittance Matrix
The admittance matrix is defined as:
[I ] = [Y ][V ], Y ij =
Ii
Vj
V k = 0 for k ≠ j
Yii: input admittance

Yij: transfer admittance between ports i and j, i≠j
For reciprocal networks (no active devices, ferrites,..), the

impedance and admittance matrices are symmetric: Zij=Zji
and Yij=Yji
If the network is lossless, Zij and Yij are purely imaginary

Example: evaluation of impedance parameters
Example: Find Z-parameters two-port T-network
Solution:
Z11: input impedance port 1 when port 2 is open circuited
V
Z11 = 1 = Z A + ZC
I1 I =0
2
Transfer impedance Z12: measure open-circuit voltage at

port 1 when current I2 applied at port 2:
V V ZC
Z12 = 1 = 2 = ZC
I 2 I = 0 I 2 Z B + ZC
1
Z12=Z21 and Z22 can be found:

V
Z 22 = 2 = Z B + ZC
I 2 I =0
1

The Scattering Matrix (S-parameters)
Impedances and admittances are easy to work with; however
these parameters cannot be measured easily:
VSWR, non-TEM complicate measurement
Short and open circuits are difficult to achieve over a
broad-band of microwave frequencies.

Active devices, such as power transistors, very often are
not open- or short-circuit stable

Scattering parameters deal directly with incident,reflected
and transmitted voltage waves.
Scattering parameters can be measured directly with a vector
network analyzer (VNA).
Conversion from scattering parameters to other matrix
parameters can be easily done.

Scattering Matrix
The scattering matrix of a N-port network with the same
characteristic impedance at all ports is defined as:
⎡V1− ⎤ ⎡ S11 S12 L S1 N ⎤ ⎡V1+ ⎤
⎢ −⎥ ⎢ ⎢ ⎥
⎢V2 ⎥ = ⎢ S 21 S 22 S 2 N ⎥ ⎢V2+ ⎥
⎥
⎢ M ⎥ ⎢ M O M ⎥⎢ M ⎥
⎢ −⎥ ⎢ ⎥⎢ ⎥
⎣⎢V N ⎦⎥ ⎣ S N 1 SN 2 L S NN ⎦ ⎣⎢V N+ ⎦⎥
− + Vi −
V = S ⋅V Sij =
V j+
Vk + = 0 for k ≠ j
Vn+ and Vn- are the amplitudes of the incident and reflected
voltage waves at the nth port
Sij is found by driving port j with incident wave of voltage
Vj+and measuring the reflected wave amplitude Vi-, coming
out of port i, incident waves on all other ports set ot zero,
terminated with matched load
Example: evaluation of S-parameters
Example: Find S-parameters 3-dB attenuator network
Solution:
Sii: reflection coefficient into port i with other ports terminated
Sij: transmission coefficient from port j to i, other terminated
So, S11: reflection coefficient port 1 when port 2 is terminated in matched
load (Z0=50Ω) − (1)
V Z − Z0
S11 = 1 = Γ (1) + = in
V1+ V + = 0 V2 = 0 Z (1) + Z
in 0 Z on port 2
2 0
(1)
Z in = 8.56 + [141.8(8.56 + 50 )] (141.8 + 8.56 + 50 ) = 50Ω
For S21:
z apply incident wave at port 1, V1+
z measure outcoming wave at port 2, V2-
V2−
S21 =
V1+ V + = 0
2 3-db attenuator
Example: evaluation of S-parameters
S11=S22=0, so V1-=0 if port 2 terminated in 50 Ω (V2+=0) ⇒V1+=V1 & V2-=V2
So applying V1 and calculating V2 (2x voltage division) :
⎛ 141.8 // 58.56 ⎞⎛ 50 ⎞
V2− = V2 = V1 ⎜ ⎟⎜ ⎟ = 0.707V1
⎝ 141.8 // 58.56 + 8.56 ⎠⎝ 50 + 8.56 ⎠
⎡ 0 0.707⎤
So, S21=S12=0.707 S=⎢
⎣0.707 0 ⎥⎦
+2
If input power is V1 2 Z0 then output power is:
−2 +2 2 +2 +2
V2 2 Z 0 = S21V1 2 Z 0 = S21 V1 2 Z 0 = V1 4 Z0
so attenuator effectively attenuates with 3-dB

(half power put into 2-port is transmitted,
other half is dissipated in resistors)
3-db attenuator
Determination [S] from [Z] or [Y] (Z0n equal)
Total voltage and current at nth port (and set Z0n=1):
V n = V n+ + V n− I n = I n+ − I n− = V n+ − V n−
then: [ Z ][ I ] = [ Z ][V + ] − [ Z ][V − ] = [V ] = [V + ] + [V − ]
rewritten as: ([ Z ] + [U ])[V − ] = ([ Z ] − [U ])[V − ]

with [U] the unit or identity matrix,
[S] can be determined as: [ S ] = ([ Z ] + [U ])−1 ([ Z ] − [U ])
z −1
for one-port this becomes: S11 = 11 in agreement with
z11 + 1
reflection coefficient
For [Z] as function of [S] [ Z ] = ([U ] + [ S ])([U ] − [ S ])−1

Generalized scattering Matrix
The scattering matrix of a N-port network with characteristic

impedance Z0n at the nth port is defined as:
⎡ b1 ⎤ ⎡ S11 S12 L S1n ⎤ ⎡ a1 ⎤ an = Vn+ Z 0n
⎢b ⎥ ⎢ S S 22 S 2 n ⎥ ⎢ a2 ⎥
⎢ 2 ⎥ = ⎢ 21 ⎥⎢ ⎥ bn = Vn− Z 0n
⎢M⎥ ⎢ M O M ⎥⎢ M ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥ Z 0n = Z 0 at port n
⎣bn ⎦ ⎣ S n1 Sn2 L Snn ⎦ ⎣an ⎦
b Vi − Z 0 j
[b] = [ S ] ⋅ [a ] Sij = i =
aj +
Vk = 0 for k ≠ j
V j + Z0i
Vk + = 0 for k ≠ j
Vn+ and Vn- are the amplitudes of the incident and reflected
voltage waves at the nth port

Power Power Delivered
an and bn can be expressed in terms of Vn and In:
an = Vn+ Z 0n = Z 0n I n+ bn = Vn− Z 0n = Z 0n I n−
an + bn = Vn Z0n an − bn = Z 0n I n
an =
1
[Vn + Z0n I n ] bn =
1
[Vn − Z0n I n ]
2 Z 0n 2 Z 0n
Average power delivered to the port n is:

⎧ *⎫
1
{ }
Pn = Re Vn I n*
2
1 ⎪ +
= Re⎨ Vn + Vn−
2 ⎪
( )
⎛ Vn+ − Vn− ⎞
⎜
⎜ Z
⎟
⎟
⎪
⎬
⎩ ⎝ 0n ⎠ ⎪⎭
1 ⎪⎧ ⎛ ⎞
a −b ⎪
*⎫
= Re⎨ Z0n (an + bn )⎜⎜ n n ⎟⎟ ⎬
2 ⎪ ⎝ Z0n ⎠ ⎪⎭ incident
⎩
{ ( )}
2 2
1 2 2 a b
= Re an − bn + bnan* − bn*an = n − n
reflected
2 2 2
Reciprocal and lossless networks
For reciprocal networks (no active elements, ferrites),

[Z] and [Y] are symmetric.
Similarly, [S]-matrix of reciprocal network is symmetric:
[S]=[S]t ([S]t is transpose matrix)
For lossless networks, [Z] and [Y] are purely imaginary

The S-parameters of a lossless network form a unitary matrix:
[S]t[S]*=[U], product any column [S] with own conjugate gives
unity, product with conjugate different column gives zero
N
∑ Ski Ski = 1
*
k =1
N
∑ S ki Skj = 0 for i ≠ j
*
k =1

Application of S-parameters
⎡ 0.1∠ 0 0.8∠ 90° ⎤
Measured S-parameters 2-port: [S ] = ⎢ ⎥
⎢⎣0.8∠ 90° 0.2∠ 0 ⎥⎦
Determine if 2-port is reciprocal or lossless, calculate
return loss at port1 for short at port 2
Solution:
[S] is symmetric, so 2-port is reciprocal
Not lossless: evaluation 1st row: S11 + S12 = (0.1)2 + (0.8 )2 = 0.65 ≠ 1
2 2

Calculation reflection coefficient for shorted port 2

(V2+= -V2-) from definition S-parameters:
V1− = S11V1+ + S12V2+ = S11V1+ − S12V2−
V2− = S 21V1+ + S22V2+ = S21V1+ − S22V2−

Application of S-parameters (2)
S21
From last equation: V2− = V1+
1 + S 22
Dividing 1st equation by V1+, inserting V2-:

V1− S12 S21
Γ= = S11 −
V1+ 1 + S22
( j 0.8)( j 0.8)
= 0.1 − = 0.633
1 + 0.2
So return loss becomes:
RL = −20 log Γ = 3.97dB

Shift in Reference Planes
Phase reference planes needed

for each port of network
S-parameters transformed
when reference planes moved
from original locations
For original & new reference:
[V − ] = [ S ] ⋅ [V + ]
[V ′ − ] = [ S ′] ⋅ [V ′ + ]
From transmission line theory:

Vn′ + = Vn+ e jθ n
Vn′ − = Vn− e − jθ n
With θn=βnln electrical length of outward shift reference plane

Shift in Reference Planes
Writing S-parameter equation in matrix form:
⎡ e jθ 1 0 ⎤ ⎡ e − jθ 1 0 ⎤
⎢ ⎥ ⎢ ⎥
⎢ e jθ 2 ⎥[V ′ − ] = [ S ]⎢ e − jθ 2 ⎥[V ′ + ]
⎢ O ⎥ ⎢ O ⎥
⎢ 0 jθ n ⎥ ⎢ 0 − jθ n ⎥
⎣ e ⎦ ⎣ e ⎦
Multiplying with inverse matrix on left:
⎡ e − jθ 1 0 ⎤ ⎡ e − jθ 1 0 ⎤
⎢ ⎥ ⎢ ⎥
−
[V ′ ] = ⎢ e − jθ 2 ⎥[ S ]⎢ e − jθ 2 ⎥[V ′ + ]
⎢ O ⎥ ⎢ O ⎥
⎢ 0 − jθ n ⎥ ⎢ − jθ n ⎥
⎣ e ⎦ ⎣ 0 e ⎦
Gives expression for new S-parameter matrix
⎡ e − jθ 1 0 ⎤ ⎡ e − jθ 1 0 ⎤
⎢ ⎥ ⎢ ⎥
[ S ′] = ⎢ e − jθ 2 ⎥[ S ]⎢ e − jθ 2 ⎥
⎢ O ⎥ ⎢ O ⎥
⎢ 0 − jθ n ⎥ ⎢ − jθ n ⎥
⎣ e ⎦ ⎣ 0 e ⎦
Note: Snń=e-2jθn, phase twice shifted by electrical length of

shift in terminal plane n (wave travels twice along length)
The Vector Network Analyzer
Critical component is directional coupler (will see later)

The transmission (ABCD) Matrix
Used to calculate cascade connection of networks by

multiplying ABCD matrices of individual two-ports
I2 flowing out port 2!
Defined as:
V1 = AV2 + BI 2
I1 = CV2 + DI 2
⎡V1 ⎤ ⎡ A B ⎤ ⎡V2 ⎤
⎢ I ⎥ = ⎢C D ⎥ ⎢ I ⎥
⎣ 1⎦ ⎣ ⎦⎣ 2 ⎦
Cascade connection:
⎡V1 ⎤ ⎡ A1 B1 ⎤ ⎡V2 ⎤ ⎡V2 ⎤ ⎡ A2 B2 ⎤ ⎡V3 ⎤
⎢ I ⎥ = ⎢C D1 ⎥⎦ ⎢⎣ I 2 ⎥⎦ ⎢ I ⎥ = ⎢C D2 ⎥⎦ ⎢⎣ I 3 ⎥⎦
⎣ 1⎦ ⎣ 1 ⎣ 2⎦ ⎣ 2
⎡V1 ⎤ ⎡ A1 B1 ⎤ ⎡ A2 B2 ⎤ ⎡V3 ⎤
So: ⎢ I ⎥ = ⎢C D1 ⎥⎦ ⎢⎣C 2 D2 ⎥⎦ ⎢⎣ I 3 ⎥⎦
⎣ 1⎦ ⎣ 1
ABCD-parameters of some useful 2-ports
Library building blocks
Not commutative
Example 1st network:
V1
A = =1
V2 I2 =0
V1 V1
B = = = Z
I2 V2 = 0
V1 Z
I1
C = = 0
V2 I2 =0
I1 I1
D = = =1
I2 V2 = 0
I1

Relation Transmission and Impedance Matrix
Z-parameters (consistent with sign convention I2 ABCD):
V1 = I1 Z11 − I 2 Z12
V2 = I1 Z 21 − I 2 Z 22
Results in calculation ABCD:
V1 I1 Z11 Z11
A= = =
V2 I2 =0
I1 Z 21 Z 21
V1 I1 Z11 − I 2 Z12 I1 Z11
B= = = − Z12
I2 V I2 I2 V
2 =0 V 2 =0 2 =0
I1 Z 22 Z Z − Z12 Z 21
= Z11 − Z12 = 11 22
I1 Z 21 Z 21
I I1 1
C= 1 = =
V2 I = 0 I1 Z 21 Z 21
2
I1 Z 22
D= =
I2 V Z 21 Reciprocal: AD-BC=1
2 =0

Conversions between 2-port parameters (p.211)

Equivalent circuits for two-ports
Discontinuity: storage
electrical-magnetic energy:
results in reactances
Example: coax-to-µstrip
transition representation:
Black-box S-parameters
Equivalent circuit with #
idealized components

T and π-equivalent networks reciprocal 2-ports
For reciprocal networks, six

independent parameters needed
(real, imag. 3 matrix elements)
Leads to two possible
equivalent networks:
using impedance: T
using admittance: π
Lossless networks: elements
purely reactive

Example: equivalent network spiral inductor
Π-equivalent circuit spiral inductor
L R
C1 C2
⎛ 1 ⎞ ⎛ 1 ⎞
L = − Im ⎜⎜ ⎟⎟ ω R = − Re ⎜⎜ ⎟⎟
⎝ Y 12 ⎠ ⎝ Y 12 ⎠ 5 50
Im (Y 11 + Y 12 ) Im (Y 22 + Y 12 )
Extracted groundcapacitance [fF]

C1 = C2 =
Extracted inductance [nH] & R [ž]

4 40
ω ω C1
Extraction procedure: 3 C2 30
R
Measure S-parameters (well- 2 20
defined reference planes) 1

L
10
Calculate L,R,C,.. afo frequency 0 0

1 10 50
Take average where constant Frequency [GHz]

Signal flow graphs
Technique for analysis of microwave networks in terms of

transmitted and reflected waves
Construction of signal flow graph: primary components are
nodes and branches
Nodes: each port I of microwave network has two nodes
ai and bi. Node ai is identified with wave entering port i,

bi with wave reflected from port I
Branches: directed path between a-node and b-node
representing signal flow, each branch has associated S-

parameter or reflection coefficient
See also Agilents application note on S-parameters posted on website

Signal flowgraph of two-port
Wave a1 incident at port 1 split, part through S11 and out port 1
as reflected wave, part transmitted through S21 to node b2. At
node b2 wave goes out port 2, can be partly reflected by load
re-enter two-port at a2 , reflected back out port 2 through S22,
part transmitted out port 1 through S12
Network for one-port network and source
Signal flow graph of microwave network can be solved for

ration of combination wave amplitudes using decomposition
rules (or using Mason’s rule control system theory)

Decomposition rules signal flow graphs
Series rule
V3 = S 32V2 = S32 S 21V1
Parallel rule
V2 = SaV1 + SbV1 = ( Sa + Sb )V1
Self-loop rule
V2 = S 21V1 + S 22V2
S21
V2 = V1
1 − S 22
Splitting rule
V4 = S42V2 = S 21 S42V1

Example signal flow graph (1)

Example signal flow graph (2)
a2 = Γl b2 so S 22a2 collapsed to S 22Γl b2
Another example in book involves application of signal flow

graphs to determine error boxes of TRL-calibration VNA
Impedance Matching and Tuning
Matching network: lossless (ideally) network matching
arbitrary load impedance (non-zero real part) to a TL
Maximum power is delivered when the load and generator

are matched to the line.
Proper input impedance transformation of sensitive receiver
components (antenna, LNA, etc.) improves the S/N ratio
For power amplifier often transformation load to optimum
load line needed to increase power output active device
Impedance matching in a power distribution network (such
as antenna array feed network) will reduce amplitude and
phase errors.
Transforming Network Selection Criteria
Complexity — A simpler impedance transformation

network is usually cheaper, more reliable, and less lossy
than a more complex design.
Bandwidth — typical matching network gives only match
at single frequency, larger BW → increase in complexity
(for instance multi-section transformers).
Implementation — Short-circuited stubs in coax and
waveguide (shorting stubs easy to implement in waveguide).
Open-circuited stubs in stripline and microstrip.
Adjustability — some applications may require adjustments
(tuning stubs with micrometer in waveguides).

Lossless Matching Network
ΓL
Z0
In general, for network matching an arbitrary load

impedance to a transmission line:
To avoid unnecessary power loss, matching network is
ideally lossless.
The impedance looking in to the matching network is Z0.
Reflections are eliminated on the transmission line to the
left of the matching network.

There will be multiple reflections between the matching
network and the load.

Homework & next lecture!!
Pozar, “Microwave Engineering” (3rd Ed.!) Will put

on site!
4.10
4.16
Due date: 2/24
Next week we’ll review finish impedance matching

and review HW2

Lecture4 New

Enviado por

Dados do documento

Descrição original:

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Lecture4 New

Enviado por

Direitos autorais:

Formatos disponíveis

E4318 - Microwave Circuit Design

 Short recap of lecture 3 (TL’s), power capacity

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 2/46

 CVN has no room available for an extended lecture, so we

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 3/46

 Course: Th 4:10-6:40 PM, 1024 Mudd

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 4/46

 the transverse fields of TEM wave are same as static fields,

2 or more conductors needed, no TEM in closed conductor

 propagation constant β dependent frequency & geometry

 quasi-TEM (different εr under and above line)

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 5/46

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 6/46

 Rectangular waveguide has limited bandwidth

 Strip-line: integrated, TEM, low dispersion and loss

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 7/46

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 8/46

 As expected, power capacity increases for guide size, for

 In practice, safety factor of two + some care for reflections

 Circuit dimensions << wavelength

 Negligible phase change throughout the circuit.

 Circuit theory — Kirchhoff’s laws and Ohm’s law.

 Circuit dimensions ≈ wavelength

 Phase depends on position. Components are characterized

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 10/46

 The voltage, current and characteristic impedance of

 TEM-type TL have unique V, I and Z0 because:

 The above integrations are independent of path.

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 11/46

 The above integrations are path dependent.

 For the dominant TE10 mode in rectangular waveguide,

 The above voltage depends on the position, x, as well as the

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 12/46

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 13/46

 Two-terminal pair ⇒ Port.

 The impedance matrix relates these voltages and currents:

Zii: input impedance

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 14/46

 The admittance matrix is defined as:

Yii: input admittance

 For reciprocal networks (no active devices, ferrites,..), the

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 15/46

 Transfer impedance Z12: measure open-circuit voltage at

 Z12=Z21 and Z22 can be found:

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 16/46

 Short and open circuits are difficult to achieve over a

broad-band of microwave frequencies.

not open- or short-circuit stable

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 17/46

so attenuator effectively attenuates with 3-dB

 then: [ Z ][ I ] = [ Z ][V + ] − [ Z ][V − ] = [V ] = [V + ] + [V − ]

 rewritten as: ([ Z ] + [U ])[V − ] = ([ Z ] − [U ])[V − ]

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 21/46

 The scattering matrix of a N-port network with characteristic

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 22/46

 Average power delivered to the port n is:

 For reciprocal networks (no active elements, ferrites),

 For lossless networks, [Z] and [Y] are purely imaginary

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 24/46

 Calculation reflection coefficient for shorted port 2

Dr. Y. Baeyens E4318-Microwave Circuit Design L.4 – 25/46

 Dividing 1st equation by V1+, inserting V2-:

Short recap of lecture 3 (TL’s), power capacity

CVN has no room available for an extended lecture, so we

Course: Th 4:10-6:40 PM, 1024 Mudd

the transverse fields of TEM wave are same as static fields,

propagation constant β dependent frequency & geometry

quasi-TEM (different εr under and above line)

Rectangular waveguide has limited bandwidth

Strip-line: integrated, TEM, low dispersion and loss

As expected, power capacity increases for guide size, for

In practice, safety factor of two + some care for reflections

Circuit dimensions << wavelength

Negligible phase change throughout the circuit.

Circuit theory — Kirchhoff’s laws and Ohm’s law.

Circuit dimensions ≈ wavelength

Phase depends on position. Components are characterized

The voltage, current and characteristic impedance of

TEM-type TL have unique V, I and Z0 because:

The above integrations are independent of path.

The above integrations are path dependent.

For the dominant TE10 mode in rectangular waveguide,

The above voltage depends on the position, x, as well as the

Two-terminal pair ⇒ Port.

The impedance matrix relates these voltages and currents:

The admittance matrix is defined as:

For reciprocal networks (no active devices, ferrites,..), the

Transfer impedance Z12: measure open-circuit voltage at

Z12=Z21 and Z22 can be found:

Short and open circuits are difficult to achieve over a

then: [ Z ][ I ] = [ Z ][V + ] − [ Z ][V − ] = [V ] = [V + ] + [V − ]

rewritten as: ([ Z ] + [U ])[V − ] = ([ Z ] − [U ])[V − ]

The scattering matrix of a N-port network with characteristic

Average power delivered to the port n is:

For reciprocal networks (no active elements, ferrites),

For lossless networks, [Z] and [Y] are purely imaginary

Calculation reflection coefficient for shorted port 2

Dividing 1st equation by V1+, inserting V2-:

Phase reference planes needed

From transmission line theory:

With θn=βnln electrical length of outward shift reference plane

Note: Snń=e-2jθn, phase twice shifted by electrical length of

Used to calculate cascade connection of networks by

Equivalent circuit with #

For reciprocal networks, six

Calculate L,R,C,.. afo frequency 0 0

Technique for analysis of microwave networks in terms of

Signal flow graph of microwave network can be solved for

Maximum power is delivered when the load and generator

Complexity — A simpler impedance transformation

In general, for network matching an arbitrary load

Reflections are eliminated on the transmission line to the

Pozar, “Microwave Engineering” (3rd Ed.!) Will put

Due date: 2/24

Next week we’ll review finish impedance matching