HighSpeed aus dem Bereich Oszilloskope & Digitaltechnik Markus Stocklas Digital Vertrieb Agilent Technologies Böblingen
Agilent Oscilloscopes - Industry s Largest Portfolio Infiniium 600MHz 80GHz New DCA-X Series 90000 Q-Series 90000 X-Series Fastest Growing Scope Company 9000 Series 90000 Series Economy 20MHz 200MHz Handheld USB InfiniiVision 70MHz 1GHz New New 7000 Series U1600A Series U2700 Series 2000X Series 3000X Series 6000/6000L Series September 1, 2010
InfiniiMax III Series Probing System Probe amps InfiniiMax III probe amps 16 GHz, 20 GHz, 25 GHz, 30 GHz 4 models 16 GHz - 30 GHz Bandwidth upgradeable Probe heads InfiniiMax III ZIF (zero insertion force) probe head 28 GHz InfiniiMax III ZIF probe tips Browser 30 GHz 2.92mm /3.5mm/SMA Probe adapter 28 GHz Solder-in Probe head 16 GHz Probe adapters Sampling scope Adapter Hi impedance probe adapter Precision BNC 50 ohm adapter Performance verification & Deskew fixture
The 90000 Q-Series Overview Orderable: February 1 st / Public Intro April 11 th DSO/DSA 92004Q 92504Q 93304Q 95004Q 95004Q 2 Channel BW 20 GHz 25 GHz 33 GHz 50 GHz 63 GHz 4 Channel BW 20 GHz 25 GHz 33 GHz 33 GHz 33 GHz Sample Rate (2/4 ch) 80 / 80 GS 160 / 80 GS Memory Depth (Std/Max) 20 Mpts DSO / 50 Mpts DSA to 2 Gpts Max Memory Depth Noise at 50 mv/div 0.375% 0.435% 0.502% 1.185% >1.185% Jitter Measurement Floor Maximum Probing BW PrecisionProbe Enabled 75 fs 30 GHz Yes to 63 GHz
Upgradeability The Q-Series Upgradable from 20 GHz to 63 GHz in the following step: 20 to 25 GHz 25 to 33 GHz 33 to 50 GHz 50 to 63 GHz The 90000 X to Q Series Upgradable at two bandwidth points: 20 to 20 GHz 33 to 33 GHz
Technology Leveraged from the 90000 X-Series Orderable: February 1 st / Public Intro April 11 th Probe Amp ADC Amp Trigger IC Indium Phosphide Chips InP Chipset Input Preamp Calibration IC Sampling DeMux IC Process Performance InP Benefits Captive process High-speed & high-voltage Flat response Extensible
Technology Leveraged from the 90000 X-Series Orderable: February 1 st / Public Intro April 11 th Probe Amp ADC Amp InP Chipset Input Preamp Trigger IC Calibration IC Sampling DeMux IC Packaging Quick Film 3D Packaging Custom Agilent technology Exceptional signal integrity Substrate keeps chipset cool and reliable
Technology Leveraged from the 90000 X-Series Orderable: February 1 st / Public Intro April 11 th Probe Amp ADC Amp InP Chipset Input Preamp Trigger IC Calibration IC Sampling DeMux The World s Fastest Most Accurate Oscilloscope Differentiating Technology Exclusive 33 GHz InP preamplifier Packaged for highest signal integrity Pipelined A/D architecture Enables Differentiating Performance True-analog bandwidth to 63 GHz Industry leading low-noise & jitter Industry s only 30 GHz probing system In The World s Fastest and Most Accurate Scope
First Demonstrated Measurements at > 60 GHz Orderable: February 1 st / Public Intro April 11 th 5.1 ps rise time (20/80) 60 GHz sine wave
Challenges In Digital Design Today Higher Data Rates Are Causing Signal Integrity (SI) Problems: Signal Integrity = Where the electrical properties of the interconnects can cause significant distortions in digital signals. >1 GHz of bandwidth <1 ns risetime Typically >2 Gb/s data rate with embedded clock Signal Integrity = Paying attention to RF effects, ie. Impedance FPGAs Are Commonplace Standards Evolve Every 2-3 Years: PCI Express 8 Gb/s 5 Gb/s 2.5 Gb/s 2009 2014 2006 2011 2003 2008 10 HIT 2011 Agilent Restricted March 2011
What is Jitter? What is an Eye Diagram? Jitter is another word for shaky, quiver, tremulous speaks of degree of instability of location. In the Digital Design world, jitter has been defined as: The short term phase variation of the significant instants of a digital signal from their ideal positions in time. Page 11
How Do Real Time Scopes Measure Jitter on Data? NRZ Serial Data Recovered Clock Jitter Trend Jitter Spectrum Units in Time Units in Time Jitter Histogram Page 12
How Do Real Time Scopes Measure Jitter on Data? NRZ Serial Data Recovered Clock Jitter Trend Jitter Spectrum Units in Time Units in Time Jitter Histogram Page 13
Where Does Jitter Come From? Transmitter Media Receiver Lossy interconnect (ISI) Impedance mismatches (ISI) Crosstalk (PJ) Thermal Noise (RJ) DutyCycle Distortion (DCD) Power Supply Noise (RJ, PJ) On chip coupling (PJ) Termination Errors (ISI) Thermal Noise (RJ) DutyCycle Distortion (DCD) Power Supply Noise (RJ, PJ) On chip coupling (PJ) Page 14
The Eye Diagram and Sampling Point E 1 Single transition Left Edge Nominal Sampling Point Eye Crossing Points Right Edge E 0 x = 0 x = 1/2 T x = T The EYE Diagram Unit Interval Oscilloscope Eye Overlaid transitions Total Jitter, J PP Ideal Sampling Point Probability Density Function Page 15
What is the Eye Diagram? 0 0 0 1 0 0 0 1 0 1 1 0 0 0 1 Eye Diagrams 1 0 1 0 1 1 1 1 1 Superimposed Bit Sequences Page 16 ADVA Training 86100C DCA-J Basics
Why Do We Care About Jitter? The only reason we analyze jitter is to Limit the Bit Error Ratio! A low Signal to Noise Ratio causes errors Voltage Noise vertical fluctuations across the sampling point Undesirable Amplitude Modulation Jitter describes the same effect but horizontally timing noise Jitter horizontal fluctuations across the sampling point Undesirable Phase Modulation Page 17
Which Eye Has Worse Jitter? A B You can t know unless you measure the Total Jitter or measure the jitter components! Page 18
Decomposing Jitter Signal jitter can be composed of several types from several mechanisms Data-Correlated j B (t) Total Jitter (TJ) Data-Uncorrelated j B (t) Deterministic Jitter (DJ) j UB (t) Random Jitter (RJ) j UB (t)= Gaussian Inter-symbol Interference (ISI) j B (t)=f(bw, data) Duty Cycle Distortion (DCD) j B (t)= d Periodic Jitter PJ j B (t)=a sin(2pft)
Agilent Fixture De-embedding & Equalization
Source of Measurement Inaccuracies impedance mismatches probing effects smaller geometries test cables and adapters fixturing device packaging, etc. SCOPE NOISE FLOOR! There are multiple ways to offset these measurement impairments. calibration methods mathematical signal processing de-embedding/embedding techniques Scope noise can be amplified by de-embedding techniques
Connector Connector De-embedding Loss Compensation or Gain Function (De-convolve) Compensate for Probing and Fixture Loss Add Margin to Transmitter Characterization PCI Express, SATA, and Custom Compliance Requirement for Gen 2, 3 Tx PHY Channel S4P Rx PHY
Connector Connector Embedding Loss Function (Convolve) Virtual Probe and Deemphasis/Equalization Connector Pin Rx TP1 TP2 TP3 Simulate Channel Loss on Signal Measured at Tx Simulate Equalization/Deemphasis at Rx Tx Signal Tx PHY Channel Channel.s4p+ conn.s4p+package.s4p Rx PHY Virtual Probe Rx Equalization TP1 TP3 TP2
Waveform Transformation Waveform Transformation maps an acquired waveform to another waveform mathematically using a transfer function model of the customer s system. The system may be real using actual customer components or may be virtual. Components may be described using RLC or S-Parameters. Look Here Look Here Look Here M Digital Source Connector Fixture Cable Cable Model And Look Here!
Product Briefing for N5465A Measure anywhere! Listen for these words: - De-Embedding - Fixture Removal - Cable Insertion - Virtual Probing - Probe Loading compensation InfiniiSim Waveform Transformation Toolset enables our customers to define the measurement environment to obtain the best measurement possible.
<Product> Position versus other Agilent Products <Visual of where <product> is positioned in product family> Probe at VIA De-embed Probe Probe at BGA RT BGA = 390 ps RT VIA = 183 ps RT De-embed = 175 ps
Serial Data Equalization Software for Infiniium Series Oscilloscopes DFE CTLE FFE automatic tap optimization Up to 40 Tabs
Transmitter De-emphasis We can account for loss through the channel at the transmitter with transmitter de-emphasis. De-emphasis off, measured at receiver De-emphasis is also called preemphasis. The amount of de-emphasis may be programmable. De-emphasis on, measured at transmitter De-emphasis on, measured at receiver
Key measure is eye quality Unequalized 1Gb/s Unequalized 3Gb/s Unequalized 5Gb/s Unequalized 8Gb/s
Feed-Forward Equalization r(t-ntd) is the input waveform n tap delays before the present time TD is the tap delay Cn is the nth coefficient (tap e(t) is the equalized waveform at time t Each voltage level is multiplied by its corresponding tap value and then all of these products are summed together to give the new equalized voltage for the location of interest
Decision Feedback Equalization r(t) is the unequalized analog waveform voltage at time t s(k n) is the logic value (either upper target or lower target) for the bit n tap delays prior to the current bit Cn is the nth coefficient (tap) V(k) is the correction voltage added to the decision threshold used when determining the logic value of bit k. Each previous bit used in the algorithm is determined to be either high or low and is then multiplied by its corresponding tap value. These voltage/tap products are summed to determine how much to shift the waveform relative to the logical decision threshold.
Differentiation between FFE & DFE Uses voltage levels of the received waveform associated with previous and current bits to correct the voltage level of the current bit 2x taps can open up the eye dramatically Makes logical decisions (zero or one) and then feeds that information back to help determinie whether the current bit is a 1 or 0. The only location of the eye that a receiver sees is at the clock (center of the eye).
N5461A SDE software: Equalization for 5Gb/s Unequalized(5Gb/s): upper left FFE: lower middle 2 taps Eye width of 1/3 DFE: lower right 3 taps Eye width of 0 FFE DFE
Equalizer Setup Window for FFE and DFE settings
Precision Probe Characterize and correct for cable, switch, and test fixture loss using only an oscilloscope
The Importance of a Flat Frequency Response Why do we care about a flat frequency response? - The flatter the response the more accurately the scope will depict the signal - Measurements become more repeatable Frequency response of the Agilent 90000 X- Series
Cables and channels are lossy Response of cable rated to 20 GHz Notice that the 3dB down point is actually at 18 GHz instead of 20 GHz Bandwidth roll-off means attenuated signal at high bandwidth Frequency response is not flat
Every input path in a switch can vary A real life example: 112 different inputs, every one with slightly different loss, phase, and response characteristics Measurements can from channel to channel Technologies continue to push us to use more inputs (for example: PCIE gen3 now has 16 inputs) How much actual bandwidth does this system have?
Every input path in a switch can vary A real life example: 112 different inputs, every one with slightly different loss, phase, and response characteristics Measurements can from channel to channel Technologies continue to push us to use more inputs (for example: PCIE gen3 now has 16 inputs) Customer thought: 12 GHz BW Actual: 4.5 GHz BW
Why is this an issue? Inaccurate, non-repeatable measurements Solution? The solution is to: 1. Understand what the characteristic of the system is, and 2. Compensate for this system variation
Traditional De-embedding Takes Time Option 1: Six steps (you would need to do the following) Find a VNA Find someone that knows how to use a VNA and measure the cable Create s- parameter file Save s-parameter file to thumbdrive and load on scope Learn waveform transformation software and correctly remove loss Analyze the data As a result, we tend to choose to ignore the cable loss and channel variation entirely
How it works Agilent s 90000 X- Series uses its world class 200 GHz Indium Phosphide technology to provide a <15ps edge to the oscilloscope Comparing the baseline measurement with the cables influence, proper characterization is done and corrections can be made Infiniium s custom InP calibration edge Fast edge or Baseline Calibration edge is then measured by the 90000 X-Series Edge with lossy cable Lossy cable is then measured against the fast edge
PrecisionProbe: 3 Easy Steps 1. Measure baseline 2. Measure loss due to cable 3. Save File
Cables: The result Applied corrected filter Corrected cable response Response of cable with no correction
Cable correction results Before PrecisionProbe After PrecisionProbe S21 cable loss is removed through compensation Rise Time improves from 67 ps to 21 ps!
Jitter results Before PrecisionProbe S21 cable loss is removed through compensation Notice how there is 50% less ISI on the corrected waveform, resulting in less Total Jitter After PrecisionProbe
The real time eye Results: More margins! 20% less jitter 33% more eye height Slightly wider eye
Cabled Environment Benefits By characterizing and compensating for cable loss increased margins. Higher accuracy Faster than traditional VNA/de-embedding method
Probe characteristics are different from probe to probe 1. InfiniiMax II browser will have a different voltage transfer function than in InfiniiMax III with a browser 2. Two InfiniiMax II s with a browser could be different 3. Even changing the span on a browser can change the voltage transfer function of a probe
Measuring the Probe No probe Probe 1. Measure baseline 3. Save File 2. Measure loss due to probe Improve your measurement quality
Probe characterization: final results Transfer function of probe with no correction 1. Transfer function is now flat for the entire bandwdith of the probe 2. 6dB of loss is now compensated
Summary: 1. Cables, probes, fixtures, switches are lossy and cause measurement errors Uncorrected 2. Traditional de-embedding technology is time-consuming and equipment intensive. 3. Precision Probe make deembedded incredibly simply. Corrected PrecisionProbe will further increase your margins without adding significant time or extra equipment
Protocol Decode on the Infiniium Series USB 2.0 Example Multi-tab protocol viewer, time aligned with analog Header view With formatted frame content Payload view Search by packet type Jump to next search The competition has no equivalent for any of these Agilent protocol features. 53
Agilent Decode Ease-of-Use Advantage Setup multiple decodes simultaneously and quickly switch back and forth between them in the Decode Listing window. This button automatically sets up the scope parameters for your specific decode. With this Auto Setup button, you can be decoding data in less than a minute. If you want to see what values were set during the Auto Setup process or if you want to change certain parameters, use this Manual Setup button.
Infiniium Protocol Analysis Timecorrelated Protocol detailled level Search and capture Multi Serial Bus Colorcoded representation Available for ALL (most) serial protocolls:
Supports both MSO Digital and Analog Channels I2C + SPI Example Using MSO channels for protocol decode Using analog channels for protocol decode
View decode in waveform area, or...
Simultaneous decode of up to 4 Serial Buses Protocol trigger/decode sources can come from any of the following: 4 analog channels +16 digital + 4 functions + 4 waveform memories Anticipated usage primarily for protocols with explicit clocks such as I 2 C, SPI, CAN, LIN, MIPI, USB 2.0, RS-232/UART...
InfiniScan the real nice way to trigger!! Apply up to 8 Zone to trigger the signal individually Trigger on Measurement parameters Find and trigger non-monotic Edges Serial Pattern Trigger
InfiniScan the real nice way to trigger!!
Market Trends 1. Analog continues to go digital Vinyl record Videotape Photo Camera Cellphone 1 0 1 1 0 1 CD DVD, Blue Ray Digital Camera Digital Cellphone 2. More integration at lower cost 3. Higher speed and lower power 4. High reliability delivered in less time
Ecosystem
Agilent Digital Test Standards Program Our solutions are driven and supported by Agilent experts involved in international standards committees: Joint Electronic Devices Engineering Council (JEDEC) PCI Special Interest Group (PCI-SIG ) Video Electronics Standards Association (VESA) Serial ATA International Organization (SATA-IO) USB-Implementers Forum (USB-IF) Mobile Industry Processor Interface (MIPI) Alliance Optical Internetworking Forum (OIF) We re active in standards meetings, workshops, plugfests, and seminars Our customers test with highest confidence and achieve compliance faster
We understand your future requirements, because we help shape them Rick Eads PCI-Sig Board Member Brian Fetz DisplayPort Phy CTS Editor VESA Board Member Jim Choate USB-IF Compliance Committee USB 3.0 Electrical Test Spec WG Min-Jie Chong SATA/SAS PHY Contributor MIPI-PHY WG Contributor Roland Scherzinger MIPI Contributor The Agilent Pyramid team maintains engagement in the top high tech standards organizations SATA / SAS Test Challenges Page 64 Agilent Restricted
A Digital system: Serial Data Link Life Cycle of a Transported bit Die Bonding wire/pins PC transmission line Standard connector Cable Standard connector PC transmission line Bonding wire/pins Die Starts in here and ends in here Page 65 HIT 2011 Agilent Restricted March 2011
ADS EMPro, and SystemVue Pulse Pattern Generator Infiniium 90000 X-Series More Info A range of essential tools measurement and simulation that will help you cut through the challenges of gigabit digital designs. www.agilent.com/find/hsd Blog: http://signal-integrity.tm.agilent.com N1930B Physical Layer Test System (PLTS) 16900 Series Logic Analyzers Bit Error Ratio Testers (BERTs) 86100D Infiniium DCA-X ENA-TDR 66 HIT 2011 Agilent Restricted March 2011
SuperSpeed Communication Physical Layer Focus Point to point communication, concurrent data flow Low power mode Link training Independent clock domains both using Spread Spectrum Clocking (SSC) Non- Super Speed TX RX TX RX Non- Super Speed Super Speed TX RX RX TX Super Speed -or- -or- Transmitter (TX) De-emphasis 8B/10B coding Data scrambling Insertion of Skip Cable / Channel Backward compatible EMI requirements Signal integrity requirements Receiver (RX) Channel equalization Clock recovery Re-timing (deletion or insertion of addition Skips) How to handle USB 3.0 physical layer test requirements October 28, 2009
Physical Layer Test Solutions Non- Super Speed Super Speed -or- -or- TX Trans- RX mitter (TX) TX RX Channel / Cable TX Non- Super RX Receiver Speed RX TX (RX) Super Speed Agilent 90000 series Infiniium Oscilloscopes Agilent U7243A USB 3.0 Transmitter Compliance Test Software Agilent E5071C Network Analyzer 86100C DCA-J TDR Agilent J-BERT N4903B with N4916A/B De-emphasis Signal Converter Agilent 81250A ParBERT All physical layer tests: test adapter Agilent U7242A USB 3.0 Test Fixture N5990A Automated Compliance and Characterization Test Software How to handle USB 3.0 physical layer test requirements October 28, 2009
SuperSpeed Measurement Requirements Transmitter Compliance Testing: Compliance will be measured at the end of the compliance channel SMA termination for TX signals, phase matched SMA cable Terminate link under test with high speed oscilloscope Measure transmitted waveform with high speed oscilloscope Use compliance pattern 1M UI of data Compute: eye diagram, Rj, Dj, Tj@10^-12 BER, average data rate, rise/fall time, Test requirement for SSC Slew Rate USB 3.0 Technical Review Page 69. 2009
Transmitter test requirements USB 3.0 Technical Review Page 70. 2009
Compliance Channels Compliance Channels are being developed to test SQ for worst case channel conditions Back panel USB route solution Channel loss will dominate Front Panel USB route solution Reflections will dominate losses Back Panel Front Panel USB 3.0 Technical Review Page 71. 2009
USB 3.0 Test fixture Support for Tx and Rx Testing SMA edge launch terminations SS A and SS B for host, device or cross hub testing USB 3.0 Test Fixtures (available now) Early Customer Needs and Development USB 3.0 Technical Review Page 72 72. 2009
Normative Transmitter Compliance Test Setup Scope SMA TP0 DSO 90K Scope SMA USB 3.0 Technical Review Page 73. 2009
USB 3.0 Technical Review Page 74. 2009
Summary report of testing with Statistics -0.35 USB 3.0 Technical Review Page 75. 2009
Agilent TX Compliance and Validation Solution Report Summary TX Compliance Test With USB Org Test Tool USB 3.0 Technical Review Page 76. 2009
USB 3.0 Protocol Decode: on scope Page 77
SS Test Adapter Eliminate All Doubt: 5Gb/s Test Signal Example No SSC, Sj / Rj shown on screen shots 1 50Ω 50Ω trigger 2 N4915A-005 Switch Channe l 3 TP 1 4 + RX - + TX - How to handle USB 3.0 physical layer test requirements October 28, 2009
Typical SuperSpeed Link Turn-on Sequence LTSSM states: Host Power-up Rx. Detect. Reset Rx. Detect. Active Polling. LFPS Polling. RxEQ Polling. Active Polling. Configuration Polling. Idle Loopback warm reset de-assert termination detected LFPS handshake TSEQ transmitted TS1 received TS2 received if directed Device Complianc e warm reset Rx. Detect. Reset Rx. Detect. Active Polling. LFPS Polling. RxEQ Polling. Active Polling. Configuration Polling. Idle Loopback multiple states Power-up How to handle USB 3.0 physical layer test requirements October 28, 2009
Characteristical Eye Closure by Sinusoidal Jitter Eye Diagram BER Scan How to handle USB 3.0 physical layer test requirements October 28, 2009
Why is Spread Spectrum Clocking Included in Receiver Compliance Test? SSC stresses clock recovery and elastic buffer, required for compliance test Max. SSC deviation is 5000ppm, modulation rate between 30kHz and 33kHz Nominal data rate: 5Gb/s Frequency downspread: 5000ppm i.e. 4.975Gb/s Receiver EQ FF CR TX Non-SS SuperSpeed RX RX loop- error back count TX read 1 2 -or- 1 2 skp s k p skp s k p skp s k p s k p s k p s kp s kp s kp s kp s kp s kp s p k p s k Original un-modulated data at 5Gb/s Data with SSC at receiver (RX) pins Receiver (RX) elastic buffer compensates for clock difference How to handle USB 3.0 physical layer test requirements October 28, 2009
Characteristical Eye Closure by Random Jitter Eye Diagram BER Scan bounded Rj unbounded Rj How to handle USB 3.0 physical layer test requirements October 28, 2009
Hands-on in der Ausstellung Herzlichen Dank für Ihr Interesse
Thank you for Attending Questions? USB 3.0 Technical Review Page 84. 2009