Vector Network Analyzer VNA Basics VNA Roadshow Budapest 17/05/2016
Content Why Users Need VNAs VNA Terminology System Architecture Key Components Basic Measurements Calibration Methods Accuracy and Uncertainty Advanced Measurements 2
Why Users Need VNAs VNAs are used to characterize RF performance of high frequency devices Used in multiple environments Characterize devices for research and development Test devices during manufacturing Verify conformance for maintenance Used with different package and/or configuration devices Coaxial connectors (most common) 2-port to multi-port Single-ended or differential input Waveguide connectors Connectorlesswith fixtures (microstrip, stripline) On-wafer with probers and probe tips 3
Why Users Need VNAs (continued) Used for multiple measurements S-Parameters (transmission and reflection characteristics) Gain Compression (replacing a power meter) Inter-Modulation Distortion (replacing a spectrum analyzer) Harmonics (replacing a spectrum analyzer) Noise Figure Time Domain (distance-to-fault, impedance measurements at specific locations, replacing a scope or TDR/TDT analyzer) Used to test many types of devices/components/sub-systems Amplifiers, Converters/Mixers, Switches, Semiconductors, ICs Filters, Diplexers, Attenuators, Adapters, Circulators, Isolators, Splitters, Couplers Cables, Connectors, Backplanes Antennas, Receivers, Transmitters Optical Modulators and Receivers 4
VNA Terminology Vector vs. Scalar Passive vs. Active Devices Passive consumes but does not produce energy; the device does not amplify the signal: Loss, Filters, cables, connectors, adapters, mixers Active the device has a DC voltage or voltages applied to amplify the signal: Gain, Amplifiers, active filters, converters Transmission vs. Reflection The decrease/increase in signal power resulting from the insertion of a device into the transmission line. A ratio of the transmitted signal to the incident signal A measure of the signal reflected at an interface such as a connector S-Parameters Scattering Matrix S-Parameters are a short hand method of describing reflection and transmission characteristics for a two-port network Phase is part of S-Parameters; only vector network analyzers can perform S- Parameter measurements It is easier to say S11 than forward reflection coefficient S11 S12 S21 S22 5
Functional Block Diagram of Test Set and Source A VNA consists of three distinct blocks: Source The frequencysource(stepped synthesized sweep) Transfer Switch The microwave test set Reflectometer components (directional couplers, bridges) Receivers (samplers, mixers) The network analyzer Digital and analog hardware to analyze the IF signals Firmware and software Front panel input, display, and memory... Reflectometer b1 a1 a2 b2 Receivers Reflectometer Port 1 Port 2 DUT 6
Key Components of a VNA Source Switches Receivers Reflectometers 7
Source Designed for optimum combination of speed and spectral purity (phase noise) Locally synthesized, no locking required through IF Allows source and receiver LO to be at different frequencies Phase noise directly impacts high level noise Trace Noise (or High Level Noise) 0.03 0.02 0.01 db 0-0.01-0.02-0.03 0 20 40 60 Frequency (GHz) 8
Switches High isolation switch technology Different switch technologies used (FET low frequency, PIN high frequency) for best insertion loss, isolation, and compression performance Leads to low leakage and better dynamic range Low Band Transfer Switch Isolation High Band Transfer Switch Isolation -20 0-60 -40 db -100 db -80-140 -120-180 0.01 0.1 1 10 100 1000 10000 Frequency (MHz) -160 0 10 20 30 40 50 60 70 Frequency (GHz) 9
Receivers Mixers used at lower frequencies, samplers used at higher frequencies Harmonic mixing approaches tend to have significant roll-off at higher frequencies compared to samplers New sampler and impulse technologies (MMIC-based) Higher LO frequencies for lower noise figure and better dynamic range Conversion loss does not noticeably increase until well beyond 110 GHz Inherent symmetry of sampler structure allows improved linearity (higher IP3) Normalized Conversion Efficiency (<100 MHz IF) Port-Referred IP3 db 0 Sampler -10-20 Mixer -30-40 10 30 50 70 Frequency (GHz) Sampler Mixer1 Mixer2 IP3 (dbm) 45 40 35 30 25 20 15 10 5 0 Sampler Sampler Mixer Harmonic Mixer System 0 20 40 60 80 Frequency (GHz) 10
Reflectometers Bridges used at lower frequencies, directional couplers used at higher frequencies Bridges do not rely on wavelength-related physical lengths to provide coupling, thus, extending down to much lower frequencies Directional couplers roll-off severely at lower frequencies, impacting dynamic range and high level noise Un-normalized coupling 0-5 -10-15 db -20-25 -30-35 coupler bridge -40 0.045 0.545 1.045 1.545 2.045 Frequency (GHz) 11
Example: Hybrid VNA Architecture Two VNAs in parallel: 6 decades of frequency coverage (from 70 khz to 70 GHz) Each receiver technology (sampler or mixer) used in its best range Each reflectometer technology (coupler or bridge) used in its best range Both share a common IF path and fully synthesized source > 2.5 GHz High Band MS4640B Block Diagram (Fully Loaded Configuration) optional < 2.5 GHz Low Band a 1 a 1 a 2 a 2 b 1 b 2 b 1 b 2 Bias 1 Bias 2 Port 1 Port 2 12 12
Basic VNA Measurements S-Parameters Transmission Reflection 13
VNA Measurement Set-Up Port 1 Test Cable Port 2 Test Cable Port 1* Device Under Test (DUT) Port 2* *When using test cables 14
What are S-Parameters? S21 -The first digit is the port number where the signal is measured (i.e., Port 2). The second digit is the port from which the signal originates (i.e., Port 1) Therefore, S21 is the measurement of gain or loss in the forward direction, magnitude and phase Examples: S11 Forward Reflection S21 Forward Transmission S12 Reverse Transmission S22 Reverse Reflection S11 S12 S21 S22 In addition, S-Parameters are in a standard format that permits engineers to perform mathematical analysis, manipulation and simulation (ex., s2p format) 15
S-Parameter Definitions S11= b1/a1 Forward Reflection S21= b2/a1 Forward Transmission S22= b2/a2 Reverse Reflection S12= b1/a2 Reverse Transmission VNA measures ratio of two signals Magnitude and phase S 11 FORWARD REFLECTION S 21 FORWARD TRANSMISSION Port 1 Port 2 a 1 b 2 b 1 DUT S 12 REVERSE TRANSMISSION b 1 S 11 a 1 + S 12 a 2 S 22 REVERSE REFLECTION a 2 b 2 S 21 a 1 + S 22 a 2 16
Measuring Forward Parameters, S11 (b1/a1), and S21 (b2/a1) Source - Receiver a 1 measures the incident signal onto the DUT Step Attenuator Forward Transfer Switch - Receiver b 1 measures the reflected signal back from the DUT - Receiver b 2 measures the transmitted signal through the DUT Direct Access Loops a 1 a 2 4 Receivers Step Attenuator Reflectometer Reflectometer b 1 b 2 Port 1 Port 2 DUT 17
Measuring Reverse Parameters, S22 (b2/a2), and S12 (b1/a2) Step Attenuator Source Transfer Switch Reverse - Receiver a 2 measures the incident signal onto the DUT s output - Receiver b 2 measures the reflected signal back from the DUT s output - Receiver b 1 measures the reverse transmitted signal through the DUT Direct Access Loops a 1 a 2 Reflectometer 4 Receivers b 1 b 2 Step Attenuator Reflectometer Port 1 Port 2 DUT 18
Transmission Measurements vs reflection measurements 19
Reflection and Transmission Basic Measurements and VNA Displays Input Reflection Coefficient Reverse Transmission (magnitude and phase) Forward Transmission (magnitude and phase) Output Reflection Coefficient 20
Reflection and Transmission Basic Measurements and VNA Displays Input Return Loss Insertion Loss Insertion Loss Output Return Loss 21
Other Display Formats Complex Impedance R + jx Short Open Circles are constant resistance (R) Arcs are constant reactance (X) Distance from the center is magnitude Rotation around the circle is phase Group delay is the rate of phase change with respect to frequency For linear devices, phase change is linear with frequency, and group delay is uniform Non-uniform group delay is one cause of distortion in wideband digital transmission systems 22
VNA Calibration Standards Types Algorithms Precision AutoCal 23
Instrument vs. Measurement Calibration Instrument Calibration performed per recommended cal interval at your local service center Ensure instrument meets published specifications Standard, Z540, Premium Measurement Calibration the procedure the VNA user goes through as part of the setup to measure the device under test Establish 50 ohm reference Zero db, zero degrees = reference plane Manual or AutoCal Short-Open-Load-Thru, Line-Reflect-Line, etc. 24
Without Calibration a VNA can t Make Accurate Measurements Calibration is fundamental to making accurate measurements VNA Port 1 VNA Port 2 DUT Defines reference planes (0 db, 0 degrees) Removes systematic errors due to frequency response, source/load match, directivity, and isolation (crosstalk) Random errors cannot be calibrated out (ex., connector repeatability, cable stability, environmental changes) 25
VNA Calibration Standards S-Parameter measurements Calibration kit (coax and waveguide) (open, short, load standards) Precision AutoCal module Microstrip kit for test fixture Wafer calibration substrate Gain compression measurements Power meter and sensor 26
Calibration Types 27
How to Calibrate a VNA Purpose: To reduce all systematic errors for both ports (forward and reverse), a full 12-term calibration is required Short-Open-Load-Thru (SOLT) The most common coax calibration algorithm Other calibration algorithms Line-Reflect-Line (LRL) Line-Reflect-Match (LRM) Thru-Reflect-Match (TRM) Offset Short, etc. Exercise good practice for best results Technique, care, knowledge, proper torque, clean connectors 28
Calibration Algorithms Calibration Algorithm Description Advantages Disadvantages SOLT (Short Open Load Thru) SSLT (Short Short Load Thru), shorts with different offset lengths SSST (Triple Short Thru), all shorts with different offset lengths SOLR, SSLR, SSSR. Like above but with reciprocal instead of thru LRL (Line Reflect Line) Also called TRL ALRM (Advanced Line Reflect Match) also called TRM Commonly coaxial Simple, redundant standards, not band limited Requires very well defined standards, poor on wafer, lower accuracy at higher frequencies Commonly waveguide Same as SOLT Same as SOLT, and band limited Commonly waveguide or high frequency coaxial Like above but when a good thru is not available High performance coaxial, waveguide or on-wafer Relatively high performance Same as SOLT but better accuracy Does not require well defined thru Highest accuracy, minimal standard definition Requires very well defined standards, poor on wafer, band limited Some accuracy degradation, but slightly less definition, other disadvantage of parent calibration Requires very good transmission lines, less redundancy so more car needed, band limited High accuracy. Only one line length, so easier to fixture/on wafer, not band limited usually Requires Load definition. Reflect standard setup may require care depending on load model used. 29
How Does Calibration Work? The VNA measures known standards Error Vectors It will compare the measured value to the known value, and calculate the difference y The difference is the error It will store an error coefficient (magnitude and phase) at every frequency/data point, and use it when making measurements x 30
Systematic Errors The six error terms below, on both ports, yield the 12-term error corrected model: Transmission Frequency Response (E TF, E TR ) Reflection Frequency Response (E RF, E RR ) Source Match (E SF, E SR ) Load Match (E LF, E LR ) Directivity (E DF, E DR ) Isolation (E XF, E XR ) These systematic errors are reduced by measurement calibration performed by the user 31
12-Term Error Corrected Model E XF RF IN E DF 1 E SF S11 A S21 A S22 A E LF E TF S21 M S11M E RF S12 A S21 A E RR S22 M E TR E LR S11 A S22 A E SR E DR S12 M S12 A 1 RF IN E XR E:Error D:Directivity S:Source Match L:Load Match F:Forward R:Reverse M:Measured A:Actual X:Isolation T:Transmission Response R : Reflection Response 32
Random Errors Connector Repeatability Cable Stability Environmental Changes Frequency Repeatability ALL of the above are not predictable, and therefore, not correctable Good measurement practice is critical to reducing random errors 33
Precision AutoCal Module Fewer connections (and fewer mistakes) than when performing a manual calibration Better accuracy than a manual SOLT calibration New switched hybrid technology Port 1 Γ2 Γ3 Port 2 New calibration algorithm Very low loss thrus -> much less degradation at higher frequencies Γ1 Switch networks Γ4 34
AutoCal Algorithm Overdetermined calibration: more standards are available in the module than absolutely necessary Makes the calibration more robust Γ1 Γ2 Γ3 Γ4 Γ5 Γ1 Γ2 Γ3 Γ4 Γ5 Τ1 Τ2 Τ3 35
AutoCal Standards The reflection standards must be far enough apart on the Smith chart that they provide independent information Designing a switching system with low loss and low parasitics makes this possible +j1.0 +j1.0 +j5.0 +j5.0 0.2 5.0 0.0 0.2 5.0 0.0 -j5.0 -j5.0 -j1.0 Too close together: unstable calibration -j1.0 Far enough apart: good calibration possible 36
VNA Accuracy Specifications Uncertainties 37
System Dynamic Range Calculated as the difference between the maximum rated source power and the specified noise floor in a specific frequency band 38
Corrected System Performance Residuals after performing user calibration with specified standards Performance measured using a metrology grade airline 39
Measurement Uncertainties The accuracy of a VNA measurement is affected by the following Dynamic range of the VNA: IFBW, averaging Corrected system performance: directivity, source/load match, etc. Error correction type used: 12-Term, 1 Path 2 Port, etc. The quality of the calibration standards: discrete standards, sliding loads, airlines, AutoCal, etc. Cable stability and connector repeatability DUT performance Uncertainty software (available on Anritsu website) to generate uncertainty curves http://downloadfiles.anritsu.com/files/en-us/software/drivers-software- Downloads/ExactUncertaintySetup_2.02.exe 40
Uncertainty Curves VNA Accuracy 41
Advanced VNA Measurements Time Domain Gain Compression 42
Time Domain Transform frequency domain data into time domain information Identify discontinuities in cables, connectors, transmission lines in either time or distance 43
Measuring Devices in the Frequency and Time Domains An unterminated low loss device provides little useful information in the frequency domain Time domain extracts useful information normally unavailable in the frequency domain 44
Time Domain Resolution and Frequency Bandwidth Two Mismatches Separated by 2mm (air) Span Resolution 1) 40 GHz 3.75mm 2) 50 GHz 3.0mm 3) 70 GHz 2.14mm 110 GHz 1.36mm Resolution determines the ability to resolve one discontinuity from another Resolution is inversely related to the frequency bandwidth Rule of thumb: 150 mm/frequency span (GHz) 45
Gain Compression Power is swept at multiple CW frequencies (up to 401 frequencies possible in VectorStar) Display of output power at each CW frequency while input power is swept Provides traditional P OUT vs. P IN and Phase vs. P IN measurements (AM/PM) 46 46
Advanced VNA Measurements More advanced measurements Intermodulation (IMD) Pulse Measurements Noise Figure Material Measurements 47
Summary VNA is an instrument that provides extremely accurate information on how the device (sub-system or component) performs in a RF/microwave system Provides a ratioed measurement relative to incident signal Provides both magnitude and phase information Wide dynamic range (~120 db) and extremely linear Many application environments; coaxial, waveguide, on-wafer, custom backplane, etc. Many calibration choices to optimize for best accuracy Additional information (application note and webinar) available at: http://downloadfiles.anritsu.com/files/en-us/application-notes/application-note/11410-00387.pdf http://www.anritsu.com/en-us/events/online-webinars/2013/wbn002166.aspx 48
Thanks 49