Notice Notice Notice Notice Notice

Similar documents
Request Request Request Request Request Request Request

SAS-2 6Gbps PHY Specification

40 AND 100 GIGABIT ETHERNET CONSORTIUM

XX.7 Link segment characteristics

Power Optimization in Stratix IV FPGAs

DKAN0008A PIC18 Software UART Timing Requirements

PLANNING AND DESIGN OF FRONT-END FILTERS

TSB-67. Transmission Performance Specifications for Field Testing of UTP Cabling Systems

High Speed Characterization Report

High Data Rate Characterization Report

Bill Ham Martin Ogbuokiri. This clause specifies the electrical performance requirements for shielded and unshielded cables.

Baseline proposals for copper twinaxial cable specifications Chris DiMinico MC Communications/PHY-SI LLC/Panduit

6.976 High Speed Communication Circuits and Systems Lecture 16 Noise in Integer-N Frequency Synthesizers

Software Defined Radio Forum Contribution

High Speed Characterization Report

High Speed Characterization Report

High Data Rate Characterization Report

3.6 Intersymbol interference. 1 Your site here

Lock-In Amplifiers SR510 and SR530 Analog lock-in amplifiers

High Data Rate Characterization Report

High Data Rate Characterization Report

IEEE 802.3bj: 100GBASE-CR4 Test Points and Parameters Chris DiMinico MC Communications/ LEONI Cables & Systems LLC

SILICON DESIGNS, INC Model 1010 DIGITAL ACCELEROMETER

Draft ETSI TR V1.2.1 (2008-xx)

Chris DiMinico MC Communications/PHY-SI LLC/Panduit NGOATH Study Group

ETSI TR V1.2.1 ( ) Technical Report

TR (draft) V0.0.0 (2003-xx)

IEEE Std 802.3ap (Amendment to IEEE Std )

Valve amplifier for proportional pressure valves

EQCD High Speed Characterization Summary

PRODUCT SPECIFICATION

2. Transceiver Design Flow Guide for Stratix IV Devices

AN EFFICIENT SET OF FEATURES FOR PULSE REPETITION INTERVAL MODULATION RECOGNITION

DELTA Test Report. Certification testing of generic cabling products: Unscreened channel, Class E, 4 connector model. Performed for Sterling

2. Transceiver Design Flow Guide

40 AND 100 GIGABIT ETHERNET CONSORTIUM

Complex RF Mixers, Zero-IF Architecture, and Advanced Algorithms: The Black Magic in Next-Generation SDR Transceivers

Solid State Relays & Its

SPICE Model Validation Report

287. The Transient behavior of rails used in electromagnetic railguns: numerical investigations at constant loading velocities

Technical Brief High-Speed Board Design Advisor Thermal Management

RF Characterization Report

SIGNAL INTEGRITY ANALYSIS AND MODELING

Potentiostat stability mystery explained

Homework 3 Solution Set

Signal Integrity Test Report

Agilent AN Applying Error Correction to Network Analyzer Measurements

APPLICATION NOTE #1. Phase NoiseTheory and Measurement 1 INTRODUCTION

Philadelphia University Faculty of Engineering Communication and Electronics Engineering. Amplifier Circuits-III

Tektronix Inc. DisplayPort Standard. Revision Tektronix MOI for Cable Tests (DSA8200 based sampling instrument with IConnect software)

Jan M. Kelner, Cezary Ziółkowski, Leszek Kachel The empirical verification of the location method based on the Doppler effect Proceedings:

AUTOMOTIVE ETHERNET CONSORTIUM

MICTOR. High-Speed Stacking Connector

TEST REPORT. DANAK Reg. no. 19. Electrical performance testing of: Connecting Hardware according to IEC

Bode Plot based Auto-Tuning Enhanced Solution for High Performance Servo Drives

Consumers are looking to wireless

High Speed Characterization Report

PRODUCT SPECIFICATION

IEEE 802.3bj Test Points and Parameters 100 Gb/s copper cable Chris DiMinico MC Communications/ LEONI Cables & Systems LLC

ECE5984 Orthogonal Frequency Division Multiplexing and Related Technologies Fall Mohamed Essam Khedr. Channel Estimation

ECEN 5014, Spring 2013 Special Topics: Active Microwave Circuits and MMICs Zoya Popovic, University of Colorado, Boulder

TR (draft) V0.0.0 (2004-xx)

Qualification testing of 100 ohm shielded channel, Class EA. Performed for Tyco Electronics Raychem N.V.

40 AND 100 GIGABIT ETHERNET CONSORTIUM

Devices selection for the construction of a microwave transmission link at 2.45 GHz

DP Array DPAM/DPAF Final Inch Designs in Serial ATA Generation 1 Applications 10mm Stack Height. REVISION DATE: January 11, 2005

Gb/s Study Group. Considerations for 25 Gb/s Cable Assembly, Test Fixture and Channel Specifications

Hybrid spectrum arrangement and interference mitigation for coexistence between LTE macrocellular and femtocell networks

IEEE 100BASE-T1 Physical Media Attachment Test Suite

25Gb/s Ethernet Channel Design in Context:

Experiment 7: Frequency Modulation and Phase Locked Loops Fall 2009

Q2 QMS-DP/QFS-DP Series 11 mm Stack Height Final Inch Designs in Serial ATA Generation 1 Applications. Revision Date: February 22, 2005

TR (draft) V0.0.0 (2004-xx)

VHDM & VHDM-L Series. High Speed. Electrical Characterization

S22P S05P Series. Application Manual. Overview. Characteristics. Use. S22P series

Richard Mellitz, Intel Corporation January IEEE 802.3by 25 Gb/s Ethernet Task Force

University of New Hampshire InterOperability Laboratory Gigabit Ethernet Consortium

Keysight Technologies Network Analysis Applying the 8510 TRL Calibration for Non-Coaxial Measurements

Independent Testing - For End User Confidence

ECE 528 Understanding Power Quality

n Proper pin termination at each end n Continuity to the remote end n Shorts between any two or more conductors n Transposed pairs: n Crossed pairs.

Finding Loop Gain in Circuits with Embedded Loops

High Speed Communication Circuits and Systems Lecture 10 Mixers

Return Loss of Test Channel for Rx ITT in Clause 136 (#72)

Building An Average Model For Primary-Side Regulated Flyback Converters

Keysight Technologies Applying Error Correction to Vector Network Analyzer Measurements. Application Note

DEVELOPMENT OF CARRIER-PHASE-BASED TWO-WAY SATELLITE TIME AND FREQUENCY TRANSFER (TWSTFT)

Cable Testing TELECOMMUNICATIONS AND NETWORKING

Frequency Hopped Spread Spectrum

Keysight MOI for USB Type-C Connectors & Cable Assemblies Compliance Tests (Type-C to Legacy Cable Assemblies)

54. Physical Medium Dependent (PMD) sublayer and baseband medium, type 10GBASE-CX4

1. Motivation. 2. Periodic non-gaussian noise

Artefact Characterisation for JPEG and JPEG 2000 Image Codecs: Edge Blur and Ringing

Level III measurement accuracy of field testers for Twisted Pair Copper Cabling Explained in Technical Detail

EXPERIMENT EM3 INTRODUCTION TO THE NETWORK ANALYZER

Chapter 6: Introduction to Digital Communication

RiseUp RU8-DP-DV Series 19mm Stack Height Final Inch Designs in Serial ATA Generation 1 Applications. Revision Date: March 18, 2005

EMC filters. 3-line filters for converters and power electronics Rated current 8 to 80 A. Date: January 2006

Where Did My Signal Go?

Fundamentals of Spectrum Analysis. Christoph Rauscher

Transcription:

TITLE: PCI-SIG ENGINEERING CHANGE NOTICE OCuLink Cable Spec ECN DATE: March, 0 AFFECTED DOCUMENT: OCuLink.0 SPONSOR: Part I:. Summary o the Functional Changes Alex Haser (Molex), Jay Neer (Molex) The IL/ itted IL requirements have been clariied. The language o this portion o the spec has been reworked to eliminate conusion and provide uniormity in the subsections included in the ollowing document.. Beneits as a Result o the Changes The signal integrity requirements necessary or OCuLink compliance have been clariied. The method or calculating itted IL has been simpliied. The equations or limit lines have been redeined to report values as losses (positive values).. Assessment o the Impact This section o the spec has been made easier to ollow and is thereore less likely to misinterpret.. Analysis o the Hardware Implications Connectors and cables must adhere to the ully speciied signal integrity requirements described in the ollowing text.. Analysis o the Sotware Implications None, this change does not aect sotware. Part II: Changes are relected in Sections. through.. (begins on the next page).

0. Mated Cable Assembly Electrical Speciications The OCuLink cable assembly contains insulated conductors terminated in a connector at each end or use as a link segment between host boards. This cable assembly is primarily intended as a point-to-point interace between host boards using controlled impedance cables. All mated cable assembly measurements are to be made between TP and TP with CCB test ixtures, which includes a connector mated to either end o the cable assembly. CCB test ixtures are to be removed beore comparing measured perormance against the speciications listed in this section. It is recommended measurements be de-embedded using a TRL calibration (see Section. or more inormation on de-embedding). All requirements are to be met when analyzed in an Ω reerence rame. The mated cable assembly speciications are based upon twin axial cable characteristics. Table - provides a summary o the cable assembly characteristics and reerences addressing each parameter. Limits apply to. GT/s,.0 GT/s, and.0 GT/s data rates and are written in terms o baud requency. Baud requencies are. GHz, GHz, and GHz or the. GT/s,.0 GT/s, and.0 GT/s data rates, respectively. Note that plots show loss (positive values). Table -. Mated Cable Assembly Dierential Characteristics Summary Description Reerence Value Unit Maximum dierential insertion loss.. Equation - db Minimum dierential return loss.. Equation - db Minimum dierential to common-mode return loss.. Equation - db Minimum dierential to common-mode conversion loss minus insertion loss.. Equation - db Minimum common-mode return loss.. Equation - db Minimum MDNEXT loss.. Equation - db Minimum MDFEXT loss.. Equation - db Maximum total cable assembly, Lane-to-Lane skew (Sc).. 0. ns 0... Characteristic Impedance and Reerence Impedance The nominal dierential characteristic impedance o the cable assembly is Ω. The dierential reerence impedance or cable assembly must be Ω.

0... Mated Cable Assembly Dierential Insertion Loss The dierential insertion loss o each pair o a mated OCuLink cable assembly, in db, must be less than the maximum limit deined in Equation -, as illustrated in Figure -.. GT/s,.0 GT/s, and.0 GT/s: Equation - ILCable() = +. + 0. or 0.0 GHz b GHz b ILCable() is the maximum mated cable assembly insertion loss, in db Acceptable Region Figure -: Maximum Mated Cable Assembly Insertion Loss This insertion loss limit is intended to accommodate cables up to m long (using awg).

0... Mated Cable Assembly Dierential Return Loss The dierential return loss o each pair o the mated OCuLink cable assembly, in db, must be greater than the minimum limit deined in Equation -, as illustrated in Figure -.. GT/s,.0 GT/s, and.0 GT/s: 0 db or 0.0 GHz GHz Equation - RLCable()= { ( ) db or GHz < b GHz b RLCable() is the maximum mated cable assembly return loss, in db Acceptable Region Figure -: Minimum Mated Cable Assembly Return Loss

... Mated Cable Assembly Dierential to Commonmode Return Loss The dierential to common-mode return loss o the mated OCuLink cable assembly, in db, must be greater than the minimum deined in Equation -, which is illustrated in Figure -.. GT/s,.0 GT/s, and.0 GT/s: 0 Equation - DCMCCable() = 0 ( ) or 0.0 GHz b GHz b DCMCRL() is the minimum mated cable assembly dierential to common-mode return loss, in db Acceptable Region Figure -: Minimum Mated Cable Assembly Dierential to Commonmode Return Loss

0 0... Mated Cable Assembly Dierential to Commonmode Conversion Loss minus Insertion Loss The dierence between the mated OCuLink cable assembly dierential to common-mode conversion loss and the cable assembly insertion loss, in db, must be greater than the minimum limit deined in Equation -.. GT/s,.0 GT/s, and.0 GT/s: Equation - DCMC-ILCable() = 0 db or 0.0 GHz b GHz b DCMC() ILCable() is the mated cable assembly dierential to common-monde insertion loss, in db is the mated cable assembly dierential insertion loss, in db... Mated Cable Assembly Common-mode Return Loss The common-mode return loss o the mated OCuLink cable assembly, in db must be greater than the minimum limit deined in Equation -.. GT/s,.0 GT/s, and.0 GT/s: Equation - CMRLCable() = db or 0.0 GHz b GHz b CMRLCable() is the minimum mated cable assembly common-mode return loss, in db

0... Mated Cable Assembly Crosstalk Crosstalk between dierential pairs inluences the data signals and any subsequent loss and jitter budgets. All system board and cable assembly designs must properly account or any crosstalk that may exist among the various pairs o dierential signals. Crosstalk is due to coupling through a channel, either at the near-end (NEXT) or at the ar-end (FEXT). The total contribution rom all aggressors on a particular victim is captured by the multidisturber or powersum crosstalk. The NEXT loss that couples within a channel is rom Lanes that transmit data in the opposite direction as the victim Lane. Thereore, MDNEXT loss on a receive Lane must be summed across contributions rom the our transmit Lanes at the near end o the mated cable assembly, yielding our aggressors total or a our-lane interace. Similarly, the FEXT loss that couples within a channel is rom the Lanes that transmit data in the same direction as the victim Lane. Thereore, MDFEXT loss on a receive Lane must be summed across contributions rom the three remaining receive Lanes at the ar end o the mated cable assembly, yielding three aggressors total or a our-lane interace. Equation - provides the calculation or MDNEXT loss; Equation - provides the calculation or MDFEXT loss. 0 0 Equation - Equation - i= MDNEXT() = 0 log 0 0 NL i()/0 or 0.0 GHz b GHz i=0 b MDNEXT() is the multi-disturber near-end crosstalk loss at requency NLi() is the NEXT loss at requency o victim-aggressor combination i, in db i is the victim-aggressor pair (0 to ) i= MDFEXT() = 0 log 0 0 FL i()/0 or 0.0 GHz b GHz i=0 b MDFEXT() is the multi-disturber ar-end crosstalk loss at requency FLi() is the FEXT loss at requency o victim-aggressor combination i, in db i is the victim-aggressor pair (0 to )

The MDNEXT loss and MDFEXT loss summed on each receive Lane o a mated OCuLink cable must be greater than the limits deined in Equations - and - respectively. These limits are illustrated in Figure -.. GT/s,.0 GT/s, and.0 GT/s: 0 Equation - MDNEXTCable() =.. log 0 ( ) db or 0.0 GHz b GHz b MDNEXTCable. GT/s,.0 GT/s, and.0 GT/s: is the minimum mated cable assembly MDNEXT loss, in db Equation - MDFEXTCable() = log 0 ( ) db or 0.0 GHz b GHz b MDFEXTCable is the minimum mated cable assembly MDFEXT loss, in db Acceptable Region Figure -: Minimum Mated Cable Assembly MDNEXT Loss and MDFEXT Loss

0... Lane-to-Lane Skew The time delay across a cable assembly at any point is measured using zero crossings o dierential voltage o the compliance pattern, while simultaneously transmitting on all physical Lanes. The compliance pattern is deined in the PCI Express Base Speciication. The skew or a mated cable assembly is quantiied as the dierence between the longest and shortest time delays measured or a given assembly. The maximum skew permitted or. GT/s,.0 GT/s, and.0 GT/s OCuLink cable assemblies in 0.ns.... Within-Lane Skew Maximum within-lane skew is not speciied in this speciication in the time domain. Within-Lane skew is instead controlled in the requency domain by means o the restrictions on dierential-to-common mode return loss and dierential-to-common mode conversion loss minus insertion loss.