Spectrum Analyzer Training - PDF Free Download

Spectrum Analyzer Training Roberto Sacchi Application Engineer roberto_sacchi@agilent.com Page 1

Agenda Introduction Overview: What is Signal Analysis? What Measurements are available? Theory of Operation Specifications Modern Signal Analyzer Designs & Capabilities Wide Bandwidth Vector Measurements Basics on digital modulation Measurements on digital modulation Slide 2

Overview What is Signal, Vector and Spectrum Analysis? Spectrum Analysis Display and measure amplitude versus frequency for RF & MW signals Separate or demodulate complex signals into their base components (sine waves) Slide 3

Overview Frequency versus Time Domain Amplitude (power) Time domain Measurements (Oscilloscope) Frequency Domain Measurements (Spectrum Analyzer) Slide 4

Overview Types of Tests Made Modulation Noise Distortion Slide 5

Overview Different Types of Analyzers Swept Analyzer A Filter 'sweeps' over range of interest LCD shows full spectral display f 1 f 2 f Slide 6

Overview Different Types of Analyzers FFT Analyzer A Parallel filters measured simultaneously LCD shows full spectral display f 1 f 2 f Slide 7

Agenda Introduction Overview Theory of Operation Specifications Modern spectrum analyzer designs & capabilities Wide Bandwidth Vector Measurements Basics on digital modulation Measurements on digital modulation Slide 8

Theory of Operation Swept Spectrum Analyzer Block Diagram RF input attenuator mixer IF gain IF filter (RBW) envelope detector Input signal Pre-Selector Or Low Pass Input Filter local oscillator Log Amp video filter sweep generator Crystal Reference Oscillator ADC, Display & Video Processing Slide 9

Theory of Operation Mixer MIXER f sig 1.5 GHz RF IF LO f sig f LO - f sig f + LO f LO f sig f LO 3,6 GHz 6.5 GHz Slide 10

Theory of Operation IF Filter (Resolution Bandwidth RBW) IF Filter Input Spectrum IF Bandwidth (RBW) Display A B C Slide 11

Theory of Operation Envelope Detector Before detector After detector Envelope Detector Slide 12

Theory of Operation Envelope Detector and Detection Types Envelope Detector Digitally Implemented Detection Types ADC, Display & Video Processing bins/buckets* Positive detection: largest value in bin displayed Negative detection: smallest value in bin displayed Sample detection: middle value in bin displayed Other Detectors: Normal (Rosenfell), Average (RMS Power) *Sweep points Slide 13

Theory of Operation Average Detector Type Envelope Detector Volts Pos Peak detection x bin ADC, Display & Video Processing Sample detection Time Power Average Detection (rms) = Square root of the sum of the squares of ALL of the voltage data values in the bin /50Ω x x Neg Peak detection Slide 14

Theory of Operation Video Filter (Video Bandwidth VBW) Video Filter Slide 15

Theory of Operation Other Components RF INPUT ATTENUATOR IF GAIN LO SWEEP GEN LCD Display, ADC & Video processing Slide 16

Theory of Operation How it All Works Together - 3 GHz spectrum analyzer f s Signal Range LO Range 0 1 2 3 (GHz) f LO - f s f LO f LO + f s f s IF filter input mixer 0 1 2 3 4 5 6 3.6 6.5 detector 3.6 GHz sweep generator f IF A LO f LO 3 4 5 6 3.6 6.5 (GHz) 0 1 2 3 LCD display (GHz) f Slide 17

SPECTRUM ANALYZER 9 khz - 26.5 GHz Key Specifications 8563A Safe spectrum analysis Frequency Range Accuracy: Frequency & Amplitude Resolution Sensitivity Distortion Dynamic Range Slide 19

Specifications Resolution What Determines Resolution? Resolution Bandwidth RBW Type and Selectivity Noise Sidebands Slide 20

Specifications Resolution: Resolution Bandwidth Mixer 3 db BW 3 db Envelope Detector Input Spectrum LO IF Filter/ Resolution Bandwidth Filter (RBW) Sweep RBW Display Slide 21

Specifications Resolution: Resolution BW 10 khz RBW 3 db 10 khz Determines resolvability of equal amplitude signals Slide 22

Specifications Resolution BW Selectivity or Shape Factor 3 db 3 db BW 60 db 60 db BW Selectivity = 60 db BW 3 db BW Determines resolvability of unequal amplitude signals Slide 23

Specifications Resolution BW Selectivity or Shape Factor RBW = 1 khz Selectivity 15:1 RBW = 10 khz 3 db 7.5 khz distortion products 60 db 60 db BW = 15 khz 10 khz 10 khz Slide 24

Specifications Resolution: RBW Type and Selectivity ANALOG FILTER Typical Selectivity Analog 15:1 Digital 5:1 DIGITAL FILTER RES BW 100 Hz SPAN 3 khz Slide 25

Specifications Sensitivity/DANL RF Input Mixer RES BW Filter Detector LO Sweep A Spectrum Analyzer Generates and Amplifies Noise Just Like Any Active Circuit Slide 26

Specifications Sensitivity/DANL Effective Level of Displayed Noise is a Function of RF Input Attenuation signal level 10 db Attenuation = 10 db Attenuation = 20 db Signal To Noise Ratio Decreases as RF Input Attenuation is Increased Slide 27

Specifications Sensitivity/DANL: IF Filter(RBW) Displayed Noise is a Function of IF Filter Bandwidth 100 khz RBW 10 db 10 db 10 khz RBW 1 khz RBW Decreased BW = Decreased Noise Slide 28

Specifications Sensitivity/DANL: Summary For Best Sensitivity Use: Narrowest Resolution BW Minimum RF Input Attenuation Sufficient Averaging (video or trace) Slide 29

Specifications Spectrum Analyzer Dynamic Range Dynamic Range The ratio, expressed in db, of the largest to the smallest signals simultaneously present at the input of the spectrum analyzer that allows measurement of the smaller signal to a given degree of uncertainty. Slide 30

Modern Spectrum Analyzer Block Diagram Pre-amp Analog IF Filter Digital IF Filter Digital Detectors FFT Attenuation Sweep vs. FFT Digital Log Amp YIG ADC Slide 32

Wide Band Block Diagram of the PSA option 122 Third Converter WB Analog IF WB Digital IF UPHB* 3 rd LO Calibrator FPGA HB 1 st LO 2 nd LO Low Band 3 rd LO NB IF *Un-preselected High Band Slide 33

Simplified Block Diagram Highband Preselected Mixer A 321.4 MHz IF ACP Module Option D 21.4 MHz IF RF INPUT Input Attenuator 0-70 db 2 db Step 3Hz - 3GHz 1st LO 3-7 GHz 3 GHz 3.9214 GHz 1st IF 3.9 GHz Lowband C 321.4 MHz IF Unpreselected Highband Mixer Digital Demod Option Preamp Option B 321.4 MHz IF 1st LO 3-7 GHz 2nd LO 3.6 GHz PMYO 1st LO 3-7 GHz A B C 321.4 MHz Out 21.4 MHz 3rd IF IF Processing ADC ASIC CPU D fs = 30 MHz 300 MHz LO 28.9 MHz LO Slide 34

rf Analog RBW filter set here RBW IF Processing Past Amplitude Only - Spectrum Analyzer (HP856x) IF2 Log Amp/ Detector ADC Bits LO1 LO2 Detected and logged level called video, mostly dc or slow moving Quadrature - Network Analyzer (HP8510A,..) Bits ADC rf/if Analog Cos(wt) RBW filter set here Sin (wt) Gain/Phase match errors ADC Bits Analog -- Digital Slide 35

IF Digitizers Now Amplitude or Vector Analyzer Push the digitizers up the rf chain ASIC I rf IF MHz ADC Bits Cos(wt) Sin (wt) RBW filters done digitally LO Fs Sample Rate ADC digitizes IF - not detected amplitude Q Q Fewer analog adjustments No temperature dependence Cheaper to manufacture Fewer components More flexible processing Analog -- Digital I Slide 36

Swept tuned measurements of broadband signals RF In IF out BPF Detectors Display processor Swept LO With narrow band measurements some information can be gained, such as amplitude and frequency range occupied by signal. Information contained within the signal will be lost because of the reduced BW. Slide 37

Digitization and FFT of broadband signals RF In IF out WBF ADC Display processor Fixed or step tuned All information is captured using a fast digitizer. An FFT is then performed to view signal in the frequency domain Slide 38

Types of wideband measurements There are basically two types of wide BW measurements: 1. Signal amplitude and lobe width for very narrow pulse radar measured in the pulse mode. 2. Phase and amplitude are needed for complete evaluation such as Chirp Radar 200 MHz linear chirp Slide 39

Instrument and System Calibration Calibration Q Amplitude error Amplitude Flatness Phase linearity Minimum Error Vector Magnitude Measured signal θ EVM Ideal Signal Phase linearity error I The goal is to measure the EVM of the DUT not the EVM introduced by the measuring system Slide 40

Signal used to calibrate IF path Amplitude and phase characterized comb covering the entire 80 MHz information BW Slide 41

Three loops of calibration Outer Loop IF cal Third Converter WB Analog IF WB Digital IF UPHB 3 rd LO HB Low Band 1 st LO 2 nd LO Calibrato r ADC cal Inner Loop IF cal FPGA 3 rd LO NB IF Slide 42

Modern Spectrum Analyzer Block Diagram Digital Detectors 3 GHz PreAmp Improve 1GHz DANL from 153 dbm to 167 dbm Analog IF Filter (Single Pole) Digital IF Filter 160 Settings 1 Hz to 8 MHz RBW 1 Hz to 50 MHz VBW Min Switching Uncertainty Normal Peak Minimum Peak Sample RMS Quasi Peak FFT Attenuator 2 db Step VCO Fast Tune Stepped for FFT Optimization for Close in PhaseNoise Far out PhaseNoise 14 Bit ADC Autoranging Dither on/off Sweep vs FFT Fast Sweep Narrow BW High Selectivity Digital Log Amp Min Linearity Contribution > 100 db Dynamic Range Slide 43

Agenda Introduction Overview Theory of Operation Specifications Modern spectrum analyzer designs & capabilities Basics on digital modulation Measurements on digital modulation Slide 44

Transmitting Information (Analog or Digital) Modify a Signal "Modulate" Detect the Modifications "Demodulate" Any reliably detectable change in signal characteristics can carry information Slide 45

The Communications Hierarchy The OSI Model: We are here Application Presentation Session Transport Network Data Link Physical e.g. Microsoft Exchange Encrypt/cross format translation E to E dialogue, billing etc. Mux, Flow and sequencing Switch, route and order packets Error detection/correction and Frames Raw bits/ Electrical specifications Slide 46

Why do we modulate? to move the signal to a frequency band where the medium has best transmitting properties radio transmission: to reduce antenna physical dimensions (f= 30 KHz = 10 Km) to multiplex multiple users in a given bandwidth Slide 47

Electromagnetic Spectrum Wave length 10 mm 1 mm f (Frequency) = (Velocity) (Wave length) Slide 48

mmwaves Atmospheric Windows Millimeter waves (30-300 GHz) have unique transmission channel characteristics of great interest for: Communications Transportation Scientific Research National Security Minimum attenuation bands 35, 94, 140, 220 GHz Maximum absorption bands 60, 120, 182 GHz Satellite Automotive RADAR National Security (imaging) Scientific Research Slide 49

Atmospheric Windows for Satellite Communications O 2 /H 2 O Minimum attenuation band: 35, 94, 140, 220 GHz Most effective for the satellite-earth signal transmissions? Maximum absorption band: 60, 120, 182 GHz Hard to be intercepted Effective for secured intersatellite transmissions Slide 50

Voltage Voltage Digital vs. Analog 1.5 1.5 1 1 0.5 0.5 0 0-0.5-0.5-1 -1-1.5 0 2 4 6 8 10 12 14 Time -1.5 0 2 4 6 8 10 12 14 Time Analog: Faithful reproduction of signal at RX Digital: Decide which symbol was sent from a pre-defined alphabet Slide 51

Response (db) Response Response Bandwidth of a Signal 1 0.8 1 0.5 0 0.6 0 1 2 3 4 5 6 7 8 0.4 0.2 0-4 -3-2 -1 0 1 2 3 4 Time (t/tb) Bandwidth of pulse of duration Tb is infinite Spectrum has sinc(x) shape extending from - to + First sidelobe -13 db down, rolls off at 20 db/dec Some form of filtering is required 0-10 -20-30 -40-50 0 1 2 3 4 5 6 7 8 Normalised Frequency (f.tb) Slide 52

Response (db) Impulse Response Nyquist Brickwall Filter 5 0-5 -10-15 -20-25 -30-35 -40-45 Pulse Response Nyquist Brickwall Filter -50 0 0.5 1 1.5 2 2.5 3 3.5 4 Normalised Frequency (f/rb) 1 0.8 0.6 0.4 0.2 0-0.2-0.4-6 -4-2 0 2 4 6 Normalised Time (t/tb) Nyquist filter - achieves zero crossings at integer multiples of symbol period e.g. brickwall filter with cut-off at RS/2 Zero crossings at symbol interval - no ISI at sample point Slide 53

The Nyquist Bandwidth In a radio transmitter the filtering is done at baseband. Ideal brick-wall filter at the minimum bandwidth Envelope of digital baseband spectrum. f n = Nyquist Frequency frequency = Symbol Rate/2 This condition gives zero ISI (Inter Symbol Interference) Slide 54

Filter Bandwidth Parameter " Practical Filter Shapes 1 0.8 0.6 0.4 0.2 = 1.0 = 0.5 = 0.3 = 0 brickwall 0 0 0.2 0.4 0.6 0.8 1 Fs : Symbol Rate Alpha describes the "sharpness" of the filter. Occupied bandwidth is approximately: Symbol rate X (1 + ) Slide 55

Single Carrier Modulation Frequency Domain View 1 carrier BW = SymRate(1+ ) Slide 56

Polar vs. "I-Q" Format Project signal to "I" and "Q" axes Polar to Rectangular Conversion "Q" Q-Value 0 deg "I" I-Value Slide 57

Creating Digital Modulation (1,0) Q +1volt (0,0) We have used the concept of Signal Space to view our modulations. We can use the same idea to engineer these modulations. 1volt +1volt I (1,1) 1volt (0,1) +1v I (0,1) f C 90º 1v Q Slide 58

How Does this Work? Putting two different messages into one signal space. (These could be independent messages.) I(t) cos( t) 90º S(t) Q(t) S(t) = I(t)cos( t) Q(t)sin( t) = A(t)[cos ( t + (t))] A I 2 Q 2 tan 1 Q I Slide 59

Separating the Components: The Receiver S(t) = I(t)cos( t) Q(t)sin( t) The composite signal is separated by multiplying (mixing) by sin( t) and cos ( t), the resulting sin 2 and cos 2 terms become [I(t) or Q(t)] ½[1 ± cos(2 t)] terms - the 2 t s are removed by LPF. LPF I(t) S(t) cos( t) +90º LPF Q(t) Slide 60

Agenda Introduction Overview Theory of Operation Specifications Modern spectrum analyzer designs & capabilities Basics on digital modulation Measurements on digital modulation Slide 61

Measurements on digital radios Swept Spectrum Analyzer (with span zero and enough ResBW) Time Domain (CCDF, pulse shaping, timing) Frequency Domain (Channel Power, spectrum mask, ) Modulation Domain Overall Modulation Quality, Modulation Quality on individual carriers Channel Response, Group Delay In Channel Spurious Search Slide 62 Vector Signal Analyzer

FFT Analyzer Block Diagram Digital Data Flow (time) Digital Filter assembly Input Signal ADC Assembly ADC 90 o phase shift Real Part (I) Imaginary Part Re-Sampling DSP D e m o d W i n d o w F F T (freq) D I S P L A Y F s (demod time) (Q) Quadrature Anti-alaising and Sampling Decimation Slide 63

Measurements in the frequency domain: - Channel power and occupied bandwidth - Adjacent channel power ratio (ACPR) Slide 64

Measurements in the frequency domain: Spurious signals In band Out of band Slide 65

- time domain Slide 66

Power Amplifier (PA) Compression How to verify? Useful measurements: ACP, CCDF With compression Without compression Compare these measurements performed: - at the input and output of the PA - at the output for decreasing values of the input level Slide 67

Modulation Quality Analysis EVM[%] MER[dB] Modulation Error Basic Concept: Q Ideal point Ideal Signal at decision time Error Vector Magnitude Error Vector Measured point Measured Signal at decision time Measured signal is never equal to ideal signal, due to noise, transmitter impairments, propagation phenomena, I Slide 68

Effect of Noise Noise adds vectorially to a signal. Noise on a QPSK constellation. Slide 69

Overlapping Probabilities There is a finite probability that adjacent states could be confused. A measure of the functioning of the system is BER (Bit Error Ratio) e.g. if 100 bits are in error in 10 8 bits. Then the BER is 10-6 Slide 70

I/Q Impairments: I/Q gain imbalance, Quadrature errors, I/Q offsets I/Q impairments are typically cause by matching problems due to component differences between the I side and Q side of the block diagram Significant measurements: constellation and EVM metrics Slide 71

Signal B: Demodulation Spectrum B Demod Points are not randomly distributed EVM High (3.6%) Mag & Phase Errors High and comparable Slide 72

Signal B: EVM Spectrum Shows Spur EVM Spectrum Spurious Signal -36dBc spur was buried under the modulated carrier Spectrum B Slide 73

Clock impairments Incorrect symbol rate The effect of symbol rate errors on the different measurements depends on the the magnitude of the errors: - different methods to verify small or large symbol errors Slide 74

Incorrect Symbol Rate + Symbol Clock Transmitter and Receiver operate with different clocks Recovered Symbol Clock Symbol Rate Error = 0.1% Slide 75

Incorrect symbol rate: small errors (2) Measurement: EVM vs time Detection and troubleshooting hint: Verify the V shape of the magnitude of the error vector versus time display Slide 76

Troubleshooting examples: QPSK transmitter with symbol rate errors Slide 77

End Roberto Sacchi Application Engineer roberto_sacchi@agilent.com Slide 78