Beruflich Dokumente
Kultur Dokumente
S-PARAMETERS
Note :
A port is a terminal pair of lines.
I have seen S-parameters described as S11, S21, etc. Can you explain?
I have seen S-parameters described as S11, S21, etc. Can you explain?
I have seen S-parameters described as S11, S21, etc. Can you explain?
I have seen S-parameters described as S11, S21, etc. Can you explain?
I have seen S-parameters described as S11, S21, etc. Can you explain?
I have seen S-parameters described as S11, S21, etc. Can you explain?
A linear network can be characterised by a set of simultaneous equations
describing the exiting waves from each port in terms of incident waves.
S11 = b1 / a1
S12 = b1 / a2
S21 = b2 / a1
S22 = b2 / a2
Note again how the subscript follows the parameters in the ratio (S11=b1/a1, etc...)
How much gain (or loss) you get is usually more important than how much
the signal has been phase shifted.
In the Si9000e
if the load impedance is changed, the S-parameters
change.
S-parameters are
quoted with source and load
if the source impedance is changed, the S-parameters change.
impedances of 50 Ohms
A little math
This is the matrix algebraic representation
of 2 port S-parameters:
A little math
This is the matrix algebraic representation
of 2 port S-parameters:
A little math
This is the matrix algebraic representation
of 2 port S-parameters:
A little math
Parameters along the leading diagonal,
S11 & S22, of the S-matrix are referred to as
reflection coefficients because they refer to
the reflection occurring at one port only.
A little math
Parameters along the leading diagonal,
S11 & S22, of the S-matrix are referred to as
reflection coefficients because they refer to
the reflection occurring at one port only.
Larger networks:
A Network may have any number of ports.
Larger networks:
A Network may have any number of ports.
Larger networks:
A Network may have any number of ports.
Larger networks:
A Network may have any number of ports.
Larger networks:
A Network may have any number of ports.
Sum up
Further reading:
Agilent papers
http://www.sss-mag.com/pdf/an-95-1.pdf
http://www.sss-mag.com/pdf/AN154.pdf
National Instruments paper
http://zone.ni.com/devzone/nidzgloss.nsf/webmain/D2C4FA88321195FE8625686B00542
EDB?OpenDocument
Other links:
http://www.sss-mag.com
http://www.microwaves101.com/index.cfm
http://www.reed-electronics.com/tmworld/article/CA187307.html
http://en.wikipedia.org/wiki/S-parameters
Online lecture OLL-140 Intro to S-parameters - Eric Bogatin
Online lecture OLL-141 S11 & Smith charts - Eric Bogatin
www.bethesignal.com
Description
TRANSMISSION LINES
V (t ) = Vo cos ( t + )
V (t ) = Re Vo e j( t + ) = Re Vo e j e j t
j=
Vo e j
is a complex phasor
36
Phasors
Im
Vo
Vo e j = Vo cos ( ) + jVo sin ( )
Vo sin ( )
Vo cos ( )
Re
q = C l xV
where Cl is the capacitance per unit length
Azimuthal magnetic flux:
= L l xI
where Ll is the inductance per unit length
38
C l x
v + v
di
v (v + v ) = L l x
dt
dv
i + i = i C l x
dt
39
B
E =
t
D
H =
t
v
i
= Ll
t
x
i
v
= C l
x
t
Solutions are traveling waves
v (t , x ) = v
v+
i (t , x ) =
Zo
x
x
t
+ v t +
vel
vel
x v
x
t +
vel Z o
vel
40
vel =
1
L lC l
1
=
L lC l
Zo =
Ll
Cl
41
RL
Pulse travels down the transmission line as a forward going wave only
(v+). However, when the pulse reaches the load resistor:
+
v
v +v
= RL =
i
v+ v
Zo Zo
Reflection Coefficient
The reverse wave can be thought of as the incident wave reflected
from the load
v R L Zo
Reflection coefficient
=
=
v + R L + Zo
Three special cases:
RL = (open)
= +1
RL = 0 (short)
= -1
RL = Z o
=0
43
Sinusoidal Waves
Experiment: Send a SINGLE frequency () sine wave into a
transmission line and measure how the line responds
v + = V + cos ( t x ) = Re V + e j x e j t
= vel
2 f 2
=
=
vel
phase velocity
wave number
di
dt
dv
i=C
dt
v=L
V = j LI
Z ind = j L
I = j CV
Z cap =
1
j C
44
Standing Waves
x
Zo
ZL
x=0
d
At x=0
ZL Zo
V = V =
ZL + Zo
+
V = V + e j x + V + e + j x
V = V + (1 )e j x + 2 V + cos ( x )
traveling wave
standing wave
45
1.5
V + (1 + ) (1 + )
Vmax
=
=
+
Vmin
V (1 ) (1 )
0.5
= VSWR
0.5
=
1.5
0.1
0.2
1
2
0.3
0.4
0.5
Position
0.6
0.7
0.8
0.9
46
1
2
Incident wave
Reflected wave
Standing wave
47
Zo
ZL
x=0
d
towards load
towards generator
L =
ZL Zo
ZL + Zo
V = V + e j x + L V + e + j x
Wave has to travel
down and back
Vforward
V + e + j ( d )
G =
= L
= L e j2 d
Vreverse gen
V + e j ( d )
48
There is a one-to-one
correspondence between G and
ZL
= 2 d
Re { }
ZG Zo
G =
ZG + Zo
1 + G
ZG = Zo
1 G
ZG = Zo
1 + L e j2 d
1 L e j2 d
49
so L = +1
jZ o
ZG =
tan ( d )
For d<<1
ZG
jZ o
1
=
d
j C l d
looks capacitive
For d = /2 or d=/4
ZG = 0
An open circuit at the load looks like a short circuit at the generator
if the generator is a quarter wavelength away from the load
50
so L = -1
Z G = jZ o tan ( d )
For d<<1
Z G jZ o d = j L l d
looks inductive
For d = /2 or d=/4
ZG
A short circuit at the load looks like an open circuit at the generator
if the generator is a quarter wavelength away from the load
51
Zo
Ps
ZL
d
x = d
x = 0
VG = V ( d ) = V + e + j d + L V + e j d
V + + j d
V + j d
I G = I( d ) =
e
L
e
Zo
Zo
The rate of energy flowing through the plane at x=-d
1
P = Re VG I G *
2
+2
forward power
P=
+2
1V
1
2 V
L
2 Zo
2
Zo
reflected power
52
L = 0
which implies:
ZL = Zo
When ZL = Zo, the load is matched to the transmission line
53
Load Matching
What if the load cannot be made equal to Zo for some other reasons?
Then, we need to build a matching network so that the source
effectively sees a match load.
Ps
Z0
ZL
=0
Typically we only want to use lossless devices such as capacitors,
inductors, transmission lines, in our matching network so that we do
not dissipate any power in the network and deliver all the available
power to the load.
54
Normalized Impedance
It will be easier if we normalize the load impedance to the
characteristic impedance of the transmission line attached to the
load.
Z
z=
Zo
z=
= r + jx
1+
1
= u + jv
r=
1 u 2 v2
2
(1 u )
+v
x=
2v
(1 u )2 + v 2
55
dB and dBm
A dB is defined as a POWER ratio. For example:
Prev
dB = 10 log
Pfor
2
= 10 log
= 20 log ( )
A dBm is defined as log unit of power referenced to 1mW:
P
PdBm = 10 log
1mW
56
Description
Z AND S PARAMETERS
I2
I1
V1
V2
V1 = Z11 I1 + Z12 I 2
V2 = Z 21 I1 + Z 22 I 2
[V ] = [Z ][I ]
58
Vi = Vi+ + Vi
Z o I i = Vi+ Vi
We can characterize the circuit with forward and reverse waves:
[V ]= [S][V ]
59
Z and S Parameters
Similar to the reflection coefficient, there is a one-to-one
correspondence between the impedance matrix and the scattering
matrix:
ai =
Vi+
2Zoi
bi =
Vi
2Zoi
61
b1 = s11a 1 + s12 a 2
b 2 = s 21a 1 + s 22 a 2
[b ] = [s ][a ]
Where:
s i, j =
Zo j
Zoi
Si, j
62
Normalized S Parameters
The s parameters can be drawn pictorially
a1
s21
s11
b1
b2
s22
s12
a2
63
Examples of S parameters
Zo
0
[s ] = j
e
e j
Transmission Line
1 0
[s ] =
Short
G
Amplifier
0
[s ] =
G
0
0
64
Examples of S parameters
1
0
[s ] = 1
0
0
1
1
0
0
0
[s ] =
1
0
0
Circulator
1
Zo
Isolator
2
65
Lorentz Reciprocity
If the device is made out of linear isotropic materials (resistors,
capacitors, inductors, metal, etc..) then:
[s ]T = [s ]
or
s j, i = s i , j
for
i j
non-reciprocal devices:
transmission line
short
directional coupler
amplifier
isolator
circulator
66
Lossless Devices
The s matrix of a lossless device is unitary:
[s ] [s] = [1]
*T
1 = s i, j
2
for all j
1 = s i, j
2
for all i
j
Lossless devices:
Non-lossless devices:
transmission line
short
circulator
amplifier
isolator
67
Network Analyzers
Network analyzers measure S
parameters as a function of
frequency
At a single frequency, network
analyzers send out forward waves a1
and a2 and measure the phase and
amplitude of the reflected waves b1
and b2 with respect to the forward
waves.
b
s11 = 1
a1 a = 0
2
b
s12 = 1
a 2 a =0
1
b
s 21 = 2
a1 a = 0
2
b
s 22 = 2
a 2 a =0
1
a1
b1
a2
b2
68
y21
y11
s21
y22
y12
Connector X
yx21
s11
x21
s22
x11
s12
x22
x12
yx12
We want to know the S parameters at
these reference planes
We measure the S parameters at these
reference planes
69
0
[s ] =
1
Short
1 0
[s ] =
Delay*
0
[s ] = j
e
*~90degrees
1
0
e j
0
71
Phase Delay
A pure sine wave can be written as:
V = Vo e j( t z )
The phase shift due to a length of cable is:
= d
=
d
v ph
= ph
The phase delay of a device is defined as:
arg (S 21 )
ph =
72
Phase Delay
For a non-dispersive cable, the phase delay is the
same for all frequencies.
In general, the phase delay will be a function of
frequency.
It is possible for the phase velocity to take on any
value - even greater than the velocity of light
Waveguides
Waves hitting the shore at an angle
73
Group Delay
A pure sine wave has no information content
There is nothing changing in a pure sine wave
Information is equivalent to something changing
m
V = Vo cos ( t ) + Vo [cos (( + )t ) + cos (( )t )]
2
V = Vo (1 + m cos ( t )) cos ( t )
74
Group Delay
The waves emanating from the source will look like
V = Vo cos ( t z )
m
+ Vo cos (( + )t ( + )z )
2
m
+ Vo cos (( )t ( )z )
2
Which can be re-written as:
V = Vo (1 + m cos ( t z )) cos ( t z )
75
Group Delay
The information travels at a velocity
v gr =
d
gr =
v gr
(arg (S 21 ))
=
76
arg (S 21 )
ph =
Group Delay:
(arg (S 21 ))
gr =
77
Description
SMITH CHART
Smith Chart
Impedances, voltages, currents, etc. all repeat
every half wavelength
The magnitude of the reflection coefficient, the
standing wave ratio (SWR) do not change, so they
characterize the voltage & current patterns on the
line
If the load impedance is normalized by the
characteristic impedance of the line, the voltages,
currents, impedances, etc. all still have the same
properties, but the results can be generalized to
any line with the same normalized impedances
79
Smith Chart
The Smith Chart is a clever tool for analyzing
transmission lines
The outside of the chart shows location on the line
in wavelengths
The combination of intersecting circles inside the
chart allow us to locate the normalized impedance
and then to find the impedance anywhere on the
line
80
Smith Chart
Thus, the first step in analyzing a transmission line is to
locate the normalized load impedance on the chart
Next, a circle is drawn that represents the reflection
coefficient or SWR. The center of the circle is the center
of the chart. The circle passes through the normalized
load impedance
Any point on the line is found on this circle. Rotate
clockwise to move toward the generator (away from the
load)
The distance moved on the line is indicated on the
outside of the chart in wavelengths
81
Smith Charts
The impedance as a function of reflection coefficient can be rewritten in the form:
r=
1 u v
(1 u )2 + v 2
x=
2v
(1 u )2 + v 2
r
1
2
u
+v =
1+ r
(1 + r )2
2
1
1
2
(u 1) + v = 2
x
x
These are equations for
circles on the (u,v) plane
82
Im {}
0.5
r=0
r=1/3
r=1
0.5
r=2.5
0.5
Re {}
0.5
83
x=1/3
x=1
x=2.5
0.5
0.5
x=-1/3
0.5
0.5
x=-1
Re {}
x=-2.5
84
Smith Chart
85
Z o = 50
What is ZL?
Z L = 50 (1 .35 + j1 .35 )
= 67 .5 + j67 .5
86
Z o = 50
What is L?
15 j25
50
= 0 .3 j0 .5
zL =
L = 0 .618 124
87
Z L = 50 + j50
Z o = 50
= 6 .78 nS
Z in = ?
zL =
L = 0 .445 64
in = L e
j2 d
= L e
j2
2 = 244
2 = 244
in = 0 .445 180
Z L = 50 (0 .38 + j0 .0 ) = 19
88
Admittance
A matching network is going to be a combination of elements
connected in series AND parallel.
Z1
Impedance is well suited when working with
series configurations. For example:
V = ZI
Z L = Z1 + Z 2
ZL =
1
Y1 =
Z1
Z1
Z2
Y1
Y2
Z1Z 2
Z1 + Z 2
I = YV
Z2
YL = Y1 + Y2
89
Normalized Admittance
Y
y=
= YZ o = g + jb
Yo
1
y=
1+
2
g=
1 u v
(1 + u )2 + v 2
b=
2v
(1 + u )2 + v 2
g
1
2
u +
+ v =
1+ g
(1 + g )2
1
1
2
(u + 1) + v + = 2
b
b
These are equations for
circles on the (u,v) plane
90
Im {}
Im { }
b=-1/3
b=-1
b=-2.5
0.5
g=2.5
1
g=1/3
g=1
0.5
0.5
g=0
0.5
Re {}1
0.5
Re {}
1
b=1/3
b=2.5
0.5
0.5
0.5
b=1
1
91
92
y = 1 + j1
What is ?
Procedure:
Plot y
0 .445 64
Read
Flip 180
degrees
= 0 .445 116
93
= 0 .5 + 45 Z o = 50
What is Y?
Procedure:
Plot
Plot
Flip vector by 180 degrees
Read coordinate
Read y
Flip 180
degrees
y = 0 .38 j0 .36
Y=
1
(0.38 j0.36 )
50
Y = (7 .6 j7 .2 )x10 3 mhos
94
Smith Chart
Constant Imaginary
Impedance Lines
Impedance
Z=R+jX
=100+j50
Normalized
z=2+j for
Zo=50
Constant Real
Impedance Circles
95
Smith Chart
Impedance divided by line impedance (50
Ohms)
Z1 = 100 + j50
Z2 = 75 -j100
Z3 = j200
Z4 = 150
Z5 = infinity (an open circuit)
Z6 = 0 (a short circuit)
Z7 = 50
Z8 = 184 -j900
96
Toward
Generator
Away From
Generator
Constant
Reflection
Coefficient Circle
Scale in
Wavelengths
Matching Example
Ps
Z 0 = 50
100
=0
Match 100 load to a 50 system at 100MHz
A 100 resistor in parallel would do the trick but of the
power would be dissipated in the matching network. We want
to use only lossless elements such as inductors and capacitors
so we dont dissipate any power in the matching network
99
Matching Example
We need to go from
z=2+j0 to z=1+j0 on
the Smith chart
We wont get any
closer by adding series
impedance so we will
need to add something
in parallel.
We need to flip over
to the admittance
chart
Impedance
Chart
100
Matching Example
y=0.5+j0
Before we add the
admittance, add a
mirror of the r=1
circle as a guide.
Admittance
Chart
101
Matching Example
y=0.5+j0
Before we add the
admittance, add a
mirror of the r=1
circle as a guide
Now add positive
imaginary admittance.
Admittance
Chart
102
Matching Example
y=0.5+j0
Before we add the
admittance, add a
mirror of the r=1
circle as a guide
Now add positive
imaginary admittance
jb = j0.5
jb = j0 .5
j0 .5
= j2 (100 MHz )C
50
C = 16 pF
16 pF
100
Admittance
Chart
103
Matching Example
We will now add series
impedance
Flip to the impedance
Smith Chart
We land at on the r=1
circle at x=-1
Impedance
Chart
104
Matching Example
Add positive imaginary
admittance to get to
z=1+j0
jx = j1 .0
( j1.0 )50 = j2 (100 MHz )L
L = 80 nH
80 nH
16 pF
100
Impedance
Chart
105
Matching Example
This solution would
have also worked
32 pF
160 nH
100
Impedance
Chart
106
Matching Bandwidth
0
80 nH
5
16 pF
100
10
100 MHz
15
20
25
30
35
40
50 MHz
50
60
70
80
90
100
110
Frequency (MHz)
120
130
140
150
Impedance
Chart
107