![5
Error
Box A
Error
Box B
DUT
Port 1 Port 2
Meas. plane 1 Meas. plane 2
Ref. plane Ref. plane
1
m
a 2
m
a
2
m
b
1
m
b
We want to measure
[S] for DUT
S-Parameter Measurements
Error boxes contain effects
of test cables, connectors, couplers,…](https://image.slidesharecdn.com/notes16-sparametermeasurements-250323015624-1a6dbb69/85/Notes-12-Surface-waves-pptx-Notes-12-Surface-waves-pptx-5-320.jpg)
Notes 12 - Surface waves.pptx Notes 12 - Surface waves.pptx
- 1. 1 Prof. David R. Jackson Dept. of ECE Notes 16 ECE 5317-6351 Microwave Engineering Fall 2011 S-Parameter Measurements
- 2. 2 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 S- parameters is measured. Most NA’s measure 2-port parameters. Some measure 4 and 6 ports.
- 3. 3 Test cables Hewlett-Packard 8510 DUT Port 1 Port 2 1 a 1 b 2 b 2 a DUT Vector Network Analyzer (VNA)
- 4. 4 Network Analysis of VNA Measurement Vector Network Analyzer Device under test (DUT) Port 1 Port 2 Measurement plane 1 Measurement plane 2 1 a Test cables 1 b 2 a 2 b
- 5. 5 Error Box A Error Box B DUT Port 1 Port 2 Meas. plane 1 Meas. plane 2 Ref. plane Ref. plane 1 m a 2 m a 2 m b 1 m b We want to measure [S] for DUT S-Parameter Measurements Error boxes contain effects of test cables, connectors, couplers,…
- 6. 6 " " . A B S S and We need to calibrate to find A B S S S and If are known we can extract from measurements. 21 12 21 12 A A B B S S S S Assume error boxes are reciprocal and 1 m a 21 A S 11 A S 12 A S 22 A S 22 B S 12 B S 11 B S 21 B S 1 1 11 S 21 S 22 S 12 S 1 1 1 Error Box A Error Box B DUT 1 m b 2 m a 2 m b S-Parameter Measurements (cont.) This is called “de-embedding.”
- 7. 7 “Short, open, match” calibration procedure , A B 2 21 11 11 22 2 21 11 11 22 11 11 1 1 SC OC match m m m S S S S S S S S S S 11 11 11 11 21 22 3 ( , , ) 3 , , SC OC match m m m S S S S S S measurements: unknowns for each port : Calibration loads 1 m a 1 m b 1 1 21 S 11 S 12 S 22 S 1 1 Connect Calibration SC -1 OC +1 Z0 0 short open match 21 12 11 22 1 L in L S S S S Recall from Notes 15:
- 8. 8 “ Thru-Reflect-Line (TRL)” calibration procedure Calibration (cont.) 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).
- 9. 9 Z-Parameter Extraction Assume a reciprocal and symmetrical waveguide or transmission-line discontinuity. Examples Discontinuity model 1 Z 0 Z 0 Z T T 1 Z 2 Z g T T Waveguide post Microstrip gap We want to find Z1 and Z2 to model the discontinuity.
- 10. 10 2 Z 1 Z 1 Z 0 Z 0 Z T T Plane of symmetry (POS) 0 Z 1 Z 2 2Z 1 Z 2 2Z 0 Z T T POS Z-Parameter Extraction (cont.)
- 11. 11 Assume that we place a short or an open along the plane of symmetry. 1 2 1 2 SC SC OC L L L Z Z Z Z Z 1 2 2 OC L Z Z Z 1 SC L Z Z 0 Z 1 Z 2 2Z SC L Z POS T T 0 Z 1 Z 2 2Z 0 Z 1 Z 2 2Z OC L Z POS 0 Z 1 Z 2 2Z Z-Parameter Extraction (cont.)
- 12. 12 The short or open can be realized by using odd- or even-mode excitation. Z-Parameter Extraction (cont.) T + - + + Odd mode excitation Even mode excitation Incident voltage waves The even/odd-mode analysis is very useful in analyzing devices (e.g., using HFSS).
- 13. 13 De-embeding of a Line Length , 2 11 11 m DUT DUT j L S S e 11 , 2 11 1 SC L m SC j L Z S S e with short circuit Replace DUT L Z L 11 m S 0 11 0 L L L Z Z S Z Z 0, Z DUT Meas. plane Ref. plane , 11 11 , 11 1 DUT m DUT L m SC S S S 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.