Berkeley

Transmission Lines in the Frequency Domain

Prof. Ali M. Niknejad

U.C. Berkeley
Copyright
c 2016 by Ali M. Niknejad

August 30, 2017

1 / 38
Why Sinusoidal Steady-State?

2 / 38
Time Harmonic Steady-State

Compared with general transient case, sinusoidal case is very

∂
easy ∂t → jω
Sinusoidal steady state has many important applications for
RF/microwave circuits
At high frequency, T-lines are like interconnect for distances
on the order of λ
Shorted or open T-lines are good resonators
T-lines are useful for impedance matching

3 / 38
Why Sinusoidal Steady-State?

Typical RF system modulates a sinusoidal carrier (either

frequency or phase)
If the modulation bandwidth is much smaller than the carrier,
the system looks like it’s excited by a pure sinusoid
Cell phones are a good example. The carrier frequency is
about 1 GHz and the voice digital modulation is about
200 kHz(GSM) or 1.25 MHz(CDMA), less than a 0.1% of the
bandwidth/carrier
Even a system like WiFi grabbing multiple channels might be
80 MHz wide, so over a 5 GHz carrier, it’s still a small
fractional bandwidth

4 / 38
Generalized Distributed Circuit Model

Z! Z! Z! Z! Z!

Y! Y! Y! Y! Y!

Z 0 : impedance per unit length (e.g. Z 0 = jωL0 + R 0 )

Y 0 : admittance per unit length (e.g. Y 0 = jωC 0 + G 0 )
A lossy T-line might have the following form (but we’ll
analyze the general case)

L! R! L! R! L! R! L! R!

G! C! G! C! G! C! G! C!

5 / 38
Telegrapher’s Eq. Again

Applying KCL and KVL to a infinitesimal section

v (z + δz) − v (z) = −Z 0 δzi(z)

i(z + δz) − i(z) = −Y 0 δzv (z)

Taking the limit as before (δz → 0)

dv
= −Zi(z)
dz
di
= −Yv (z)
dz

6 / 38
Sin. Steady-State Voltage/Current

Taking derivatives (notice z is the only variable) we arrive at

d 2v di
2
= −Z = YZv (z) = γ 2 v (z)
dz dz
d 2i dv
2
= −Y = YZi(z) = γ 2 i(z)
dz dz
Where the propagation constant γ is a complex function
p
γ = α + jβ = (R 0 + jωL0 )(G 0 + jωC 0 )

The general solution to D 2 G − γ 2 G = 0 is e ±γz

7 / 38
Lossless Lines

8 / 38
Lossless Line for Sinusoidal Steady State

V + −γz V − γz
v (z) = V + e −γz + V − e γz i(z) = e − e
Z0 Z0
The voltage and currentqare related (just as before, but now
0
easier to derive). Z0 = YZ 0 is the characteristic impedance
of the line (function of frequency with loss).
For a lossless line we discussed before, Z 0 = jωL0 and
Y 0 = jωC 0 . Propagation constant is imaginary
p √
γ = jωL0 jωC 0 = j L0 C 0 ω
The characteristic impedance is real
r
L0
Z0 =
C0
β is like the spatial frequency, also known as the wave number
You might prefer to think of it in terms of wavelength λ,
β = 2π λ
9 / 38
Back to Time-Domain

Recall that the real voltages and currents are the < and =
parts of
v (z, t) = e ±γz e jωt = e jωt±βz
Thus the voltage/current waveforms are sinusoidal in space
and time
Sinusoidal source voltage is transmitted unaltered onto T-line
(with delay)
If there is loss, then γ has a real part α, and the wave decays
or grows on the T-line

e ±γz = e ±αz e ±jβz

The first term represents amplitude response of the T-line

10 / 38
Passive T-Line/Wave Speed

For a passive line, we expect the amplitude to decay due to

loss on the line
The speed of the wave is derived as before. In order to follow
a constant point on the wavefront, you have to move with
velocity
d
(ωt ± βz = constant)
dt
q
dz
Or, v = dt = ± β = ± L01C 0
ω

11 / 38
Lossless T-Line Termination

Z0 , β ZL

z = −! z=0

Okay, lossless line means γ = jβ (α = 0), and =(Z0 ) = 0 (real

characteristic impedance independent of frequency)
The voltage/current phasors take the standard form

v (z) = V + e −γz + V − e γz

V + −γz V − γz
i(z) = e − e
Z0 Z0

12 / 38
Lossless T-Line Termination (cont)

v (0) V + +V −
At load ZL = i(0) = Z
V + −V − 0
The reflection coefficient has the same form
ZL − Z0
ρL =
ZL + Z0
Can therefore write
 
v (z) = V + e −jβz + ρL e jβz

V +  −jβz 
i(z) = e − ρL e jβz
Z0

13 / 38
Standing Waves and VSWR

14 / 38
Power on T-Line (I)

Let’s calculate the average power dissipation on the line at

point z
1
Pav (z) = < [v (z)i(z)∗ ]
2
Or using the general solution

1 |V + |2  −jβz  
Pav (z) = < e + ρL e jβz e jβz − ρ∗L e −jβz
2 Z0
The product in the < terms can be expanded into four terms

1 + ρL e 2jβz − ρ∗L e −2jβz −|ρL |2

| {z }
a−a∗

Notice that a − a∗ = 2j=(a)

15 / 38
Power on T-Line (II)

The average power dissipated at z is therefore

|V + |2

Pav = 1 − |ρL |2
2Z0
Power flow is constant (independent of z) along line (lossless)
No power flows if |ρL | = 1 (open or short)
Even though power is constant, voltage and current are not!

16 / 38
Voltage along T-Line

When the termination is matched to the line impedance

ZL = Z0 , ρL = 0 and thus the voltage along the line
|v (z)| = |V + | is constant. Otherwise

|v (z)| = |V + ||1 + ρL e 2jβz | = |V + ||1 + ρL e −2jβ` |

The voltage magnitude along the line can be written as

|v (−`)| = |V + ||1 + |ρL |e j(θ−2β`) |

The voltage is maximum when the 2β` is a equal to θ + 2kπ,

for any integer k; in other words, the reflection coefficient
phase modulo 2π

Vmax = |V + |(1 + |ρL |)

17 / 38
Voltage Standing Wave Ratio (SWR)

Similarly, minimum when θ + kπ, where k is an integer k 6= 0

Vmin = |V + |(1 − |ρL |)

The ratio of the maximum voltage to minimum voltage is an

important metric and commonly known as the voltage
standing wave ratio, VSWR (sometimes pronounced viswar),
or simply the standing wave ratio SWR

Vmax 1 + |ρL |
VSWR = =
Vmin 1 − |ρL |

It follows that for a shorted or open transmission line the

VSWR is infinite, since |ρL | = 1.

18 / 38
SWR Location

Physically the maxima occur when the reflected wave adds in

phase with the incoming wave, and minima occur when
destructive interference takes place. The distance between
maxima and minima is π in phase, or 2βδx = π, or
π λ
δx = =
2β 4
VSWR is important because it can be deduced with a relative
measurement. Absolute measurements are difficult at
microwave frequencies. By measuring VSWR, we can readily
calculate |ρL |.

19 / 38
VSWR → Impedance Measurement

By measuring the location of the voltage minima from an

unknown load, we can solve for the load reflection coefficent
phase θ
ψmin = θ − 2β`min = ±π
Note that
|v (−`min )| = |V + ||1 + |ρL |e jψmin |
Thus an unknown impedance can be characterized at
microwave frequencies by measuring VSWR and `min and
computing the load reflection coefficient. This was an
important measurement technique that has been largely
supplanted by a modern network analyzer with built-in digital
calibration and correction.

20 / 38
VSWR Example

Consider a transmission line terminated in a load impedance

ZL = 2Z0 . The reflection coefficient at the load is purely real
zL − 1 2−1 1
ρL = = =
zL + 1 2+1 3

Since 1 + |ρL | = 4/3 and 1 − |ρL | = 2/3, the VSWR is equal

to 2.
Since the load is real, the voltage minima will occur at a
distance of λ/4 from the load

21 / 38
Impedance of T-Lines (“Ohm’s Law in Freq Domain”)

22 / 38
Impedance of T-Line (I)

We have seen that the voltage and current along a

transmission line are altered by the presence of a load
termination. At an arbitrary point z, wish to calculate the
input impedadnce, or the ratio of the voltage to current
relative to the load impdance ZL

v (−`)
Zin (−`) =
i(−`)

It shall be convenient to define an analogous reflection

coefficient at an arbitrary position along the line

V − e −jβ`
ρ(−`) = = ρL e −2jβ`
V + e jβ`

23 / 38
Impedance of T-Line (II)

ρ(z) has a constant magnitude but a periodic phase. From

this we may infer that the input impedance of a transmission
line is also periodic (relation between ρ and Z is one-to-one)

1 + ρL e −2jβ`
Zin (−`) = Z0
1 − ρL e −2jβ`
The above equation is of paramount important as it expresses
the input impedance of a transmission line as a function of
position ` away from the termination.

24 / 38
Impedance of T-Line (III)

This equation can be transformed into another more useful

form by substituting the value of ρL
ZL − Z0
ρL =
ZL + Z0

ZL (1 + e −2jβ` ) + Z0 (1 − e −2jβ` )
Zin (−`) = Z0
Z0 (1 + e −2jβ` ) + ZL (1 − e −2jβ` )
Using the common complex expansions for sine and cosine, we
have
sin(x) (e jx − e −jx )/2j
tan(x) = = jx
cos(x) (e + e −jx )/2

25 / 38
Impedance of T-Line (IV)

The expression for the input impedance is now written in the

following form

ZL + jZ0 tan(β`)
Zin (−`) = Z0
Z0 + jZL tan(β`)

This is the most important equation of the lecture, known

sometimes as the “transmission line equation”

26 / 38
Shorted Line I/V

The shorted transmission line has infinite VSWR and ρL = −1.

Thus the minimum voltage Vmin = |V + |(1 − |ρL |) = 0, as
expected. At any given point along the transmission line

v (z) = V + (e −jβz − e jβz ) = −2jV + sin(βz)

whereas the current is given by

V + −jβz
i(z) = (e + e jβz )
Z0
or
2V +
i(z) = cos(βz)
Z0

27 / 38
Shorted Line Impedance (I)

The impedance at any point along the line takes on a simple

form
v (−`)
Zin (−`) = = jZ0 tan(β`)
i(−`)
This is a special case of the more general transmision line
equation with ZL = 0.
Note that the impedance is purely imaginary since a shorted
lossless transmission line cannot dissipate any power.
We have learned, though, that the line stores reactive energy
in a distributed fashion.

28 / 38
Shorted Line Impedance (II)

A plot of the input impedance as a function of z is shown

below
Zin (λ/4)
10

8
! !
! Zin (z) ! 6
! !
! Z0 !
4 Zin (λ/2)
2

-1 -0.8 -0.6 -0.4 -0.2 0

z
λ

The tangent function takes on infinite values when β`

approaches π/2 modulo 2π

29 / 38
Shorted Line Impedance (III)

Shorted transmission line can have infinite input impedance!

This is particularly surprising since the load is in effect
transformed from a short of ZL = 0 to an infinite impedance.
A plot of the voltage/current as a function of z is shown below
v(−λ/4)
v/v +
2

1. 5
v(z) i(z) Z0

1
i(−λ/4)
0. 5

0 z/λ
-1 -0.8 -0.6 -0.4 -0.2 0

30 / 38
Shorted Line Reactance

`  λ/4 → inductor
10
` < λ/4 → inductive
7. 5
reactance 5

` = λ/4 → open (acts jX(z)

2. 5
0
like resonant parallel
-2.5
LC circuit) -5

` > λ/4 but ` < λ/2 -7.5

→ capacitive reactance .25 .5 .75 1.0 1.25

z
And the process λ
repeats ...

31 / 38
Open Line I/V

The open transmission line has infinite VSWR and ρL = 1. At

any given point along the transmission line

v (z) = V + (e −jβz + e jβz ) = 2V + cos(βz)

whereas the current is given by

V + −jβz
i(z) = (e − e jβz )
Z0
or
−2jV +
i(z) = sin(βz)
Z0

32 / 38
Open Line Impedance (I)

The impedance at any point along the line takes on a simple

form
v (−`)
Zin (−`) = = −jZ0 cot(β`)
i(−`)
This is a special case of the more general transmision line
equation with ZL = ∞.
Note that the impedance is purely imaginary since an open
lossless transmission line cannot dissipate any power.
We have learned, though, that the line stores reactive energy
in a distributed fashion.

33 / 38
Open Line Impedance (II)

A plot of the input impedance as a function of z is shown

below
Zin (λ/2)
10

! !
! Zin (z) ! 6
! !
! Z0 !
4 Zin (λ/4)
2

-1 -0.8 -0.6
z -0.4 -0.2 0

The cotangent function takes on zero values when β`

approaches π/2 modulo 2π

34 / 38
Open Line Impedance (III)

Open transmission line can have zero input impedance!

This is particularly surprising since the open load is in effect
transformed from an open
A plot of the voltage/current as a function of z is shown below
i(−λ/4)
v/v +
2

1. 5 v(z) i(z)Z0

1 v(−λ/4)

0. 5

0 z/λ
-1 -0.8 -0.6 -0.4 -0.2 0

35 / 38
Open Line Reactance

`  λ/4 → capacitor
10
` < λ/4 → capacitive
7. 5
reactance 5

` = λ/4 → short (acts jX(z)

2. 5
0
like resonant series LC
-2.5
circuit) -5

` > λ/4 but ` < λ/2 -7.5

→ inductive reactance .25 .5 .75 1.0 1.25

z
And the process λ
repeats ...

36 / 38
λ/2 Transmission Line

Plug into the general T-line equaiton for any multiple of λ/2

ZL + jZ0 tan(−βλ/2)
Zin (−mλ/2) = Z0
Z0 + jZL tan(−βλ/2)
2π λm
βλm/2 = λ 2 = πm
tan mπ = 0 if m ∈ Z
Zin (−λm/2) = Z0 ZZL0 = ZL
Impedance does not change ... it’s periodic about λ/2 (not λ)

37 / 38
λ/4 Transmission Line

Plug into the general T-line equaiton for any multiple of λ/4
2π λm
βλm/4 = λ 4 = π2 m
tan m π2 = ∞ if m is an odd integer
Z02
Zin (−λm/4) = ZL
λ/4 line transforms or “inverts” the impedance of the load

38 / 38