ECE 5317-6351
Microwave Engineering
Fall 2011
Prof. David R. Jackson
Dept. of ECE
Notes 16
S-Parameter
Measurements
1
S-Parameter Measurements
S-parameters are typically measured, at microwave frequencies,
with a network analyzer (NA).
These instruments have found wide, almost universal, application
since the mid to late 1970’s.
Vector network analyzer: Magnitudes and phases of the S
parameters are measured.
 Scalar network analyzer: Only the magnitudes of the Sparameters is measured.
2
Vector Network Analyzer (VNA)
Port 1
a1
b1
DUT
Hewlett-Packard 8510
DUT
b2
a2
3
Network Analysis of VNA Measurement
Port 1
Vector Network
Analyzer
Measurement
plane 1
a1
Measurement Port 2
plane 2
a2
b1
b2
Device
under
test
(DUT)
Test cables
4
S-Parameter Measurements
We want to measure
[S] for DUT
m
a
Port 1 1
b1m
Error
Box A
Meas. plane 1
Error
Box B
DUT
Ref. plane
Ref. plane
a2m
m
2
Port 2
b
Meas. plane 2
Error boxes contain effects
of test cables, connectors, couplers,…
5
S-Parameter Measurements (cont.)
Error Box A
m
1
a
A
S21
S11A
S12A
S21
1
1
S12
S12B
1
1
1
b2m
S11B
B
S22
S22
S11
A
S22
b1m
Error Box B
DUT
B
S21
a2m
Assume error boxes are reciprocal
A
B
 S21
 S12A and S21
 S12B
A
B
W e n e e d to " ca lib ra te " to fin d  S  a n d  S  .
A
B
If  S  a n d  S  a re k n o w n  w e ca n e xtra ct  S  fro m m e a su re m e n ts .
This is called “de-embedding.”
6
Calibration
“Short, open, match” calibration procedure

S21
1
a1m
1

S22

S11
SC
-1
Connect
OC
+1
Z0
0
b1m

S12
1
short
1
open
  A,B
Calibration loads
S 

S
m
11 SC

 S 11 
S
S
m
11 m atch

 S 11 

 S 11
2
Recall from Notes 15:
21

1  S 22
S 

m
11 O C
match
2
21

1  S 22
3 m e a s u re m e n ts :
m
m
( S 11 , S 11
SC
m
OC
, S 11
)
m atch
 in  S 11 
S 21 S 12  L
1   L S 22
3 u n k n o w n s fo r e a c h p o rt :
S



, S 21 , S 22
11

7
Calibration (cont.)
“ Thru-Reflect-Line (TRL)” calibration procedure
This is an improved calibration method that involves three types of connections:
1) The “thru” connection, in which port 1 is directly connected to port 2.
2) The “reflect” connection, in which a load with an (ideally) large (but
not necessarily precisely known) reflection coefficient is connected.
3) The “line” connection, in which a length of matched transmission
line (with an unknown length) is connected between ports 1 and 2.
The advantage of the TRL calibration is that is does not requires precise short, open,
and matched loads.
This method is discussed in the Pozar book (pp. 193-196).
8
Z-Parameter Extraction
Assume a reciprocal and symmetrical waveguide or
transmission-line discontinuity.
T
Examples
g
T
Microstrip gap
Waveguide post
Discontinuity model
T
We want to find Z1
and Z2 to model
the discontinuity.
T
Z1
Z1
Z0
Z2
Z0
9
Z-Parameter Extraction (cont.)
T
T
Z1
Z1
Z0
Z2
Plane of
symmetry
(POS)
Z0
POS
T
T
Z1
Z0
Z1
2Z 2
2Z 2
Z0
10
Z-Parameter Extraction (cont.)
Assume that we place a short or an open along the plane of symmetry.
T
T
POS
Z1
Z0
Z1
2Z 2
2Z 2
SC
Z0
ZL
Z0
ZL
 Z1
Z LSC
POS
Z1
Z1
OC
2Z 2
Z0
2Z 2
 Z1  2 Z 2
Z LOC
Z1  Z L
SC
Z2 
1
2
Z
SC
L
 ZL
OC

11
Z-Parameter Extraction (cont.)
The short or open can be realized by using odd- or even-mode excitation.
T
Incident voltage waves
+
-
Odd mode excitation
+
+
Even mode excitation
The even/odd-mode analysis is very useful in analyzing devices (e.g., using HFSS).
12
De-embeding of a Line Length
We wish the know the reflection coefficient of a device under test (DUT), but the DUT is
not assessable directly – it has an extra length of transmission line connected to it.
S11m
S11   L 
L
m , DUT
S 11
 S 11
DUT
e
Z0 , 
 j2 L
Z L  Z0
Z L  Z0
ZL
DUT
Ref. plane
Meas. plane
R e p la ce D U T  Z L  w ith short circuit
 S 11
m , SC
 e
S
SC
11
  1
 j2 L

L  S
DUT
11
 S
m ,D U T
11
 1 
 m , SC 
 S 11 
13