Professional Documents
Culture Documents
Electromagnetics:
Microwave Engineering
Numerical Analysis of
Transmission Lines
Outline
• Governing equations
• Numerical representation
• Finite‐difference approximations
• Matrix solution of electric potential 𝑉 𝑥, 𝑦
• Calculating the transmission line parameters
1
6/8/2020
Governing Equations
Slide 3
Maxwell’s Equations
Start with Maxwell’s equations.
B 0
D 0
H J D t
E B t
2
6/8/2020
Electrostatic Approximation
The dimensions of a transmission line are typically much smaller than the operating
wavelength so the wave nature of electromagnetics is less important to consider.
Therefore, Maxwell’s equations are essential be solved in the limit as 𝑑⁄𝑑𝑡 → 0.
B 0
D 0
H J D t
E B t
Electrostatics & Magnetostatics
Maxwell’s equations have decoupled into two sets of equations. One describes
electrostatics while the other describes magnetostatics.
Electrostatics
B 0 D 0
D 0 E 0
H J
Magnetostatics
E 0
B 0
H J
6
3
6/8/2020
Governing Equations
Maxwell’s equations:
D 0 Eq. (1)
E 0 Eq. (2)
In addition, there is the constitutive relation
D rE Eq. (3)
It is not preferred to solve vector equations if there is a way to avoid it. Electrostatic fields
are completely characterized by the electric potential 𝑉 𝑥, 𝑦 .
E V Eq. (4)
7
Differential Equation to Solve
D 0 Eq. (1)
E 0 Eq. (2)
D rE Eq. (3)
E V Eq. (4)
1. Substitute Eq. (3) into Eq. (1) to eliminate 𝐷.
rE 0 Eq. (5)
2. Substitute Eq. (4) into Eq. (5) to eliminate 𝐸.
4
6/8/2020
Differential Equation in a Homogeneous Medium
When the dielectric is homogeneous, our differential equation simplifies to
r V 0 V 0 2V 0
In Cartesian coordinates, this expands to
2V 2V 2V
0
x 2 y 2 z 2
2 z 2 0
Transmission lines are uniform in the z direction so and Laplace’s equation
reduces to
2V 2V
0
x 2 y 2
Numerical Representations
Slide 10
5
6/8/2020
Discrete Function for Electric Potential 𝑉 𝑥, 𝑦
Analytical functions contain an infinite amount of information.
In order to store 𝑉 𝑥, 𝑦 on a computer, it is stored in an array where function values are
only known at discrete points.
Start with the analytical
function 𝑉 𝑥, 𝑦 .
V x, y
x 11
Discrete Function for Electric Potential 𝑉 𝑥, 𝑦
Analytical functions contain an infinite amount of information.
In order to store 𝑉 𝑥, 𝑦 on a computer, it is stored in an array where function values are
only known at discrete points.
Divide space into a grid.
y
x y
V x, y
x 12
6
6/8/2020
Discrete Function for Electric Potential 𝑉 𝑥, 𝑦
Analytical functions contain an infinite amount of information.
In order to store 𝑉 𝑥, 𝑦 on a computer, it is stored in an array where function values are
only known at discrete points.
Store the analytical
function only at a
discrete points inside of
each cell on the grid.
j
V i, j
i 13
Discrete Function for Electric Potential 𝑉 𝑥, 𝑦
Analytical functions contain an infinite amount of information.
In order to store 𝑉 𝑥, 𝑦 on a computer, it is stored in an array where function values are
only known at discrete points.
The discrete function tends to
be thought of as being uniform
throughout the cell, but this is
technically incorrect.
j
V i, j
i 14
7
6/8/2020
Discrete Function for Electric Potential 𝑉 𝑥, 𝑦
Analytical functions contain an infinite amount of information.
In order to store 𝑉 𝑥, 𝑦 on a computer, it is stored in an array where function values are
only known at discrete points.
The discrete function
𝑉 𝑖, 𝑗 is usually
visualized this way.
V i , j V x, y
x 15
A Fundamental Tradeoff for Discrete Functions
There is always a fundamental trade‐off between accuracy and speed of simulation. The
purpose of almost all work in computational methods is to get improved accuracy with
fewer points.
Fewer points More points
Less memory More memory
Faster simulations Slower simulations
Poorer accuracy “Sweet spot” Better accuracy
Best compromise between
accuracy, memory, and speed
of simulation.
16
8
6/8/2020
Finite‐Difference
Approximations
Slide 17
Finite‐Difference Approximations (1 of 2)
Suppose the first‐order derivative of 𝑓 𝑥 is
to be numerically calculated at 𝑥 𝑥 . f3
f2
rise = f3 – f1
The first‐order derivative is slope. The slope can
be estimated as riserun using information from f1
surrounding points.
Slide 18
9
6/8/2020
Finite‐Difference Approximations (2 of 2)
The derivative at the midpoints between data
points can be estimated. f3
f 2.5
f 2 f1 f3 f 2 f2
f x1.5 f x2.5
x x
f1.5
The second‐order derivative is the slope of the
slope. f1
f x2.5 f x1.5
f x2
x
f 3 f 2 f 2 f1
x x
x
f 2 f 2 f1 x1 x2 x3
3
x 2
Slide 19
Finite‐Difference Approximation of Laplace’s
Equation
2V 2V
0
x 2 y 2
2V V i 1, j 2V i, j V i 1, j 2V V i, j 1 2V i, j V i, j 1
x 2 x y 2 y
2 2
V i 1, j 2V i, j V i 1, j V i, j 1 2V i, j V i, j 1
0
x y
2 2
20
10
6/8/2020
Rearrange Terms in the Finite‐Difference Equation
V i 1, j 2V i, j V i 1, j V i, j 1 2V i, j V i, j 1
0
x y
2 2
V i 1, j 2V i, j V i 1, j V i, j 1 2V i, j V i, j 1
0
x x x y y y
2 2 2 2 2 2
1 1 2 2 1 1
V i 1, j V i 1, j V i, j V i, j 1 V i, j 1 0
x x x y y y
2 2 2 2 2 2
21
Matrix Solution of Electric
Potential
Slide 22
11
6/8/2020
Write Large Set of Equations
The final form of the finite‐difference equation is
1 1 2 2 1 1
V i 1, j V i 1, j V i, j V i, j 1 V i, j 1 0
x x x y y y
2 2 2 2 2 2
This equation is written once for every point on the grid.
Nx 4 x 0.5
Ny 4 y 0.5
23
Consider Boundary Conditions
Nx 4 x 0.5
Ny 4 y 0.5
All of the highlighted terms will be set to zero.
Dirichlet Boundary Conditions
24
12
6/8/2020
Build Matrix Equation 𝐿 𝑣 0
16 4 0 0 4 0 0 0 0 0 0 0 0 0 0 0 V 1,1 0
4 16 4
0 0 4 0 0 0 0 0 0 0 0 0 0 V 2,1 0
0 4 16 4 0 0 4 0 0 0 0 0 0 0 0 0 V 3,1 0
0 0 4 16 0 0 0 4 0 0 0 0 0 0 0 0 V 4,1 0
4 0 0 0 16 4 0 0 4 0 0 0 0 0 0 0 V 1, 2 0
0 4 0 0 4 16 4 0 0 4 0 0 0 0 0 0 V 2, 2 0
0 0 4 0 0 4 16 4 0 0 4 0 0 0 0 0 V 3, 2 0
0 0 0 4 0 0 4 16 0 0 0 4 0 0 0 0 V 4, 2 0
0
0 0 0 4 0 0 0 16 4 0 0 4 0 0 0 V 1,3 0
0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 0 V 2, 3 0
0 0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 V 3, 3 0
0 0 0 0 0 0 0 4 0 0 4 16 0 0 0 4 V 4, 3 0
0 0 0 0 0 0 0 0 4 0 0 0 16 4 0 0 V 1, 4 0
0 0 0 0 0 0 0 0 0 4 0 0 4 16 4 0 V 2, 4 0
0 0 0 0 0 0 0 0 0 0 4 0 0 4 16 4 V 3, 4 0
0 0 0 0 0 0 0 0 0 0 0 4 0 0 4 16 V 4, 4 0
25
Is the Matrix Equation Solvable?
16 4 0 0 4 0 0 0 0 0 0 0 0 0 0 0 V 1,1 0
4 16 4
0 0 4 0 0 0 0 0 0 0 0 0 0 V 2,1 0
0 4 16 4 0 0 4 0 0 0 0 0 0 0 0 0 V 3,1 0
0 0 4 16 0 0 0 4 0 0 0 0 0 0 0 0 V 4,1 0
4 0 0 0 16 4 0 0 4 0 0 0 0 0 0 0 V 1, 2 0
0 4 0 0 4 16 4 0 0 4 0 0 0 0 0 0 V 2, 2 0
0 0 4 0 0 4 16 4 0 0 4 0 0 0 0 0 V 3, 2 0
0 0 0 4 0 0 4 16 0 0 0 4 0 0 0 0 V 4, 2 0
0 V 1,3
0 0 0 4 0 0 0 16 4 0 0 4 0 0 0 0
0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 0 V 2,3 0
0 0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 V 3,3 0
0 0 0 0 0 0 0 4 0 0 4 16 0 0 0 4 V 4,3 0
0 0 0 0 0 0 0 0 4 0 0 0 16 4 0 0 V 1, 4 0
0 0 0 0 0 0 0 0 0 4 0 0 4 16 4 0 V 2, 4 0
0 0 0 0 0 0 0 0 0 0 4 0 0 4 16 4 V 3, 4 0
0 0 0 0 0 0 0 0 0 0 0
4 0 0 4 16 V 4, 4 0
v
L 0
Lv 0 v L 0 0
1
2V 0
Trivial Solution 26
13
6/8/2020
Force Known Potentials
Force values here to 5.0
Force values here to 2.0
16 4 0 0 4 0 0 0 0 0 0 0 0 0 0 0 V 1,1 0
4 16 4
0 0 4 0 0 0 0 0 0 0 0 0 0 V 2,1 0
0 4 16 4 0 0 4 0 0 0 0 0 0 0 0 0 V 3,1 0
0 0 4 16 0 0 0 4 0 0 0 0 0 0 0 0 V 4,1 0
4 0 0 0 16 4 0 0 4 0 0 0 0 0 0 0 V 1, 2 0
V 2, 2 5 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 V 2, 2 5
V 3, 2 5 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 V 3, 2 5
0 0 0 4 0 0 4 16 0 0 0 4 0 0 0 0 V 4, 2 0
0
0 0 0 4 0 0 0 16 4 0 0 4 0 0 0 V 1,3 0
0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 0 V 2,3 0
0 0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 V 3,3 0
0 0 0 0 0 0 0 4 0 0 4 16 0 0 0 4 V 4,3 0
V 1, 4 2 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 V 1, 4 2
V 2, 4 2 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 V 2, 4 2
V 3, 4 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 V 3, 4 2
V 4, 4 2
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 V 4, 4 2
27
Solve for Electric Potential 𝑣
16 4 0 0 4 0 0 0 0 0 0 0 0 0 0 0 V 1,1 0
4 16 4
0 0 4 0 0 0 0 0 0 0 0 0 0 V 2,1 0
0 4 16 4 0 0 4 0 0 0 0 0 0 0 0 0 V 3,1 0
0 0 4 16 0 0 0 4 0 0 0 0 0 0 0 0 V 4,1 0
4 0 0 0 16 4 0 0 4 0 0 0 0 0 0 0 V 1, 2 0
0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 V 2, 2 5
0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 V 3, 2 5
0 0 0 4 0 0 4 16 0 0 0 4 0 0 0 0 V 4, 2 0
0 V 1,3
0 0 0 4 0 0 0 16 4 0 0 4 0 0 0 0
0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 0 V 2,3 0
0 0 0 0 0 0 4 0 0 4 16 4 0 0 4 0 V 3,3 0
0 0 0 0 0 0 0 4 0 0 4 16 0 0 0 4 V 4,3 0
0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 V 1, 4 2
0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 V 2, 4 2
0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 V 3, 4 2
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 V 4, 4 2
v
L b
L v b v L b
1
28
14
6/8/2020
Calculating the Transmission
Line Parameters
Slide 29
Calculating the Electric Fields
Once the scalar potential 𝑉 𝑥, 𝑦 is found, the electric field intensity 𝐸 𝑥, 𝑦 is
E x, y V x, y
It follows that the electric flux density 𝐷 𝑥, 𝑦 is
D x, y r x, y E x, y
30
15
6/8/2020
Distributed Capacitance 𝐶
In the electrostatic approximation, the transmission line is a capacitor. The total energy U stored in a
capacitor is
U 1
2 D E dA
A
CV02
U V0 is the voltage across the capacitor.
2
If the above equations are set equal and solved for C, the answer is
1
C
V02 D E dA
A
31
Distributed Inductance 𝐿
The voltage signal 𝑣 along the transmission line travels at the same velocity as the electromagnetic
field 𝑣 . This means
1 1
vV vE LC r 2 r
LC c0
Solving this for L gives
r r This means the distributed inductance 𝐿 can be calculated
L 2 directly from the distributed capacitance 𝐶.
cC
0
Dielectric materials should not alter the inductance. However if the value of C calculated on the
previous slide is used, it will. This is incorrect. The solution is to calculate the distributed capacitance
Ch with a homogeneous dielectric and then calculate the distributed inductance 𝐿 from this.
1
L
c02Ch
32
16
6/8/2020
Calculating the Transmission Line Parameters
The characteristic impedance Zc is calculated from the distributed inductance L and distributed
capacitance C through
L
Zc
C
It follows that the phase constant is
Note: This simple model did not consider loss, so 𝑅 𝐺 0 (lossless transmission line)
Slide 33
17