數位示波器的應用 抖動(jitter)測量 · pdf file2 y抖動的定義...
TRANSCRIPT
1
(Jitter)
2
Agenda
3
Jitter, what is it?
What is jitter?
the deviation of an edge from where it should bethe deviation of an edge from where it should be
Jitter is caused by:(among other things) Thermal noise Injected noise (EMI/RFI) Circuit instabilities Transmission Losses
4
()
()
5
Things you should already know:
Period Jitter
Cycle-to-Cycle Jitter
Time Interval Error (TIE)
Clock Jitter
Data Jitter Clock recovery
BER
Three Methods of Jitter Measurement
6
Three Methods of Jitter Measurement
1ns
0.8ns 0.8ns 0.8ns 1.2ns 1.2ns 1.2ns 1.2ns 0.8ns 0.8ns
-0.2ns -0.4ns -0.6ns -0.4ns -0.2ns 0ns 0.2ns 0ns -0.2ns
0.8ns
1.2ns
1.0ns
+0.4ns0ns
-0.4ns
-0.2ns
-0.6ns
0ns 0.2ns
-0.4ns -0.4ns
-0.2ns
0ns 0ns
0.8ns
-0.4ns
1.2ns
-0.2ns
1.2ns
0ns
0.8ns
1.2ns
+0.4ns
-0.2ns
-0.4ns
-0.2ns
0ns
()
()
Period Jitter
Time Interval Error (TIE)Jitter
Time Trend forPeriod Jitter
Time Trend forCycle-Cycle Jitter
P(n+1)-P(n)
Time Trend forTIE Jitter()
7
TjTotal Jitter in a signal
Estimated jitter for a large population: 1012 bits
8
Random Jitter (RJ)
()
Peak-to-Peak(unbounded)!
1-sigma or RMS 7-sigma
9
Deterministic Jitter (DJ)
Peak-to-Peak(bounded)!
Peak-to-Peak
10
Periodic Jitter (PJ)
RF
Peak-to-Peak
11
Duty Cycle Distortion (DCD)
rise-timefall-timeDuty Cycle
0.0v
Peak-to-Peak
12
Data-Dependent Jitter (DDJ or ISI)High/Low(Pattern)
DDJ
13
BER
Bit Error Rate Ratio of good bits to bad bits
Send 1,000,000,000,000,000,000,000 bits
Receive 999,999,999,999,999,999,999 bits correctly
Yields a BER of 1:10-21 (1x10-21 or 1E-21)
Requires sending as many bits as you are testing for!
You must capture all bits in sequence for this test to be valid, only a BERT can do this!
14
Unit Interval and Bit Error Rate
UI: the nominal period of one transmitted bit
Serial Data Source
Transmitter
Connector Cable Connector Amplifier
Channel
Clock DataRecovery
(CDR)
Receiver
END END
BERT System
BER: Ratio of good bits to bad bits Generally specified at 1012 bits
15
(01)
1-sigma or RMS 7-sigma
Z
Z = 50.999999 BER=1:10-6
Z = 60.999999999 BER=1:10-9
Z = 70.999999999999 BER=1:10-12
16
Total Jitter @ BER
Random components
Deterministic components
Peak-to-Peak
1-sigma or RMS
pdf : Tj = Dj Rj (convolution)
Pk-Pk: Tj = (N*Rj) + Dj , where N =2*Z
17
Statistics
Theorem: Sum of Two Independent RVs(Random Variable)
If two RVs are independent, the pdf of the sum is the convolution of the respective pdfs
Central Limit Theorem
If N RVs are independent, the pdf of the sum is Gaussian as N infinity
=
18
Histograms vs. Eye Diagrams
How open is the eye, anyway?
(depends how long you watch)
Eye Opening=?
19
Eye Opening vs. Bit Error Rate
Cumulative Distribution Function (cdf)
Probability Density Function (pdf)
Eye Open for BER 1:10-12
20
BER Estimationusing Eye Diagram Statistics
21
Rj/Dj @ Bath Tub Curve
Start with TIE PLL TIE
Perform FFT Determine
frequency and pattern rate
Sum pattern related bins
Sum unrelated periodic bins
Measure RMS of remaining bins
Estimate BER
22
TDSJIT3 Factory Correlation
4 BERT Correlation0 A tool is useless if it doesnt agree with accepted standards0 TDSJIT3 correlates to a BER Tester within 2% to 3%
Results for PRBS-7 Data Pattern
Raw Noise (10ps rms) Sinusoidal (100KHz) Sinusoidal (50MHz) JIT3 BERT Diff JIT3 BERT Diff JIT3 BERT Diff JIT3 BERT Diff
Freq 1.0625G 1.0625G 0.0 1.0625G 1.0625G 0.0 1.0625G 1.0625G 0.0 1.0625G 1.0625G 0.0 PRBS-7 BER eye open eye open eye open eye open
(127 bits) 1.00E-02 97.4 97.1 0.3 95.2 95.2 0.0 95.0 95.4 -0.4 95.1 95.4 -0.3 1.00E-03 96.6 96.2 0.4 93.2 93.2 0.0 93.5 93.8 -0.3 93.7 93.7 0.0 1.00E-04 96.0 95.6 0.4 91.8 91.6 0.2 92.3 92.1 0.2 92.5 92.4 0.1 1.00E-05 95.5 95.4 0.1 90.5 90.1 0.4 91.4 91.1 0.3 91.6 91.3 0.3 1.00E-06 95.1 95.2 -0.1 89.4 88.8 0.6 90.6 90.1 0.5 90.8 90.7 0.1 1.00E-07 94.7 95.1 -0.4 88.3 88.3 0.0 89.8 89.4 0.4 90.2 90.0 0.2 1.00E-12 93.2 94.2 -1.0 84.2 82.1 2.1 87.0 85.5 1.5 87.5 85.9 1.6 RJ 3.34ps Noise/Com 22db 7.79ps Noise/Com 12db 5.93ps Noise/Com 17db 5.78ps Noise/Com 17db DJ 24.34ps Sine 27.88ps Sine 56.78ps Sine 53.86ps Sine PJ 5.23ps Freq 100K 3.96ps Freq 100K 37.90ps Freq 100K 33.31ps Freq 50M DCD 4.60ps Mag 1mv 6.10ps Mag 1mv 4.08ps Mag 150mv 4.63ps Mag 150mv DDJ 14.52ps 17.82ps 14.80ps 15.92ps difference in %UI exterpolated
23
(setup and hold)
(PLL)
(data)
?
-(Cycle-Cycle)
24
Vcc Noise
Ground Bounce
PLL
Dead Band
(Crystal)
(Crosstalk)
25
(Histogram)
.
26
.
27
28
29
8
30
JavaJavaTM
31
T X X X X X X X X
T X X X X X X X X
P1 P2 Period = X1-T,X2-X1,X3-X2,,Xn-Xn-1
32
TDS7000/6000/5000TDSJit3
425
:
(Histogram)
(Time Trend)
(Cycle Trend)
(Spectrum)
(BathTub)
33
TDS7/6/5000,DPO/DSATDSJit3
: (10) (4) - (3) (8) 425
6
34
: (Mean) (Max) (Min) (Pk-Pk) (Std Dev) (Population) (Max+) (Max-)
6
35
(SSC)
32MB
36
(Span)
37
38
39
40
BER Eye Opening
BER Eye Opening
BER Eye Opening 530.91m Ui
41
:
(TDS7000B, 1.2ps RMS)
()
(5.9343ps)
(Delta Time Accuracy)
42
(Delta Time Accuracy)
-
DTA(interpolation)
43
4 TDS740420GS/s(SR)
1.5ppm
400MHz (2.5ns )
DTA = 0.03 / SR + (DTA = 0.03 / SR + () X () X ()) DTA = 0.03 / 20GS + 1.5ppm x 2.5ns DTA = 1.5ps + 0.00375ps
0.00375ps
DTA
44
TDSJIT3 Software Overview
45
TDSJIT3 User Interface
Measurement Selection Clock Data Clk-Data General
Users can select measurements from any source
Channel, Math, Ref
User can select the same measurement on multiple channels:
Period(Ch1), Period(Ch2), Period(Ch3), Period(Math1)
46
TDSJIT3 Configuration
Configuration Meas Params Source Autoset Gate/Qualify Ref Levels Population Limit
Users configure each Source and Measurement independently Each source has its own settings Each measurement has its own
settings
47
TDSJIT3 Autosetting
Vertical Autoset Intended to maximize vertical
dynamic range Each source can be set
independently
Horizontal Autoset Adjusts horizontal settings to
ensure reliable results based on signal risetime
Vert and Horiz Autoset are really fine tuning controls, I normally recommend customers set scope manually for best results
48
TDSJIT3 Gates and Qualifiers
Gate Measurements are made only
inside gated region Gate by Cursor Gate by Zoom Limits single shot population
hence Rj/Dj capability
Qualify Measurements are made only
when Qualify input has transitions and state meets criteria
49
TDSJIT3 Reference Levels
Sets thresholds for making measurements Each source has its own
settings for rising and falling edges
Autoset-Setup allows changing levels by percentage (e.g., 10-90%)
Autoset All Active is a quick way to go and ensures good results 99% of the time
Hysteresis is an important setting, especially in noisy environments
50
TDSJIT3 Population Limit
Sets population limit for making measurements Measurement stops when limit
reached In Run mode (continuous),
acquisitions stop when population limit reach
A handy and customer desired feature
51
TDSJIT3 Summaries
Shows a summary of measurements, their configurations and reference levels Handy place to review how the
software is configured
52
TDSJIT3 Results Views
Results are displayed differently in TDSJIT3 Results are shown in different
views: All Statistics Min/Max Mean/StdDev TIE Rj/Dj BER
53
TDSJIT3 All Statistics
Each measurement is selected independently only results for one measurement are shown All statistics are shown for the
selected measurement The first column is the last
acquisition The second column is the
accumulation of all acquisitions Clear erases all data
Max +/- show the largest pos/negchange on a cycle to cycle basis for every measurement type!
54
TDSJIT3 TIE: Rj/Dj - BER
Displays results for Rj/Dj Rj/Dj results are only available
for TIE and PLL TIE measurements
Rj/Dj is a two step process: make the TIE measurement; then calcula