RF and Microwave Test and Design Roadshow 5 Locations across Australia and New Zealand

RF and Microwave Test and Design Roadshow 5 Locations across Australia and New Zealand

Advanced VNA Measurements

Agenda Overview of the PXIe-5632 Architecture SW Experience Overview of VNA Calibration SOLT Technique Reference Plane extension Typical VNA Measurements Advanced Features of the 5632 Source Loops Frequency Translation Mode Time Domain Analysis Pulsed S-Parameters Intro to non-linear VNA s 3

NEW! NI PXIe-5632 Vector Network Analyzer Specifications NI PXIe-5632 NI PXIe-5630 Frequency Range 300 khz to 8.5 GHz 10 MHz to 6 GHz Architecture 2-port, full S-parameters PXI Express, 3 slots 2-port, 1-path (T/R) PXI Express, 2 slots Power Range -30 to +15 dbm, 0.01 db steps -30 to +5 dbm, 0.5 db steps IF Bandwidth 10 Hz to 500 khz 10 Hz to 30 khz Dynamic Range Measurement Time >110 db (115 db typ) below 6 GHz >105 db (110 db typ) below 8 GHz <65 µsec/pt (100 khz IFBW) Forward or Reverse S-parameters Number of Points 20,001 3,201 >110 db below 3 GHz >100 db below 6 GHz < 200 µsec/pt (30 khz IFBW) Forward S-parameters only 4

Port 1 Port 2 Port 1 Port 2 Architectures of NI VNA s PXIe-5632:Full S-Parameter PXIe-5630: T/R Test Set a 1 a 2 a 1 LO LO b b 1 2 Front Panel 1 2 DUT b b 1 2 Front Panel 1 2 DUT Measure S11, S21, S22, S12 Measure S11, S21 5

NI PXIe-5632 VNA Architecture Dual Sources Independently tuned Improved isolation, increased performance Flexible measurement capabilities Source Loops Flexible measurement capabilities Pulsed S parameters, extended power range, two tone measurements Power Level Control -30 to +15 dbm source power range Very fine (0.01 db) power resolution Low Latency Trigger Improved determinism for pulsed S-parameter measurements Fast IF Architecture 10 Hz to 500 khz IF BW Fast Test Times Linear, Segmented, and Power Sweeps 6

NI-VNA API Supports LabVIEW, CVI and C/C++ ADEs Very Simple API Fewer Vis Than Other RF APIs Setup Sweep, Cal, and Measure OR Load Cal and Measure Can Perform Calibration via SFP or API Recommend Calibrating in SFP, Saving Cal, and Loading Cal for NI-VNA Shipping Examples Demonstrates SFP and LabVIEW / NI- VNA API NI-VNA Shipping Examples nivna_sparammeasure.vi 7

nivna_sparammeasure.vi 8

Understanding VNA Calibration

First Step in Any VNA Application is Calibration Always ask your customer about a calibration kit! NI Automatic VNA Calibration Kit [K-type (MF, MM, FF)] NI Manual VNA Calibration Kit [K-Type, N-type)] Customer s Manual Calibration Kit Enter in calibration data into VNA SW to re-use customer s existing manual calibration kit 10

Calibration Families SOLT = Short, Open, Load, Thru Based on the 12-term error model Popular for end of the cable testing Susceptible to human error TRL = Thru, Reflect, Line Based on an 8-term error model Frequently used when SOLT is not practical Common with probing and de-embedding Highly accurate Often users buy or create customer standards Automatic or Electronic (E-Cal) Internal temperature stabilized Based on modeled impedances Cal constants stored in an internal EPROM Fast, repeatable, convenient and eliminates virtually all human error 12

SOLT Calibration Standards Open Short Incident Wave Incident Wave Open Short Reflected Wave Reflected Wave Γ = ρ = 1 Z = Reflected wave is in-phase with the incident wave Γ = ρ = 1 Z = 0 Reflected wave is 180 out-of-phase with the incident wave 13

SOLT Calibration Standards Load Thru Incident Wave Incident Wave Open Load Short Thru No Reflected Wave Γ = ρ = 0 Z = Z 0 Impedance is equal to the characteristic impedance Typically 50 Ω for RF systems Incident wave measured to correct for frequency response errors 14

Automated SOLT Calibration Short Short Port 1 Open Load Open Load Port 2 Thru Connecting in an automated way different known impedances (not necessarily Short, Open, Load ) M Electronic Calibration Module F 15

Dealing with Non-Insertables Ideal to calibrate with same connector type and gender as the DUT next to never a reality ƒ = 6 GHz λ = 5 cm Phase-Equal Adapters At 6 GHz 1 mm = 7.2 Thru Port 1 F M DUT M M Port 2 16

Test Fixtures Required for majority of automated test Wide variety of surface mount components (not a simple SMA connection) ƒ = 6 GHz λ = 5 cm At 6 GHz 1 mm = 7.2 To Port 1 To Port 2 Unknown phase and magnitude response DUT SMA or other standard connection type 17

Custom Calibration Standards Short Open Load Thru 19

Source and Receiver Calibration Used when absolute power accuracy is important 1. Calibrate VNA Source with NI USB Power Meter 2. Use Calibrated Source to Calibrate VNA Receivers 20

Advanced Features of the PXIe- 5632 Access Loops Frequency Translation Mode Time Domain Analysis Pulsed S-Parameters

Feature: Access Loops Access loops allow customers to reconfigure the VNA. Source access loops Receiver access loops o Reference receiver (a x ) access o Measurement receiver (b x ) access HW in the loops becomes part of the VNA s signal path The idea is to add HW behind the directional devices where it does NOT impact system performance. HW in the loop gets corrected by user calibration and does not degrade accuracy. 31

Access Loops in VNAs 4 loop positions per port S1, S2, a1& b1 Box VNAs have plenty of room Most higher end offer 3 loops PXIe-5632 VNA Limits space so we chose source loop, S2. 32

What can we do with the source access loops? Source access loop can be used to add the following: Attenuator: Lower Power Measurements Pulse modulator: Pulsed Measurements Combiner: Combined Dual Source for TOI measurements ATTN Pulse Mod 33

Port 1 Port 2 Feature: Frequency Translation Mode? PXIe-5632:Full S-Parameter a1 a2 Each source can be tuned to a different frequency, enabling us to stimulate multi-port devices LO b1 b2 Front Panel 1 2 DUT Measure S11, S21, S22, S12 34

Applications with Dual Sources PXIe-5632 VNA PXIe-5632 VNA Source 1 Source 2 Port 1 Loop Loop Port 2 Source 1 Source 2 Port 1 Loop Loop Port 2 RF LO IF DUT Two-Tone Stimulus Mixer Measurements IP3/TOI Measurements 35

Intermodulation Distortion (IM3) Third-Order Distortion Product Intermodulation Distortion (IM3) Second-Order Distortion Products Third-Order Distortion Products ƒ 1 + ƒ 2 2ƒ1 + ƒ2 ƒ1 + 2ƒ2 ƒ 2 - ƒ 1 2ƒ 1 - ƒ 2 2ƒ 2 ƒ 1 ƒ 1 ƒ 2 2ƒ 1 2ƒ 2 3ƒ 1 3ƒ 2 Frequency Two-tone stimulus at frequencies ƒ 1 and ƒ 2 36

Intermodulation Distortion with VNA s VNA sources are tuned to be slightly offset in frequency Receiver measures 3 rd order distortion products ajacent to stimulus frequencies Two-Tone Stimulus from VNA w/combiner DUT (RF PA) Intermodulation Distortion (IM3) Measures Ratio 37

Feature: Time Domain Analysis VNAs acquire mag and phase data across a range of frequencies Can use this data as inputs to an Inverse FFT (IFFT) Resulting IFFT output is the time domain waveform whose frequency domain content is the previously acquired VNA data Use the speed of light (through the DUT) to calculate distance Provide the VNA with the DUT s velocity factor: D = c * V f * t Concepts you have learned about FFTs apply, just in the opposite direction Aliasing, resolution, etc More on this in ensuing slides. 41

Time Domain Analysis - Example Cable Fault Location A cable normally has low return loss A break causes complete reflection - looks like an open circuit High return loss (~0 db) Time Domain allows the location of the open to be determined Open @ 8.55 cm 20*log(0.932) = -0.6 db 45

Time Domain Analysis - Resolution Resolution in time/distance is one figure of merit for TDA Resolution determines the degree to which closely-spaced impedance mismatches can be resolved Resolution in time domain (sec) determined by the bandwidth of the VNA frequency sweep (Hz) Just like frequency resolution (Hz) of forward FFT is determined by amount of input time data (sec) Rule of thumb: resolution 150 (mm) / Span (GHz) Example: 6 GHz sweep span resolution 25 mm Example: 20 GHz sweep span resolution 8 mm Example: 70 GHz sweep span resolution 2 mm 46

FFT Time Domain Analysis - Resolution Increasing frequency domain bandwidth leads to finer time domain resolution Bandwidth (Hz) IFFT Resolution (Δt) Δt Analogous to sample rate determining bandwidth of forward FFT 47

Time Domain Analysis Alias-Free Range Alias-free range calculation example Frequency sweep from 10 MHz to 6 GHz with 600 points Δf = 10 MHz Alias-free Range (m) = (1/Δf) * c * Velocity Factor Use 0.7 Velocity Factor for Teflon cable dielectric Alias-free Range (m) = 20.99 m Maximum DUT length for transmission measurements is 20.99 m Maximum DUT length for reflection measurements is 10.50 m 51

Introduction to Non-linear VNA s

Terms and Definitions Small-scale network analysis is the analysis of RF devices in the linear operating region (S-parameters) Large-scale network analysis is the analysis of RF devices in the full operating region, both linear and non-linear (PHD measurements) Polyharmonic Distortion (PHD) measurement systems are describing functions that relate the phase and magnitude of harmonics to the phase and magnitude of the input signal. X-Parameters are PHD measurements from Agilent Cardiff Model are PHD measurements offered by Cardiff/Mesuro S-Functions are PHD measurements by NMDG (now NI) 67

S-Functions, X-Parameters, and the Cardiff Model RF devices are inherently non-linear Non-linearity produces harmonics These are not capture in S-Parameters PHD Measurements (S-Functions, X-Parameters, and the Cardiff model) capture & describe harmonically related products Input PA Output S-Parameters capture this 68 PHD Measm ts also capture this

Polyharmonic Distortion Models Accurate phase/magnitude measurements require thousands of measurements PXI solution reduced measurement time from several hours to 6 seconds 71

Instrumentation Architecture for PHD Measurements PHD measurements require multi-channel, phasesynchronized generators & receivers. Each generator and receiver is tuned to a different harmonic of fundamental tone. Cardiff system uses PXIe-5673 s as sources, and PXIe- 5663 s as the receivers. PXIe-5663 Syn LO s f 1 f 1 f 2 PXIe-5673 PXIe-5673 PXIe-5673 PA DUT PXIe-5663 PXIe-5663 PXIe-5663 Syn LO s f 3 PXIe-5673 PXIe-5663 f 1 PXIe-5652 Cardiff phase reference standard 72

PHD Models Summary PHD models are describing functions that include: S-parameters for for every harmonic that are also a function of input power Measuring them requires: Generation and analysis of multi-tone signals Phase-synchronized multi-frequency stimulus Phase-synchronized multi-frequency receiver PHD models include: X-parameters (Agilent) Cardiff Model (Cardiff/Mesuro) S-Functions (NMDG now part of NI) 73

Summary Calibration of VNA s is relatively straightforward Use SOLT technique and make your DUT insertible NI Software-defined VNA s can solve a wide range of applications Standard passive component testing Cable fault detection Pulsed S-parameters Industry is moving towards non-linear VNA measurements as well Cardiff model S-Functions 74