1 Important Characteristics of Digital Oscilloscopes Important Characteristics of Digital Oscilloscopes and RADAR Pulse Measurements with Digital Oscilloscopes

1

Important Characteristics of Important Characteristics of Digital OscilloscopesDigital Oscilloscopes

andand

RADAR Pulse Measurements RADAR Pulse Measurements with Digital Oscilloscopeswith Digital Oscilloscopes

Important Characteristics of Important Characteristics of Digital OscilloscopesDigital Oscilloscopes

andand

RADAR Pulse Measurements RADAR Pulse Measurements with Digital Oscilloscopeswith Digital Oscilloscopes

5:30 – 6:00 Pizza and Refreshments6:00 – 7:00 Technical Presentation

This is a FREE event. Non-Members Welcome!

and Southeastern Michigan present…

Vince Woerdeman, Agilent Technologies

Marty Gubow, Agilent Technologies

http://www.ieee.org/portal/site/iportals

Click to edit Master subtitle style

Page 2

AgendaAgenda

Evaluating a Scope’s Performance Characteristics What Bandwidth is needed? What Sample Rate is needed? How does Nyquist’s Theorem and

aliasing apply to oscilloscopes? Acquisition Errors and Interleave

Distortion What are other important

characteristics?


Page 3

Evaluating Performance CharacteristicsEvaluating Performance Characteristics

Is Full Scope Functionality Retained?

Required Number of Channels?

Required Bandwidth/Acquisition Performance?

Waveform Update Rate, Decode Update Rate, Probing, Ease-of-use, Display Quality, Triggering, etc.?


Page 4

“Rule-of-thumb” Bandwidth Suggestion“Rule-of-thumb” Bandwidth Suggestion

Suggested Bandwidth = 5X Highest Clock Rate

Allows capture of the 5th harmonic with minimum attenuation.

Scope Bandwidth


Page 5

Accurate Bandwidth DeterminationAccurate Bandwidth Determination

Step #2: Determine highest signal frequency content (fKnee).fKnee = 0.5/RT (10% - 90%)fKnee = 0.4/RT (20% - 80%)

Step #3: Determine degree of required measurement accuracy.

Required Accuracy

Gaussian Response

Maximally-flat Response

20% BW = 1.0 X fKnee BW = 1.0 X fKnee



Step #4: Calculate required bandwidth.

Step #1: Determine fastest rise/fall times of device-under-test.

Source: Dr. Howard W. Johnson, “High-speed Digital Design – A Handbook of Black Magic”


Page 6

System Bandwidth CalculationSystem Bandwidth Calculation

fKnee = (0.5/500ps) = 1GHz

3% Accuracy: Scope Bandwidth = 1.9 x 1GHz = 1.9GHz


Example

Determine the minimum required bandwidth of an oscilloscope with an approximate Gaussian frequency response to measure a 500ps rise-time (10-90%):



Page 7

Analog Bandwidth ComparisonsAnalog Bandwidth Comparisons

What does a 100 MHz clock signal really look like?

100MHz Scope

Rise Time = 2.5ns500MHzScope

Rise Time = 750ps

1GHzScope

Rise Time = 550ps2GHzScope

Rise Time = 495ps


Page 8

How Much Sample Rate is Required?How Much Sample Rate is Required?

The truth lies somewhere in between!

Engineer Fred has total trust in Dr. Nyquist and says:

“2X over the scope’s bandwidth.”

Engineer Betty doesn’t trust Dr. Nyquist and says:

“10X to 20X over the scope’s bandwidth.”


Page 9

Nyquist’s Sampling TheoremNyquist’s Sampling Theorem

Nyquist’s sampling theorem states that for a limited bandwidth (band-limited) signal with maximum frequency fmax, the equally spaced sampling frequency fs must be greater than twice of the maximum frequency fmax, i.e.,

fs > 2·fmax

in order to have the signal be uniquely reconstructed without aliasing.

The frequency 2·fmax is called the Nyquist sampling rate (fS). Half of this value, fmax, is sometimes called the Nyquist frequency (fN).

Dr. Harry Nyquist


Page 10

Nyquist’s Basic Rules…Nyquist’s Basic Rules…

1.fMAX < fS/2

The highest frequency sampled MUST be less than fS/2…

This is NOT the same as oscilloscope bandwidth.

2.Samples MUST be equally spaced

The forgotten rule!

But not-so-simple for DSO technologyBut not-so-simple for DSO technology


Page 11

Ideal Brick-wall Response w/ BW @ Nyquist (fN)Ideal Brick-wall Response w/ BW @ Nyquist (fN)

fSfN

-3dB

Frequency

0dB

Att

en

uat

ion


Page 12

Gaussian Response w/ BW @ fS/2 (fN)Gaussian Response w/ BW @ fS/2 (fN)

Frequency

-3dB

Att

en

uat

ion

0dB

fN fS

Aliased Frequency Components


Page 13

500MHz scope sampling @ 1GSa/s (BW = fS/2 = fN)500MHz scope sampling @ 1GSa/s (BW = fS/2 = fN)


Page 14

Frequency

-3dB

Att

en

uat

ion

0dB

fN fS fS/4

Gaussian Response w/ BW @ fS/4 (fN/2)Gaussian Response w/ BW @ fS/4 (fN/2)

fS/2.5

Maximally-Flat Response w/ BW @ fS/2.5 (fN/1.25)Maximally-Flat Response w/ BW @ fS/2.5 (fN/1.25)Gaussian Response w/ BW @ fS/2 (fN)Gaussian Response w/ BW @ fS/2 (fN)




Page 15

500-MHz scope (2 GSa/s vs. 4 GSa/s)500-MHz scope (2 GSa/s vs. 4 GSa/s)

2 GSa/s (fBW = fS/4 = fN/2)

4 GSa/s (fBW = fS/8 = fN/4)

Input = 100 MHz clock with 1 ns edge speeds


Page 16

6-GHz scope (20 GSa/s vs. 40 GSa/s)6-GHz scope (20 GSa/s vs. 40 GSa/s)

20 GSa/s (fBW = fS/3.3)

40 GSa/s (fBW = fS/6.6)

Input = 1.25 GHz clock with 100 ps edge speeds


Page 17

Complying with Nyquist’s Rule #1 (fS > 2 x fMAX)Complying with Nyquist’s Rule #1 (fS > 2 x fMAX)

2X sampling violates Rule #1

2.5X to 5X sampling sufficiently satisfies Rule #1

> 5X sampling provides further compliance with Rule #1… IF additional error sources are not introduced that violate Rule #2

Engineers often overlook Rule #2…

“Samples MUST be evenly spaced”


Page 18

Real-time Non-interleaved ADC SystemReal-time Non-interleaved ADC System

ADC #1 ACQMEM

SampleClock

To CPUInput

AnalogAmplifier


Page 19

Sample Rate > 4 x fBW (Non-interleaved)Sample Rate > 4 x fBW (Non-interleaved)

Sample Clock

= Input Signal

= Sample Clock

= Sin(x)/x Interpolated Waveform

= Real-time Digitized Point

Sin(x)/x Interpolated WaveformInput Signal


Page 20

Real-time Interleaved ADC SystemReal-time Interleaved ADC System

ADC #1 ACQMEM

SampleClock

To CPU

AnalogAmplifier

ADC #2 ACQMEM

½ ClockDelay

Input

To CPU

Input

Accurate ADC interleaving requires:

1. Matched vertical response of each ADC

2. Precise phased-delayed clocking


Page 21

SR > 8 x fBW (Perfectly Interleaved)SR > 8 x fBW (Perfectly Interleaved)

Clock #1

Clock #2

= Input Signal

= Sample Clock





Page 22

SR > 8 x fBW (Poorly Interleaved)SR > 8 x fBW (Poorly Interleaved)

Clock #1

Clock #2

= Input Signal

= Sample Clock





Page 23

Testing for Interleave DistortionTesting for Interleave Distortion

1. Effective bits analysis using sine waves

2. Visual sine wave test

3. Spectrum analysis

4. Measurement stability/repeatability

Interleave distortion violates Nyquist’s Rule #2:

“Samples must be evenly spaced”


Page 24

1-GHz Sine Wave on 1-GHz BW Scopes1-GHz Sine Wave on 1-GHz BW Scopes4 GSa/s (non-interleaved)4 GSa/s (non-interleaved)

20 GSa/s (interleaved)20 GSa/s (interleaved)

4 GSa/s produces superior results compared to 20 GSa/s

Interleave Distortion


Page 25

2.5-GHz Sine Wave on a 3-GHz Scope2.5-GHz Sine Wave on a 3-GHz Scope20 GSa/s (Single-chip ADC)20 GSa/s (Single-chip ADC)

40 GSa/s (Dual-interleaved ADC chip-set)40 GSa/s (Dual-interleaved ADC chip-set)

Precision ADC interleaving technology produces improved measurements

Vp-p (σ) = 2.4 mV

Vp-p (σ) = 1.8 mV


Page 26

Interleave Sampling Distortion

2.5-GHz Sine Wave on a 2.5-GHz Scope2.5-GHz Sine Wave on a 2.5-GHz Scope10 GSa/s (Single-chip ADC)10 GSa/s (Single-chip ADC)

Poor ADC interleaving technology produces degraded measurements

40 GSa/s (Quad-interleaved ADC chip-set)40 GSa/s (Quad-interleaved ADC chip-set)

Vp-p (σ) = 9.1 mV

Vp-p (σ) = 12.0 mV


Page 27

FFT Analysis of 2.5-GHz Sine Wave at 40 GSa/sFFT Analysis of 2.5-GHz Sine Wave at 40 GSa/s

3-GHz Scope3-GHz Scope

2.5-GHz Scope2.5-GHz Scope

10-GSa/s Distortion (-32 dB)40-GSa/s Distortion


Page 28

400-MHz Clock Sampled @ 40 GSa/s400-MHz Clock Sampled @ 40 GSa/s



Rise Time (avg.) = 254psRise Time (range) = 60psRise Time (σ) = 10ps

Rise Time (avg.) = 250psRise Time (range) = 35psRise Time (σ) = 3.3ps


Page 29

FFT Analysis of 400-MHz Clock at 40 GSa/sFFT Analysis of 400-MHz Clock at 40 GSa/s



40-GSa/s Distortion10-GSa/s Distortion

(27 dB below 5th harmonic)


Page 30

Other Oscilloscope Characteristics to ConsiderOther Oscilloscope Characteristics to Consider

Waveform Update Rate

Advance Analysis

Display Quality

Ease-of-use

Probing

Price


Page 31

InfiniiMax Active Probe ExtensionInfiniiMax Active Probe Extension

Allows for environmental chamber testing up 105 degrees C.


Page 32

Questions and AnswersQuestions and Answers

33

Oscilloscope Radar

Measurement Basics


Page 34

AgendaAgenda

• Introduction

• Pulsed Power and Power Spectrum

Measurements

• Noise Measurements

• Component Measurements

• Evaluating I/Q Demodulator Errors

• Pulsed Component Measurements

• Time Domain Measurements

• Jitter Measurements

Radar Measurement Basics


Page 35

Introduction



Page 36

Some Typical Radar ApplicationsSome Typical Radar Applications

• Surveillance

• Search and track

• Fire control

• Navigation

• Missile guidance

• Proximity fuses

• Altimeter

• Terrain avoidance

• Weather mapping

• Space



Page 37

The Wide Range of Measurement RequirementsThe Wide Range of Measurement Requirements

Parameter Typical Range

• Frequency………………………………..100MHz - 95GHz

• Pulse Width (PW)……………………….10nsec to Infinite (CW)

• Pulse Repetition Frequency …………30Hz to 300KHz

• Rise Time………………………………...1nsec - 100nsec

• Duty Cycle……………………………….0.01% - 100%

• Peak Power………………………….…..1W - 50MW

• Pulse Compression…………………….FM, Phase Coded

• Frequency Agility……………………….100MHz - 2GHz (BW)



Page 38

Simplified Pulse Doppler Radar Block DiagramSimplified Pulse Doppler Radar Block Diagram

PRFGENERATOR

DISPLAY

ADC S/H LPFVIDEO

AMP

COHO LIMITER LPF

ADC S/H LPF VIDEOAMP

0

SPLITTER

o

2ndIFA

IFBPF

2nd

L.O.

1stIFA

IFBPF

LNA

STALO

COHO BPF AMPRFBPF

Doppler

and

Range

FFT

Processor

PREDRIVERAMP

PULSEDPOWER

TRANSMITTER

DUPLEXER

Transmitter/Exciter

Receiver/Signal Processor

Antenna

90o

PULSEMODULATOR

RECEIVERPROTECTION

FREQUENCYAGILE L.O.



Page 39

Active Electronically Steered AntennaActive Electronically Steered Antenna

Transceiver

Wave Front


Animation


Page 40


Page 41


Page 42

AgendaAgenda

• Introduction


Measurements









Page 43

Pulsed Power and Power Spectrum Measurements



Page 44

Why Measure Power?Why Measure Power?

• High peak power influences the expense of the system

R

$ $

Modulator,PFN, etc.

OutputStage

• Power determines the absolute range

RPt

4



Page 45

Instruments Used to Measure PowerInstruments Used to Measure Power

• Vector Signal Analyzer

• Spectrum Analyzer

• Power Meter



Page 46

AgendaAgenda

• Introduction


Measurements









Page 47

Noise FigureNoise Figure

-the degradation in the signal-to-noise ratio as the signal passes through the network

Noise Figure, F =

SN in

SN out

(S/N)in

(S/N)out

T = 290°K


G


Page 48

N8975A Noise Figure AnalyzerN8975A Noise Figure Analyzer

• Wide frequency range (1.5GHz/3GHz/26.5GHz)

• Graphical data display

• Ease of use

• Variable IF bandwidths

• Intuitive user interface

• Smart Noise Source (cal files stored in EEPROM and internal temperature sensor)



Page 49

AgendaAgenda

• Introduction


Measurements









Page 50

Component Test



Page 51

Why make Network Analyzer measurements on a RadarWhy make Network Analyzer measurements on a Radar

• Verify specifications of “building blocks” for more complex RF systems

• Ensure distortionless transmission of communications signals

• Linear: constant amplitude/linear phase / constant group delay

• Non-linear: harmonics, intermodulation, compression, AM-to-PM conversion

• Ensure a good match when absorbing power (e.g. an antenna)



Page 52

The Need for Both Magnitude and PhaseThe Need for Both Magnitude and Phase

4. Time-domain characterization

Mag

Time

5. Vector-error correction

Error

MeasuredActual

2. Complex impedance needed to design matching circuits

3. Complex values

needed for device modeling

1. Complete characterization of linear networks

High-frequency transistor model

Collector

Base

Emitter

S21

S12

S11 S22



Page 53

PNA Performance Network Analyzer FamilyPNA Performance Network Analyzer Family

• Up to 35 s/point measurement speed

• 143 dB dynamic range with direct receiver access

• 128 dB dynamic range at test ports

• 0.005 dB trace noise (10 kHz IF bandwidth

• 3, 6, 9, 20, 40, and 50 GHz microwave models

• 4 mixer-based receivers enable TRL/LRM calibration



Page 54

AgendaAgenda

• Introduction


Measurements









Page 55

Evaluating I/QDemodulator Errors



Page 56

•Magnitude is an absolute value•Phase is relative to a reference signal

Phase

Ma

g

0 deg

Polar Display -- Magnitude and Phase Represented TogetherPolar Display -- Magnitude and Phase Represented Together


Animation


Page 57

Magnitude Change Phase Change

Frequency ChangeBoth Change

0 deg

0 deg

Phase

Mag

0 deg 0 degPhase

Signal Changes or ModificationsSignal Changes or Modifications



Page 58

89640 Vector Signal Analyzer89640 Vector Signal Analyzer

• Tuners covering dc to 6.0 GHz Frequency Range

• 36-78 MHz bandwidth for broadband signal formats.

• >200MHz bandwidth with 54832B Infiniium scope

• I/Q display formats

• Analog AM/FM/PM demodulation

• Time Gated measurements

• 1.2Gbytes of capture memory

• Tight integration with ADS (PC based design applications).



Page 59

PSG Performance Signal Generator FamilyPSG Performance Signal Generator Family

• 250 kHz to 20 or 40 GHz Frequency Range in Coax

• Extension to 110 GHz with the 83550 Series Multipliers

• High power

• Excellent phase noise

• AM/FM/PM and pulse modulation capabilities



Page 60

AgendaAgenda


• Introduction

• Power Measurements


• Component Test






Page 61

Pulsed ComponentMeasurements

AgendaAgenda



Page 62

Pulsed Transfer Functions in the Time DomainPulsed Transfer Functions in the Time Domain

Amplifier



Page 63

AgendaAgenda


• Introduction



• Component Test






Page 64

Time Domain Measurements



Page 65

Why Measure Pulse Parameters?Why Measure Pulse Parameters?

PW determines resolving ability (small is better) -- BW ~ 1 / PW

PW affects average power (absolute range, large is better)

Unintentional AM and fast risetimes can reduce the life expectancy

PRI determines unambiguous range

of transmitter

R=P G G

P (4 )3

r

t t r

2

4

• PW determines resolving ability (small is better) - BW ~ 1/PW

• PW affects average power (absolute range, large is better)

• Unintentional AM and fast rise times can reduce the life expectancy of transmitter

• PRI determines unambiguous range

PW determines resolving ability (small is better) -- BW ~ 1 / PW

PW affects average power (absolute range, large is better)

Unintentional AM and fast risetimes can reduce the life expectancy

PRI determines unambiguous range

of transmitter

R=P G G

P (4 )3

r

t t r

2

4



Page 66

What are Important Pulse Parameters?What are Important Pulse Parameters?

Pulse RepetitionInterval (PRI),

PulseWidth

Risetime Falltime

Also: Duty Cycle

Pulse Shape (over and preshoot, droop)

Pulse Width Stability

PRI Stability

PulseOff Time

1

PRF

• Duty cycle

• Pulse shape (over and pre-shoot, droop)

• Pulse width stability

• PRI stability

Also:



Page 67

Definition of Pulse WidthDefinition of Pulse Width

Average Power During On-time of Pulse

-6dB

Pulse Width

Pulse Repetition Interval



Page 68

Time Domain MeasurementsTime Domain Measurements

Envelop parameters

• Rise time

• Fall time

• Pulse width

• Period

• On/off ratio

Modulation in the pulse

• Unintentional

- AM to PM

- Phase noise

• Intentional

- Chirp

- Barker coding

- Frequency agility



Page 69

Measuring with a Digital OscilloscopeMeasuring with a Digital Oscilloscope

Advantages

• General measurement tool

• Wide bandwidth

• Easy to understand

• Option to post process signal

Considerations

• Aliasing

• Dynamic range

• Flatness

• Memory depth



Page 70

Time Domain MeasurementsTime Domain Measurements

PRFGENERATOR

DISPLAY

ADC S/H LPFVIDEO

AMP

COHO LIMITER LPF

ADC S/H LPF VIDEOAMP

0

SPLITTER

o

2ndIFA

IFBPF

2nd

L.O.

1stIFA

IFBPF

LNA

STALO

COHO BPF AMPRFBPF

Doppler

and

Range

FFT

Processor

PREDRIVERAMP

PULSEDPOWER

TRANSMITTER

DUPLEXER

Transmitter/Exciter

Receiver/Signal Processor

Antenna

90o

PULSEMODULATOR

RECEIVERPROTECTION

FREQUENCYAGILE L.O.



Page 71

Infiniium OscilloscopeInfiniium Oscilloscope

• 4 channels

• Up to 64 MB deep memory

• Up to 40 GSa/s sample rate/channel

• Infiniium award-winning usability

• Full upgradeability



Page 72

AgendaAgenda


• Introduction



• Component Test






Page 73

Jitter Measurements



Page 74

What is Jitter?What is Jitter?

Threshold

Threshold

Threshold

Jitter

Noise

-creates ambiguity in threshold crossing

(a)

(b)

(c)



Page 75

Jitter FunctionJitter Function

t1 t2t3

t4 t5

Ideal PulseTrain

Jitter signal viewed atinstants in time

Jitter magnitudeJitter function



Page 76

What is Jitter? What is Jitter?

Ideal clock:

Jittered clock:

)2sin( tfc

)2sin(2sin 101

34 tftf cc

)2sin(101

34 tfcJitter:

UI32

• Jitter is the deviation of a timing event of a signal from its ideal position.

• This is the traditional description of jitter,

• commonly referred to Time Interval Error (TIE), or phase jitter.


Page 77

Radar Jitter MeasurementRadar Jitter Measurement

PULSEDPULSEDRADARRADAR

JitterSource

Pulse Envelope

Trigger

Ch1

DigitizingOscilloscope

PRI Reference

Crystal Detector



Page 78

Time Interval vs. Time ProfileTime Interval vs. Time Profile

Time Interval

Time

Jitter Periodic Rate

17.9 KHz

3.8 ns

Peak-to-Peak Jitter Amplitude



Page 79

Histogram of Clock PeriodHistogram of Clock Period

Peak-to-Peak

% Probability

Period

Probability Analysis

Jitter Distribution

- +MEANMIN MAX



Page 80

Histogram of Edge JitterHistogram of Edge Jitter


Page 81

“Real World” Jitter is Complex

Random Jitter (RJ) is unbounded• Due to thermal noise, shot

noise, etc.• Follows Gaussian distribution• Requires statistical analysis

to be quantified• RJpp = 14.1 x Jrms for 10-12

BER

Deterministic Jitter (DJ) is bounded and composed of:• Duty-Cycle-Distortion (DCD)• Inter Symbol Interference

(ISI)• Periodic Jitter (PJ) RJ

DJ

Jitter is composed of random and deterministic

components


Page 82

Jitter Probability: BERJitter Probability: BER

randomticdeterminispkpk nJJ

=


Page 83

How Do Real Time Scopes Measure Jitter on Data? How Do Real Time Scopes Measure Jitter on Data?

Jitter Trend

NRZ Serial Data

Recovered Clock

Jitter Spectrum

Units in Time

Units in Time

Jitter Histogram


Page 84

Agilent EZJIT Jitter Measurement ApplicationAgilent EZJIT Jitter Measurement Application

Signal

Trend

Histogram

Spectrum


Page 85

Total Jitter ComponentsTotal Jitter Components

• TJ: Total Jitter (Convolution of RJ & DJ)

• RJ: Random Jitter (rms)

• DJ: Deterministic Jitter (p-p)

PJ: Correlated & uncorrelated Periodic Jitter due to cross-talk and EMI

DCD: Duty Cycle Distortion due to threshold offsets and slew rate mismatches

ISI: Inter-Symbol Interference due to BW limitation and reflections

TJ

RJDJ

PJ DCD ISI

Unbounded (RMS)Bounded (p-p)


Page 86

Where Does Jitter Come From?Where Does Jitter Come From?

Transmitter Receiver

•Thermal Noise (RJ)•Voltage Offsets (DCD)•Power Supply Noise (RJ, PJ)•On chip coupling (PJ)

•Lossy interconnect (ISI)•Impedance mismatches (ISI)•Crosstalk (PJ)

•Termination Errors (ISI)•Thermal Noise (RJ)•Incorrect Threshold (DCD)•Power Supply Noise (RJ, PJ)•On chip coupling (PJ)

Media


Page 87

Into this…Into this…


Page 88

Questions and AnswersQuestions and Answers

Documents

1 Important Characteristics of Digital Oscilloscopes Important Characteristics of Digital Oscilloscopes and RADAR Pulse Measurements with Digital Oscilloscopes