N5990A User Guide for PCI
Keysight N5990A
Test Automation
Software Platform
for PCIe
User Guide
Notices
Copyright Notice
© Keysight Technologies 2015
No part of this manual may be reproduced in any form or by any means
(including electronic storage and
retrieval or translation into a foreign
language) without prior agreement
and written con- sent from Keysight
Technologies, Inc. as governed by
United States and interna- tional
copyright laws.
Manual Part Number
N5990-91040
Edition
Edition 5.0, September 2015
Published by:
Keysight Technologies
Deutschland GmbH,
Herrenberger Str. 130,
71034 Böblingen, Germany
Technology Licenses
The hardware and/or software
described in this document are
furnished under a license and may be
used or copied only in accordance
with the terms of such license.
U.S. Government
Rights
The Software is “commercial computer
software,” as defined by Federal
Acquisi- tion Regulation (“FAR”)
2.101. Pursuant to FAR 12.212 and
27.405-3 and Depart- ment of
Defense FAR Supplement (“DFARS”)
227.7202, the U.S. govern-ment
acquires commercial computer
software under the same terms by
which the software is customarily
provided to the public. Accordingly,
Keysight pro- vides the Software to
U.S. government customers under its
standard commercial license, which is
embodied in its End User License
Agreement (EULA), a copy of which
can be found at http://
www.keysight.com/find/sweula. The
li- cense set forth in the EULA
represents the exclusive authority by
which the U.S. government may use,
modify, distribute, or disclose the
Software. The EULA and the license
set forth therein, does not require or
permit, among other things, that
Keysight: (1) Furnish technical information related to commercial
computer software or commercial
computer soft- ware documentation
that is not custom- arily provided to
the public; or (2) Relin- quish to, or
otherwise provide, the gov- ernment
rights in excess of these rights
customarily provided to the public to
use, modify, reproduce, release,
perform, dis- play, or disclose
commercial computer software or
commercial computer soft- ware
documentation. No additional government requirements beyond those
set forth in the EULA shall apply,
except to the extent that those terms,
rights, or licenses are explicitly
required from all providers of
commercial computer soft- ware
pursuant to the FAR and the DFARS
and are set forth specifically in writing
elsewhere in the EULA. Keysight shall
be under no obligation to update,
revise or otherwise modify the
Software. With respect to any
technical data as defined by FAR
2.101, pursuant to FAR 12.211 and
27.404.2 and DFARS 227.7102, the
U.S. government acquires no
greater than Limited Rights as
defined in FAR
27.401 or DFAR 227.7103-5 (c), as
appli- cable in any technical data.
Warranty
THE MATERIAL CONTAINED IN
THIS DOCUMENT IS PROVIDED “AS
IS,” AND IS SUBJECT TO BEING
CHANGED, WITHOUT NOTICE, IN
FUTURE EDITIONS. FURTHER, TO
THE MAXIMUM EXTENT
PERMITTED BY APPLICABLE LAW,
KEYSIGHT DISCLAIMS ALL
WARRANTIES, EI- THER EXPRESS
OR IMPLIED, WITH REGARD TO
THIS MANUAL AND ANY
INFORMATION CONTAINED
HEREIN, INCLUDING BUT NOT LIMITED TO THE IMPLIED
WARRANTIES OF
MERCHANTABILITY AND FIT- NESS
FOR A PARTICULAR PUR- POSE.
KEYSIGHT SHALL NOT BE LIABLE
FOR ERRORS OR FOR INCIDENTAL OR CONSEQUENTIAL
DAM- AGES IN CONNECTION WITH
THE FURNISHING, USE, OR
PERFOR- MANCE OF THIS
DOCUMENT OR OF ANY
INFORMATION CONTAINED
HEREIN. SHOULD KEYSIGHT AND
THE USER HAVE A SEPARATE
WRITTEN AGREEMENT WITH WARRANTY TERMS COVERING THE MATERIAL IN THIS DOCUMENT THAT
CONFLICT WITH THESE TERMS,
THE WARRANTY TERMS IN
THE SEPARATE AGREEMENT
SHALL CONTROL.
Safety Information
CAUTION
A CAUTION notice denotes a hazard.
It calls attention to an operating
procedure, practice, or the like that,
if not correctly performed or adhered
to, could result in damage to the
product or loss of important data.
Do not proceed beyond a CAUTION
notice until the indicated conditions
are fully understood and met.
WARNING
A WARNING notice denotes a haz- ard.
It calls attention to an operating
procedure, practice, or the like that, if
not correctly performed or adhered to,
could result in personal injury or death.
Do not proceed beyond a WARNING
notice until the indicated conditions are
fully understood and met.
Contents
Contents
1
Introduction
1.1
What’s in This Chapter 7
1.1.1
1.2
2
Document History 7
N5990A Overview
2.1
3
Test Automation Software Platform 9
Using Software
3.1
Test Station Configuration 11
3.1.1
3.2
3.3
Using the Software 32
Trouble Shooting 34
Computer PCI Express Bus Test Application
4.1
PCI Express (Peripheral Component Interconnect Express)
4.1.1
4.1.2
4.1.3
4.2
37
Supported Hardware Configurations 37
Configuring a PCI Express Station 37
Configuring a PCI Express DUT 42
PCIe3 Module Procedure Descriptions 43
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5
4.2.6
6
Configuring DUT 19
Selecting, Modifying, & Running Tests 22
Results 26
Oscilloscope Transmitter Test Integration 32
3.3.1
3.3.2
5
Using Keysight IO VISA Connection Expert 16
Starting Test Station 18
3.2.1
3.2.2
3.2.3
4
Overview of This Guide 7
Configure DUT 44
Gen1 & Gen2 ASIC Tests 57
Gen3 ASIC Tests 87
Gen1 & Gen2 CEM Tests 196
Gen3 CEM Tests 230
Rx Switch Matrix 305
Troubleshooting and Support
5.1
Log List and File 311
6.1
Data Structure and Backup 313
Appendix
6.1.1
6.1.2
ValiFrame Data Straucture 313
ValiFrame Backup 315
Keysight N5990A Test Automation Software Platform for PCIe User Guide
5
Contents
6.2
Remote Interface 316
6.2.1
6.2.2
6.2.3
6.2.4
6.3
Controlling Loop Parameters and Looping Over Selected Tests 324
6.3.1
6.3.2
6.3.3
6.3.4
6.3.5
6.3.6
6.4
6
Connect() 326
SetToDefault() 326
Init() 326
GetParameterList() and GetParameterValues()
SetNextValue() 326
Disconnect() 327
326
IBerReader328
6.4.1
6.5
Introduction 316
Interface Description 317
Using the Remote Interface 319
Results Format 322
IBerReader Interface 329
Main Power Switch Control 331
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Keysight N5990A Test Automation Software Platform for PCIe
User Guide
1
1.1
Introduction
1.1
What’s in This Chapter / 7
1.2
Document History / 7
What’s in This Chapter
This chapter provides an introduction of this user guide.
1.1.1 Overview of This Guide
This guide provides a detailed description of the N5990A Test Automation Software
Platfom.
1.2
Document History
First Edition
(September, 2014)
The first edition of this user guide describes functionality of software version N5990A
ValiFrame_2.23_PCI-Express3_1.40 or higher.
Second Edition
(October, 2014)
The second edition of this user guide describes functionality of software version
N5990A ValiFrame_2.23_PCI-Express3_1.41 or higher.
Third Edition
(January, 2015)
The third edition of this user guide describes functionality of software version N5990A
ValiFrame_2.23_PCI-Express3_1.50 or higher.
Fourth Edition
(May, 2015)
The fourth edition of this user guide describes functionality of software version
N5990A ValiFrame_2.23_PCI-Express3_1.55 or higher.
Fifth Edition
(September, 2015)
The fifth edition of this user guide describes functionality of software version N5990A
ValiFrame_2.23_PCI-Express3_1.60 or higher.
Keysight N5990A Test Automation Software Platform for PCIe
User Guide
2
N5990A Overview
2.1
2.1
Test Automation Software Platform / 9
Test Automation Software Platform
The Keysight Technologies N5990A Test Automation Software Platform “ValiFrame” is
an open and flexible framework for automating tests such as electrical compliance
tests for PCI Express digital bus.
The product runs on a standard PC that controls a wide range of test
hardware. Typically, the hardware comprises of instruments for
stimulus and response tests such as pattern generators, bit errror
ratio testers (BERTs), and oscilloscopes. Key elements of the
software platform are a test sequencer, receiver test libraries, and
interfaces to oscilloscope applications for transmitter tests.
Additional options are available, e.g. User Programming.
N5990A is impemented in C# within the Microsoft .NET Framework.
The software platform is specified in the data sheet 5989-5483EN,
incl. the PC requirements. Application examples for PCI Express®
are given in the application notes 5989-5500EN.
Keysight N5990A Test Automation Software Platform for PCIe
User Guide
3
3.1
Using Software
3.1
Test Station Configuration / 11
3.2
Starting Test Station / 18
3.3
Oscilloscope Transmitter Test Integration / 32
Test Station Configuration
Test Station Selection
The set of test instruments used for a specific application is referred to in the
following as "Test Station" or short "Station". The test station is controlled by a
suitable PC and the N5990A Test Automation Software Platform. At first, ValiFrame
Station Configuration
(Start > All Programs > BitifEye> <Application>) needs to be started prior to
“ValiFrame” (see Figure 3-1 and Figure 3-6).
When the ValiFrame Station Configuration is started, a window appears as shown in
Figure 3-2.
Figure 3-1: ValiFrame station configuration icon
3
Using Software
Figure 3-2: N5990A station selection window
The available Test Stations are listed in a drop-down menu (highlighted in Figure
3-2). Multiple entries can be generated by User Programming (N5990A opt. 500) and
select the station to be used.
The N5990A Test Automation Software Platform supports the PCI Express
applications.
The N5990A primarily provides physical layer test automation.
The N5990A opt. 001 is the interface to SQL databases (and web browsers). In case
this option was purchased, the connection to the database application server is
established by unchecking the default "Database Offline" selection and entering the
IP address of the server. Proceed with “Next” or quit with “Cancel”.
The test results can be displayed in two different ways: in MS-Excel worksheets or in
an HTML viewer.
Pressing the “Next” Button opens a ValiFrame Station Configuration window as given
in Figure 3-3.
12
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
Figure 3-3: N5990A station configuration window
Test Station
Configuration
Depending on the selected application in Figure 3-2. The ValiFrame Station
Configuration window shows the instruments or instrument combinations that are
needed. All the required instruments can be selected using the drop-down menus
here and click on “Next” button to continue.
The user must ensure that all the selected instruments for the test station are
connected to the test station PC controller by remote control interfaces such as LAN,
USB or GPIB.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
13
3
Using Software
Figure 3-4: N5990A instrument configuration window
14
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
After all required instruments have been selected in Figure 3-3, those are listed in
the ValiFrame Instrument Configuration Window (Figure 3-4). In order to control
instruments for use with the test station, instrument connections need to be
established by using specific hardware addresses as described in the following
section. The "Mode" check box needs to be checked to set a specific instrument
status from "Offline" to "Online".
When starting a specific test station configuration for the first time, all instruments
are set to the “Offline” mode. In this mode the test automation software does not
connect to any instrument. This mode can be used for demonstrations or checks.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
15
3
Using Software
3.1.1 Using Keysight IO VISA Connection Expert
Introduction
The Keysight IO VISA Connection Expert is recommended to setup new connections
or verify existing connections. Start the Connection Expert with a right-click on the
VISA icon in the task bar and select “Keysight Connection Expert”. A window popsup as shown in Figure 3-5.
Figure 3-5: Keysight IO VISA connection expert window
Under “Instruments”, select “Rescan” (highlighted in Figure 3-5). For each
instrument that is needed, verify that an entry exists in the list for the instrument and
that before the VISA Address there is a green checkmark. The connection to
instruments can be verified by using the Keysight Visa Assistant, which is available in
the same menu.
Once all the instruments to be used are listed properly, their address strings can be
entered in the ValiFrame Instrument Configuration Window (Figure 3-4). The
recommended way of doing this is by copying and pasting instrument addresses as
follows:
Click on the “VISA Address” field next to an instrument in the Connection Expert.
Copy the address, highlight the same instrument in the Test Station Connection
window, paste the address in the text field (highlighted in Figure 3-4) and click
“Apply Address”. Repeat this process for all instruments being used, except the
ParBERT and standard specific applications running on the oscilloscope.
16
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
The applications running on the oscilloscope use a different technology to provide
remote access to ValiFrame, called .NET Remoting. Communication is only possible
using the LAN connection of the oscilloscope and for this reason the IP address
needs to be used with this type of instrument.
Enter the instrument address in the text field as shown in Figure 3-4 and press
“Apply Address” to set it. Once all the instruments are set with the appropriate
addresses that may be used, select the instruments that will be used by the Test
Automation Software by checking the tick box next to “Offline” in the “Mode” column.
Use the “Check Connections” button to verify that the instrument addresses are
valid.
Once you click the “Configure” button, the changes will be implemented and the Test
Station Configuration window will be closed.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
17
3
Using Software
3.2
Starting Test Station
Start the ValiFrame test station with a double click on the icon on the desktop
(example for PCIe test station given in Figure 3-6). Alternatively, start the ValiFrame
station from “Start / All Programs / BitifEye”.
Figure 3-6: ValiFrame PCIe test station icon
The ValiFrame N5990A will connect automatically to all instruments that are not set
to Offline mode in the Test Station Configuration (see Figure 3-4). It is ready for use
once all connections have been initialized successfully and the main menu will
appear as shown in Figure 3-7 .
18
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
3.2.1 Configuring DUT
Once the N5990A main menu appeared, the DUT needs to be configured in order to
proceed with testing. Click on the “Configure DUT” button in the tool bar or select
the “Configure DUT” option from the File menu (see Figure 3-7). A window will
appear as given in Figure 3-8.
Figure 3-7: ValiFrame main window
Keysight N5990A Test Automation Software Platform for PCIe User Guide
19
3
Using Software
Figure 3-8: Configure DUT panel
20
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
The parameter selections available in the “Configure DUT” panel depend on the
specific application. Select all parameters which apply to the particular application to
configure the DUT. The selected DUT parameters and the information entered by the
user will be shown in the measurement reports. It is also stored with the
measurement data in case a connection to a SQL database exists. As this information
will be used to retrieve data from the database, select unique identifiers and
descriptions.
In most applications either Compliance or Expert Mode must be selected. In
compliance mode the tests run according to the specific test specification (such as
PCI Express 3.0). In expert mode the DUT can be characterized to determine
performance margins. It is provided for advanced users and includes additional tests
as well as additional parameters to run tests differently than in compliance mode.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
21
3
Using Software
3.2.2 Selecting, Modifying, & Running Tests
After the DUT has been configured, press the “OK” button in Configure DUT Panel.
The ValiFrame main window is displayed with the procedure tree as shown in Figure
3-9. It contains the list of calibration and test procedures, typically in the following
groups:
1. Calibration
2. Receiver tests
3. Transmitter tests
Figure 3-9: N5990A main window with the procedures
22
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
Specific calibration or test procedures can appear multiple times as they might be
required for testing the DUT under various conditions. A typical example are multiple
data rates supported by the same DUT.
Use the “Properties” and “Log List” buttons of the main menu (highlighted in Figure
3-9) to display additional information on the right side and bottom of the ValiFrame
main window. The parameter grid on the right side of the window shows the
parameters which are related to the selected calibration or test procedure subgroups
or to individual procedures, These parameters can only be set before the execution of
the procedure subgroup or procedure is started. The log list in the bottom of the
window shows calibration and test status messages (regular progress updates as
well as warnings and error messages).
3.2.2.1 System Calibration
It is necessary to calibrate the test system before running the first test, in order to
ensure that test results are consistent from run to run. Provided the equipment has
achieved thermal stability before the calibration is started (typically after 30 min of
warm-up), the thermal environment is stable, and no system elements have been
exchanged, the calibration is very stable and may only have to be repeated once a
week or even less frequently. The calibration interval depends on the degree of
accuracy desired. If the station is not calibrated prior to a DUT test, the results of the
previous calibration will be used for the current tests.
3.2.2.2 Selecting Procedures
The calibration, receiver, and transmitter test procedure groups can be selected
globally by clicking on the check box at the top of the group. Alternatively, an
individual test procedure can be selected by checking the specific selection boxes in
front of the tests. Only the test procedures which are selected will be executed.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
23
3
Using Software
3.2.2.3 Modifying Parameters
Most calibration and test procedures as well as the groups containing them have
parameters that control the details of how the procedures are run. In compliance
mode most of these parameters are read-only. In expert mode almost all parameters
can be modified. First, select a specific calibration or test procedure or one of the
groups containing them in the ValiFrame procedure tree. The parameters should be
displayed in a property list on the right side of the screen. If they are not displayed,
press the “Properties” button. Depending on the user selection on the left side of the
top of the list, the list is either ordered alphabetically or in categories. The test
parameters available can be changed individually (see Figure 3-10). The test
parameters selected are listed in the MS Excel test results worksheets, see Figure
3-12.
Figure 3-10: Editing the test parameters
24
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
3.2.2.4 Running Procedures
To run the selected procedure, press the “Start” button (see Figure 3-9). The
procedures are run in the order shown in the procedure selection tree. Some
procedures may require user interaction such as changing cable connections or
entering DUT parameters. The required action is prompted in pop-up dialog boxes
prior to the execution as shown in Figure 3-11.
Figure 3-11: Connection diagram pop-up window
Keysight N5990A Test Automation Software Platform for PCIe User Guide
25
3
Using Software
3.2.3 Results
3.2.3.1 Run-Time Data Display
Most procedures generate data output. While the procedure is running, the data is
displayed in a temporary MS Excel worksheet or HTML page (depending on the
selected viewer in the Station),which opens automatically for each individual
procedure. An example is given in Figure 3-12 . See the Appendix for more details
about the file directories.
26
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
Figure 3-12: Result MS-Excel worksheet example
The MS-Excel worksheet or the HTML page are opened during the procedure run
and closed once the specific procedure is finished. As long as the N5990A Software
is running, each worksheet or page can be reopened with a double-click on the
respective procedure. However, the individual worksheets or pages will be lost when
the N5990A main window is closed, unless they were saved by the user.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
27
3
Using Software
If a test or calibration procedure was run more than once, the list of results is visible
below the particular procedure after expanding the tree below the procedure (see
Figure 3-13).
Figure 3-13: Selecting the repeated procedure and show test results
28
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
3.2.3.2 Results Workbook
For user convenience, all individual results are combined in a summary MS Excel
workbook or HTML document at the end of the test run. The workbook must be
saved explicitly (File > Save Results as Workbook...) as shown in Figure 3-14,
otherwise it will be lost! After all tests have been run, a test report document can be
generated additionally for easy documentation and printing with the standard Print
function of the File menu (see Figure 3-14). An example test report for PCIe is shown
in Figure 3-15.
Figure 3-14: Save results as workbook
Keysight N5990A Test Automation Software Platform for PCIe User Guide
29
3
Using Software
Figure 3-15: Test report example
30
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
3.2.3.3 Smiley's Representation
Once the selected procedures are run successfully, the smiley at the individual
procedure indicates the result (Pass / Fail / Incomplete) by displaying its face in
specific ways as given below (see Table 1).
Table 1: Smiley's result description table
Smiley
Description
It indicates that the procedure passed successfully at the previous run and the results are available.
It indicates that the procedure passed successfully at present run.
It indicates that the procedure was aborted/disturbed somehow and failed at the previous run.
It indicates that the procedure was aborted/disturbed somehow and failed at the present run.
It indicates that the procedure failed at the previous run.
It indicates that the procedure failed at the present run.
Generally this kind of smiley displays two results such as the first half indicates that the result of the present run and the
second half shows the result of the previous run. In this example, the first half indicates that the procedure passed
successfully at the present run and the second half means that it was not completely run at the previous run.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
31
3
Using Software
3.3
Oscilloscope Transmitter Test Integration
Keysight Technologies provides a range of transmitter test applications for highspeed digital interfaces. The transmitter test applications run on real-time
oscilloscopes of the Keysight 90000 series such as a DSO (Digital Sampling
Oscilloscope). The transmitter test applications can be used stand-alone, without the
N5990A Test Automation Software Platform. For this use model, please refer to the
user documentation of the specific application.
The transmitter test applications however can be run through the N5990A Test
Automation Software Platform too. A remote interface is used to execute the
transmitter test procedures. For this model a test controller PC with the N5990A
software needs to be connected to the oscilloscope via Ethernet, e.g. through a LAN
switch.
3.3.1 Using the Software
In the N5990A Test Station Configuration, the available transmitter test applications
are listed as instruments (see Figure 3-16 for the example of PCIe). The IP address of
the oscilloscope has to be used as the instrument address. After entering the
address, the transmitter test application instrument needs to be set to “Online” with
a click on its check-box. Push the “Check Connections” button to verify that the
connection works properly. If the transmitter test application is not already running
on the oscilloscope the N5990A Test Automation Software automatically starts it via
the oscilloscope firmware.
Figure 3-16: Setting the TX scope application online
32
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
The N5990A Main Window lists the transmitter tests in the test tree just like the
receiver tests, typically in a separate “Transmitter” group.
During the test run, the oscilloscope transmitter test application sends its connection
diagrams and pop-up dialog windows to the controller PC on which the N5990A Test
Automation Software is running. Once the oscilloscope application finished the test
run, the N5990A software will save the test results in a MS Excel worksheet which
includes screen shots, data graphs, data tables and specification limits similar to the
Rx test report.
Figure 3-17: N5990A main window with transmitter tests
Keysight N5990A Test Automation Software Platform for PCIe User Guide
33
3
Using Software
3.3.2 Trouble Shooting
This section provides solutions for the following problems:
1. Wrong version of the transmitter test application
2. Error message at start-up and connection failures
3. Transmitter test application and oscilloscope seem to hang
34
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Using Software
3
3.3.2.1 Wrong version of the transmitter test application
When starting the transmitter tests, the N5990A software compares the version of
the transmitter test application which is currently installed on the oscilloscope with
the version which was tested with N5990A. In case the versions do not match, an
error message will appear in the N5990A log file and a warning dialog will show
details about the latest tested version. The appropriate version of the transmitter test
application should be installed on the oscilloscope to avoid problems. Even if the
versions do not match, the N5990A Test Automation Software can try to run the
transmitter tests. This may work if the changes between the transmitter test
application versions are small, but installing the officially supported version is always
strongly recommended.
3.3.2.2 Error message at start-up and connection fails
The connection to the transmitter test application must be established through
Ethernet (LAN), however the firewall settings might not be set properly on the
oscilloscope or the controller PC. This might result in error messages when the
N5990A Test Automation Software tries to start the oscilloscope transmitter test
application In this case, check whether the following applications are added to the
firewall exception list:
1. Transmitter test application on the oscilloscope
2. N5990A Test Automation Software and N5990A Station Configuration on
the controller PC
In case the controller PC has more than one LAN adapter, the .Net remoting back
channel which displays the dialogs may not work and the oscilloscope application
may try to open the remoting back channel to an invalid address. To recover from
this, the LAN adapter which is connected to the oscilloscope should be set to be the
primary adapter. This might require help from a network administrator as the specific
setting depends on the Windows version.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
35
3
Using Software
If the connection and information dialogs from the oscilloscope are not displayed
properly, check the firewall settings first and then make sure that the LAN adapter
connected to the oscilloscope network is set to the primary one.
3.3.2.3 Transmitter test application and oscilloscope seem to hang
In general the transmitter test application expects a valid signal that can be used as
a trigger for the sampling but sometimes the signal is missing or too small, i.e. below
the threshold.In this case the oscilloscope may appear to be frozen. This is expected
oscilloscope behavior because the oscilloscope trigger hardware stops the execution
of oscilloscope firmware as long as the trigger signal is missing. To exit from this
state, apply a valid signal or reboot the oscilloscope and restart the N5990A software
to check the signals before starting transmitter tests if the required trigger signal is
unknown. Please report the test and test conditions to [email protected]
36
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Keysight N5990A Test Automation Software Platform for PCIe
User Guide
4
4.1
Computer PCI Express
Bus Test Application
4.1
PCI Express (Peripheral Component Interconnect Express) / 37
4.2
PCIe3 Module Procedure Descriptions / 43
PCI Express (Peripheral Component Interconnect Express)
4.1.1 Supported Hardware Configurations
For PCI Express testing, N5990A supports hardware configurations based on the
Keysight J-BERT N4903B and Keysight J-BERT M8020A.
4.1.2 Configuring a PCI Express Station
Follow the instructions given in Section 3.1. An example test station configuration
window is given in Figure 4-1.
4
Computer PCI Express Bus Test Application
Figure 4-1: PCI-Express station configuration window
38
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.1.2.1 PCI Express Station Configuration Window Options
4.1.2.1.1
Data Generator
The data generator is used to create patterns with specified stress parameters. The
following instruments can be selected as data generator:

JBERT-B (Keysight N4903B High Perfomance Serial BERT)

JBERT-M8020A
The selected generator will also be used as error detector to check if the data looped
back from the DUT contains errors.
Figure 4-2: Configuration with JBERT-N4903B
Keysight N5990A Test Automation Software Platform for PCIe User Guide
39
4
Computer PCI Express Bus Test Application
Figure 4-3: Configuration with JBERT M8020A
4.1.2.1.2
Main Power Switch Control
It can be selected as:

Manual


Netlo 230 B (It is a PDU (Power Distribution Unit) that integrates one 230 V
input and four 230 V outlets which allow to connect virtually any 230 V
powered device)
SynAccessNP
If it is chosen as Manual, the user needs to power cycle the DUT manually. When it is
selected as NetIo230B or SynaccessNP the DUT is power cycled automatically.
In order to use the Power Switch, the ValiFrame option 008: Remote Power
Management Support is required
For more details please refer to the Appendix section Main Power Switch Control.
4.1.2.1.3
Use ext. 100 MHz Reference Clock Source
An external 100MHz reference clock source can be used as clock source for the data
generator and the DUT to obtain a constant clock signal. It is only necessary in
Common Reference Clock architecture for Add in Card and ASIC DUTs; in Systems
the reference clock is provided by the DUT.
40
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.1.2.1.4
4
Reference Clock Multiplier (only for JBERT-B)
The reference clock can be multiplied to the datarate internally in the JBERT-B or
externally with the N4880A. The options are:

None: In this case, the JBERT-B will provide the clock to the DUT. This
option is only valid for AddIn Card or ASIC DUTs. For System DUTs the
Internal PLL will be used instead. This setting cannot be selected if “Use ext.
100 MHz Reference Clock Source” is enabled.

N4880A (Keysight N4880A Reference Clock Multiplier)

4.1.2.1.5
Internal: Internal multiplication can only be used if the reference clock is
clean without SSC because of the low loop bandwith of the internal
multiplying PLL.
Common Mode Interference Source (only for JBERT-B)
It is required only for the DUT type ASIC. For CEM tests a CMSI (Common Mode
Sinusoidal Interference) source is not required, so the CMSI source can be selected
as “None”. The options available to select as the CMSI are:

Keysight 81150A Pulse Function Arbitrary Waveform Generator

None
4.1.2.1.6
Use Switch for Tx Test
If selected, Tx tests will include a switch to test more than one lane without changing
connections.
4.1.2.1.7
Use Switch for Rx Test
If selected, Rx tests will include a switch to test more than one lane without changing
connections.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
41
4
Computer PCI Express Bus Test Application
4.1.3 Configuring a PCI Express DUT
For PCI Express 1.0 a, 1.1, 2.0, and 3.0 the following DUT types are supported, and
accessories are recommended (see Figure 4-4):

Add-In Card
o Compliance Base Board (CBB) available from PCISIG

System
o Compliance Load Board (CLB) available from PCISIG

ASIC
o Custom DUT load board
The PCI-Express DUT configuration panel is shown in Figure 4-4. Select the DUT
type as mentioned above, the other options such as Spec version and Clock
Architecture can also be selected. As for all applications, the descriptions and
comments along with user name can be entered, time stamps are shown. As well as
the “Compliance Mode” and “Expert Mode” is also selected. The latter is for userdefined characterization, margin tests, and debugging.
42
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-4: PCI-Express configure DUT panel
4.2
PCIe3 Module Procedure Descriptions
This section describes the test procedures and calibration of the PCIe3 module based
on version 1.40.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
43
4
Computer PCI Express Bus Test Application
4.2.1 Configure DUT
Figure 4-5: Configure DUT window
44
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.2.1.1 Product Parameters:
DUT Type:
The DUT type can be chosen as:

Add-In Card: for example, a graphics card is tested according to the CEM
specification

System: for example, a mother board is tested according to the CEM
specification

ASIC: for example, a PCIe chip is tested according to the Base
specification
Version:
The available PCI Express specification versions available are:

1.0a: supports 2.5 GT/s data rate.

1.1: supports 2.5 GT/s data rate.

2.0: supports 2.5 GT/s and 5.0 GT/s data rates.

3.0: supports 2.5 GT/s, 5.0 GT/s and 8 GT/s data rates.
Spec Version:
The calibrations and tests are defined according to the spec version selected.
Clock Architecture:
It can be selected as:

Common Ref Clock: The default clock architecture where all parts of the
system use the same clock.

SRNS (Separate Ref Clks No Ssc)

SRIS (Separate Ref Clks Independent Ssc)
Keysight N5990A Test Automation Software Platform for PCIe User Guide
45
4
Computer PCI Express Bus Test Application
4.2.1.2 Test Parameters
Compliance/Expert Mode:
In Compliance mode only the plain compliance tests are shown with very little
customization parameters. In Expert mode additional debugging tests are added and
the compliance tests can be run with customized settings.
Pressing the “Show Parameters” button it is possible to select more test parameters:
Figure 4-6: PCI express ASIC parameters
There are some parameters that are common for all data rates:
46
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.1.2.1
4
Generator SSC
Figure 4-7: Generator SSC
It is only possible to enable the Generator SSC for SRIS (Separate Ref Clks
Independent Ssc) architecture. In that case the following parameters can be set:

SSC Frequency

SSC Deviation
4.2.1.2.2
Power Switch Automation
Figure 4-8: Power switch automation
If a Power Switch is used, ValiFrame will power on/off automatically the DUT and the
loopback training will run without user interaction.

Channel: Number of the switch channel to use.

Off-On Duration: Time span between power off and power on the switch.

Settling Time: Time span after power on.

Max. Retries for LB training: Maximum number of times that ValiFrame will
try to train the DUT into loopback mode. If it is not possible within these
tries the test will be aborted automatically. When Power Switch Automation
is unselected, ValiFrame asks the user to retry every time loopback fails.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
47
4
Computer PCI Express Bus Test Application
4.2.1.2.3
External Reference Clock
An external reference clock it is only necessary in Common Reference Clock
architecture for AddIn Card and ASIC DUTs. This option must be previously selected
in the Station Configuration to be visible in the Test Parameters Dialog.
Figure 4-9: External reference clock ref clock multiplier internal


Use External 100 MHz Reference Clock: If this is selected, an external
source will provide the clock to the DUT and the data generator. If it is not
selected, the clock output of the data generator will be directly connected
to the DUT.
Ref Clock Multiplier: When using an external reference clock there are two
options to connect the external clock to the data generator:
o
Internal: Just connect both directly and the data generator will
use his internal PLL.
o
N4880A: Use an external clock multiplier. (only for
J-BERT-B)
For System setups the reference clock is provided by the DUT. It can be multiplied
internally in the data generator or with the external N4880A clock multiplier (only for
J-BERT N4903B).
Figure 4-10: External reference clock ref clock multiplier N4880A
48
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.1.2.4
4
BER Reader
Figure 4-11: BER reader J-BERT-B analyser
The error measurement can be done using the J-BERT N4903B/M8020A error
detector or with an external BER reader.
There are other parameters that need to be selected independently for each data
rate:
4.2.1.2.5
Channels
Figure 4-12: Channels PCIe1 M8048A ISI channel Ch7 24inch


PCIe1 J-BERT-B ISI Channel, PCIe2 J-BERT-B ISI Channel, PCIe1 M8048A
ISI Channel, PCIe2 M8048A ISI Channel: For 2.5GT/S and 5.0GT/s ASICs,
Inter Symbol Interference is added to the signal.
Emulate ISI Channel. If the M8020A J-BERT option M8041A-0G5 to
generate ISI internally is available, selecting the checkbox will allow
ValiFrame to use it.
Figure 4-13: Channels 3.5 dB link and 6.0 dB link

3.5 dB Link/6.0 dB Link: For System 5.0GT/s setups a 3.5dB Link, a 6.0dB
Link, or both can be tested.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
49
4
Computer PCI Express Bus Test Application
Figure 4-14: Channels CBB Rev. 3 riser card

CBB rev.2/CBB rev.3 riser card: For CEM 8.0GT/s DUTs the Rx tests must
be done with compliance base board gen3, but they can also be done with
the CBB gen2 additionally.
Figure 4-15: Without channel short channel long channel

50
Without Channel, Short Channel, Long Channel: For ASICs 8.0GT/s, in
order to emulate different target applications, three test cases with different
channel lengths are defined. For a JBERT-N4903B setup, only the N4915A014 ISI channels box can be selected. For a JBERT-M8020A setup, the
previous or the M8048A-002 box can be selected. If the Option M8041A0G5 is available, it can be selected too.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.1.2.6
4
Calibration
4.2.1.2.6.1 For CEM and ASIC 2.5GT/s-5.0GT/s
Figure 4-16: Calibration for CEM and ASIC 2.5 GT/s – 5.0 GT/s



Scope Connection: It can be differential (channel 1, 2, 3 or 4) or direct
connect (channels 1-3 or 2-4). The default is direct connect channels 1-3
when using M8020A and differential channel 1 when using J-BERT N4903B.
Use Transfer Function: Calibration boards or fixtures or additional cables
can be embedded or de-embedded using the transfer function in the scope.
Transfer Function on Scope: The transfer function file on the scope to
embed or deembed additional components.
4.2.1.2.6.2 For CEM 8.0 GT/s
Figure 4-17: Calibration for CEM 8.0 GT/s with embedded replica channel



Scope Connection: It can be differential (channel 1, 2, 3 or 4) or direct
connect (channels 1-3 or 2-4). The default is direct connect channels 1-3
when using M8020A and differential channel 1 when using J-BERT N4903B.
Use Transfer Function: calibrations boards or fixtures or additional cables
can be by embedded or de-embedded with its transfer function in the
scope.
Transfer Function on Scope: The transfer function file on the scope to
embed or deembed additional components.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
51
4
Computer PCI Express Bus Test Application
4.2.1.2.6.3 For ASIC 8.0 GT/s
Figure 4-18: Calibration for ASIC 8.0 GT/s
Figure 4-19: Calibration for ASIC 8.0 GT/s with embedded replica channel







52
Embed Replica Channel: If true, use a tranfer funtion to embed the replica
channel.
Package Model on Scope: The transfer function file of the package model.
Replica Ch. + Package Model on Scope: The transfer function file that
combines the package model and replica channel, if replica channel is
embedded.
Replica Channel Model on scope: The transfer function file of the replica
channel.
Step Low Time: This is the time span from the falling edge of the step to the
end.
Step High Time: This is the time span from the rising edge of the step to the
end.
Number of averages for step response: The number of times that the step
response is averaged to minimize the noise.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.1.2.7
4
Receiver
Figure 4-20: Receiver




PCIe3 Link Training Mode:
o
J-BERT_Loopback_Training: J-BERT sends a loopback training
sequence to the DUT.
o
Vendor_Specific: The vendor can place his DUT in loopback with
his own tools like JTEC, I2C…
PCIe3 Link Training Suite Setting File: The Link training suite setting file
(script file) which will be used for loopback training.
PCIe3 Link Training Suite Lane Number: The lane number which will be
encoded in the TS1s/TS2s sent to the DUT during loopback training. If Auto
each lane will be encoded with its own number.
Relax Time: Time span between the point the stress signal is changed and
the BER measurement begins.
Error Detector

Use CDR: If true, CDR is used to generate a clock signal for the J-BERT
error detector. If false, the clock is supplied by the J-BERT data generator.
With CDR the J-BERT error detector performance is better.

Loop Bandwidth: The loop bandwidth of the J-BERT error detector CDR.

Enable SSC tracking: Enable SSC tracking if DUT transmits with SSC. Only
for J-BERT N4903B.

Expected Deviation: Select the expected deviation the DUT has. Only for JBERT N4903B.

Peaking. Only for J-BERT M8020A.

Analyzer Equalization. Only for J-BERT M8020A.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
53
4
Computer PCI Express Bus Test Application
4.2.1.2.8
Transmitter
There are specific parameters for Tx tests:
Figure 4-21: PCI express ASIC parameters
54
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.1.2.9
4
Lanes Configuration
To select which lanes are going to be tested, press the button “Lanes Configuration”
in the Configure Dialog.
Figure 4-22: Lanes configuration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
55
4
Computer PCI Express Bus Test Application
Lanes for Rx Test:
To select which receiver lanes are going to be tested. Depending on the option
chosen in the station configuration more or less lanes can be selected:

If Use Switch for Rx test is unselected: All lanes can be tested. During the
tests the user must switch the cables from lane to lane manually.

If Use Switch for Rx test is checked and at least one module is type SPT4:
only four lanes can be tested.

If Use Switch for Rx test is selected and all modules are type SPT6: six lanes
can be selected.
Lanes for Calibration:

If Use Lane0 calibration for all lanes in Rx tests is selected, only lane 0 has
to be calibrated and its results will be used in the Rx tests for all lanes.

If Calibrate all Rx lanes is selected, it will be necessary to run the
calibrations for each Rx lane checked. Then each Rx test will use the
specific calibrations of the lane to test.
Lanes for Tx Tests:
To select which transmitter lanes are going to be tested. Depending on the option
chosen in the station configuration more or less lanes can be selected:

If Use Switch for Tx test is unselected: All lanes can be tested. During the
tests the user must switch the cables from lane to lane manually.

If Use Switch for Tx test is checked and module type is SPT4: only lanes 0 to
3 can be tested.

If Use Switch for Tx test is selected and all modules are type SPT6: only
lanes from 0 to 5 can be selected.
56
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.2.2 Gen1 & Gen2 ASIC Tests
4.2.2.1 Calibration
4.2.2.1.1
Common Calibration Parameters:
Scope Connection:
All calibrations can be done with either a differential probe or a single ended direct
connection on two channels, with one exception: The Common Mode Optimization
can only be done with a differential probe.
Jitter Unit:
All the jitter parameters can be displayed in time or in unit interval. The graphics
results will be represented in the chosen unit.
PCIe1 M8048A ISI Channel:
The M8048A ISI Channel used for PCIe1 ASIC Rx Calibration.
PCIe2 M8048A ISI Channel:
The M8048A ISI Channel used for PCIe2 ASIC Rx Calibration.
PCIe1 J-BERT-B ISI Channel:
The J-BERT-B ISI Channel used for PCIe1 ASIC Rx Calibration.
PCIe2 J-BERT-B ISI Channel:
The J-BERT-B ISI Channel used for PCIe2 ASIC Rx Calibration.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
57
4
Computer PCI Express Bus Test Application
4.2.2.1.2
Random Jitter Calibration
Purpose and Method:
In the Rx tests the input signal will be stressed with a combination of jitter sources to
simulate the possible impairments expected at the Rx input when operating in a
target system. Random jitter is added to simulate the effects of thermal noise. Due to
system intrinsic jitter the effective jitter level is different from the value set in the data
generator, so the jitter amplitude is calibrated.
The test automation calibrates six equally spaced RJ values (from 0 to 10ps). The JBERT sends a clock pattern during this calibration procedure. The actual jitter is
measured on a DSO using the RJ/DJ-separation software EzJIT.
The calibration data is stored in a caltable
(Pcie<generation>_ASIC_RandomJitter.txt). There is a caltable for gen1 ASIC and
one for gen2 ASIC. For the measurements these calibration tables will be used to
calculate the RJ amplitude that needs to be set on the generator to get the desired
RJ amplitude at the test point.
Figure 4-23: Connection diagram M8020A
58
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-24: Connection diagram J-BERT-B
Parameters in Expert Mode: none
Used Calibrations: none
Keysight N5990A Test Automation Software Platform for PCIe User Guide
59
4
Computer PCI Express Bus Test Application
Figure 4-25: Result description
Column 1 Set Jitter: The jitter amplitude set in the instrument
Column 2 Actual Jitter: The measured jitter amplitude
60
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.2.1.3
4
ISI Calibration
Purpose and Method:
In ASIC Rx tests, Inter Symbol Interference is generated to provide a close to real
environment. Due to system intrinsic jitter the effective jitter level is different from
the value set in the data generator, so the jitter amplitude is calibrated.
The test automation calibrates the ISI traces number 0, 1, 2, and 3. ISI is injected
routing the signal though J-BERT-B opt 020 traces. The actual value is calculated as
the difference between the eye width when the J-BERT-B sends a clock pattern and
the eye width when it sends the compliance pattern. The eye width is measured with
a DSO using horizontal histograms.
The calibration data is stored in a caltable (Pcie<generation>_ASIC_ISI.txt). There is a
caltable for gen1 ASIC and one for gen2 ASIC. For the measurements these
calibration tables will be used to display the ISI amplitude.
Connection Diagram: The same as for “Random Jitter Calibration”
Parameters in Expert Mode: none
Used Calibrations: none
Figure 4-26: Result description
Column 1 Measured ISI: The measured ISI.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
61
4
Computer PCI Express Bus Test Application
4.2.2.1.4
CM Optimization
Purpose and Method:
Common Mode Sinusoidal Interference is generated by coupling two sine waves with
the same phase and amplitude to Normal and Complement of the test signal. If there
is a skew between the two channels or the amplitudes are not equal some amount of
the CM signal will transfer to Differential Mode which is not wanted. This calibration
minimizes this effect by adjusting the delay and the amplitude correction factor
between ch1 and ch2 of the 81150A function generator.
The differential amplitude between the channels of the 81150A is measured with a
DSO for several different delay and amplitude correction factor values. The channels
are perfectly adjusted if the differential amplitude between them is 0. There are
several steps to find the optimal delay and amplitude corrector factor.
The first step sets delays from -12 ns to 12 ns and measures the differential
amplitudes. It uses a linear search algorithm to look for the delay that generates the
minimum differential amplitude, “delay1”. The second step uses a binary search
algorithm to find the optimum delay in the range “delay1” ±100ps, “delay2”. In the
next step delay is set to “delay2” and a binary search algorithm is used to find the
optimum amplitude corrector factor in the range 0.9 to 1.1. Then this optimal
amplitude corrector factor is set and a binary search algorithm is used again to find
the optimum delay in the range “delay2”±10ps. The last step is to verify that the
differential amplitude between the channels is below the defined maximum (7.5 mV
for gen 1, 5 mV for gen2) with the optimal delay and amplitude corrector factor
values.
The calibration data is stored in a caltable (Pcie<generation>_ASIC_CMResidual.txt).
There is a caltable for gen1 ASIC and one for gen2 ASIC. For the measurements
these calibration tables will be used for the amplitude correction and skew
adjustment of the 81150A function generator.
Available for following configuration:
Data Generator: J-BERT N4903B
Connection Diagram: The same as for “Random Jitter Calibration”
Parameters in Expert Mode:
Number of Averages: To reduce the noise floor the differential signal has to be
averaged when measured with the scope.
Used Calibrations: none
62
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-27: Result description
Column 1:
Delay Ch1: The optimum delay between channel 1 and channel 2 of the 81150A
Column 2:
Amplitude Correction Factor Ch1: Amplitude Correction Factor for channel 1.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
63
4
Computer PCI Express Bus Test Application
4.2.2.1.5
CM Sinusoidal Interference Calibration
Purpose and Method:
Common Mode Sinusoidal Interference (CMSI) is generated with a 81150A function
generator (s. CM Optimization). It is coupled into to the data signal with pick off tees,
so the resulting amplitude at the input of the Rx is attenuated and needs to be
calibrated.
First the output amplitude of the 81150A is set to 1V. The resulting amplitude is
measured and used to calculate a scale factor (set amplitude / measured amplitude).
This scale factor is used to calculate the necessary input value to have an output
signal of 300 mV (the level used in Rx tests), EndVoltage. Then CMSI is calibrated
from 0 to EndValue + 500mV with steps of 500mV. In every step the CMSI is
generated with the 81150A and measured with a DSO. When the scope connection is
differential the amplitude is only set in one of the channels of the 81150A. When the
scope connection is single ended the amplitude is set in both channels.
The calibration data is stored in a caltable
(Pcie<generation>_ASIC_CmInterference.txt). There is a caltable for gen1 ASIC and
one for gen2 ASIC. For the measurements these calibration tables will be used to
adjust the voltage amplitude to the desired output CMSI.
Connection Diagram: The same as for “Random Jitter Calibration”
Parameters in Expert Mode:
Number of Averages: To reduce the noise floor the differential signal has to averaged
when measured with the scope.
Used Calibrations: none
64
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-28: Result description
Column 1: Set CM SI Amplitude: CM SI set at the 81150A
Column 2: Measured CM SI Amplitude: The measured jitter amplitude
Keysight N5990A Test Automation Software Platform for PCIe User Guide
65
4
Computer PCI Express Bus Test Application
4.2.2.1.6
Eye Height Calibration
Purpose and Method:
The test fixtures attenuate the data signal. To compensate for this, the data signal
differential swing is calibrated.
The test automation calibrates five equally spaced differential voltage amplitudes.
The minimum amplitude is 100mV and the maximum is the maximum value that the
data generator can generate.
For this calibration the data generator sends the compliance pattern. It also adds
random jitter, ISI, swept sinusoidal jitter and SSC residual to the signal. The height is
measured in the scope using horizontal histograms.
The calibration data is stored in a caltable (Pcie<generation>_ASIC_EyeHeight.txt).
There is a caltable for gen1 ASIC and one for gen2 ASIC. For the measurements
these calibration tables will be used to adjust the differential voltage amplitude to
the desired eye height.
Connection Diagram: The same as for “Random Jitter Calibration”
Parameters in Expert Mode: none
Used Calibrations:

Random Jitter Calibration

Isi Calibration

CM Sinusoidal Interference
66
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-29: Result description
Column 1
Set Diff Voltage: The differential voltage amplitude set in the instrument
Column 2
Measured Eye Height: The measured eye height amplitude
Keysight N5990A Test Automation Software Platform for PCIe User Guide
67
4
Computer PCI Express Bus Test Application
4.2.2.1.7
Eye Height Verification
Purpose and Method:
This procedure checks if it is possible to generate a signal with an eye height value
that is inside a range of +-10% tolerance with respect to the eye height target value.
The target eye height is the minimum value specified in the CTS.
For this verification the data is generator sends the compliance pattern. It also adds
random jitter, ISI, HF sinusoidal jitter and SSC residual to the signal CMSI. The height
is measured in the scope using horizontal histograms.
Connection Diagram: The same as for “Random Jitter Calibration”
Parameters in Expert Mode: none
Used Calibrations:

Random Jitter Calibration

Isi Calibration

CM Sinusoidal Interference
Figure 4-30: Result description
Column 1
Set Diff Voltage: The differential voltage amplitude set in the instrument
Column 2
Measured Eye Height: The measured eye height amplitude
68
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.2.2.2 Receiver
4.2.2.2.1
Common Receiver Parameters
Data Rate Specific:
1
Data rate deviation

Deviation in ppm that will be applied to the generator data rate.
2
Link Training

Link Training Mode:
o
JBERT _Link_Training: J-BERT sends a loopback training
sequence to the DUT.
o
Vendor_Specific: The vendor can place his DUT in
loopback
with his own tools like JTEC, I2C…

Link Training Suite Setting File:
The Link Training Suite settings file (script file) which will be used for
loopback training.

Default Link Training Lane Number for every Lane:
The lane number which will be encoded in the TS1s/TS2s sent to the
DUT during loopback training. This value will be set by default for all
lanes. If Auto each lane will be encoded with its own number.

Suppress Loopback Training Messages:
When set to true this hides all popup messages related to loopback
training.
3
Error Detector

Use CDR: If true, CDR is used to generate a clock signal for the JBERT error detector. If false, the clock is supplied by the J-BERT data
generator. With CDR the J-BERT error detector performance is better.



Loop Bandwidth: The loop bandwidth of the J-BERT error detector
CDR.
Enable SSC tracking: Enable SSC tracking if DUT transmits with SSC.
Only for J-BERT N4903B.
Filter Gen1/Gen2 SKPOs for BER test: Ignore SKP symbol in the error
detector. Only for J-BERT N4903B.
BER Measurement

Relax Time: Time span between the point the stress signal is changed and
the BER measurement begins.
Lane Specific

Link Training Lane Number: The lane number which will be encoded in the
TS1/TS2s sent to the DUT during loopback training.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
69
4
Computer PCI Express Bus Test Application
4.2.2.2.2
Receiver Compliance
Purpose and Method:
This test determines if the DUT meets the receiver specifications. The procedure
measures the BER when all jitter types and the eye height are set to their spec limit
values (maximum for jitter, minimum for eye height). In expert mode these values can
be changed.
Figure 4-31: Connection diagram M8020A common clock architecture no external ref clock
70
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-32: Connection diagram M8020A common clock architecture external ref clock
Keysight N5990A Test Automation Software Platform for PCIe User Guide
71
4
Computer PCI Express Bus Test Application
Figure 4-33: Connection diagram M8020A separated clock architecture
72
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-34: Connection diagram J-BERT-B common clock architecture no external ref clock
Keysight N5990A Test Automation Software Platform for PCIe User Guide
73
4
Computer PCI Express Bus Test Application
Figure 4-35: Connection diagram J-BERT-B, common clock architecture external clock internal clock multiplier
74
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-36: Connection diagram J-BERT-B, common clock architecture external ref clock N4480A clock multiplier
Keysight N5990A Test Automation Software Platform for PCIe User Guide
75
4
Computer PCI Express Bus Test Application
Figure 4-37: Connection diagram J-BERT-B separated clock architecture
Parameters in Expert Mode:
1
2
76
Compliance Test

ISI: The amount of ISI introduced by the selected trace.

CMSI Frequency: Frequency of the Common Mode Sinusoidal
Interference.

CMSI Amplitude: Common Mode Sinusoidal Interference
added to the signal.
amplitude
Generator Jitter
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application



4
Random Jitter: The amount of random jitter (rms) added to the test
signal.
Swept sinusoidal Jitter: The amplitude of the sinusoidal jitter
component that is swept continuously from 1.5 to 100MHz during the
test.
SSC Residual (only for 5.0Gbit/s): The SSC Residual emulates the
residual which is caused by path length differences in the clock
distribution and SSC modulation in real world systems. The residual
SSC is triangular. It should be >= 75 ps.
3
Eye Height

Eye Height: The eye height used in this procedure.
4
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.

Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if
no bit error occurs.

Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.


Used Calibrations:

Random Jitter Calibration

ISI Calibration

CM Sinusoidal Interference Calibration

Eye Height Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
77
4
Computer PCI Express Bus Test Application
Figure 4-38: Result description
78
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:

Pass/Fail Result: The BER measured should be smaller than the Target
BER.
Column 2:

BER: Measured BER.
4.2.2.2.3
Receiver Jitter Tolerance
Purpose and Method:
This test procedure searches the maximum sinusoidal jitter at which the DUT passes
the BER test. The test first uses relatively large steps to go linearly from “Start Jitter”
up. When BER test fails it goes back down with mid-sized steps until it passes again.
From there it steps up again with small steps until an error is found again. The
maximum passed value is the last test point that did not return an error. All of this
happens separately for each frequency.
Connection Diagram:
The same as for “Receiver Compliance”
Parameters in Expert Mode:
1
Jitter Tolerance Test

ISI Trace: The ISI trace to set.

ISI: The amount of ISI introduced by the selected trace.

CMSI Frequency: Frequency of the Common Mode Sinusoidal
Interference.

CMSI Amplitude: Common Mode Sinusoidal Interference amplitude
added to the signal.
2
Sinusoidal Jitter Variation

Frequency Mode: Specifies the distribution of the frequency points to
test. It can be:
o
Compliance Frequencies: The frequencies defined in the
specification for compliance testing.
o
Equally Spaced Frequencies.
o
User Defined Frequencies.
o
Single Frequency.

Frequency Scale:
The results can be represented in logarithmic or linear scaling. For
“Equally Spaced Frequencies” the frequency points depend on the
chosen frequency scale.

Start Frequency:
Start frequency for “Equally Spaced Frequencies”.

Stop Frequency:
Stop frequency for “Equally Spaced Frequencies”.

Number of Frequency Steps:
Number of frequency points for “Equally Spaced Frequencies”.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
79
4
Computer PCI Express Bus Test Application
Search Algorithm (HyteresisUp or Binary):
The Binary search is not recommended for
devices with a long recovery time.





3
Parameter

Force Retraining on Each Frequency:
Re-train the DUT at each tested jitter frequency.
4
Generator Jitter

Random Jitter:
The amount of random jitter (rms) added to the test signal.
5
80
Frequency Points:
Frequency points for “User Defined Frequencies”.
Start Jitter Points:
Start jitter amplitudes for “User Defined Frequencies”. In the
other frequency modes start jitter amplitude is always 0.
Jitter Frequency: Single frequency point for “Single Frequency”.
Jitter Step Size:
The size of the smallest sinusoidal jitter amplitude step used to search
the “Max passed jitter” at each frequency.
Show Min Failed Points:
The results graph can show the minimum failed jitter in
addition to the maximumpassed jitter for each tested frequency.

HF Sinusoidal Jitter (only for 5.0Gbit/s):
The amout of high frequency sinusoidal jitter added to
the test signal. It should be >= 27ps.

HF Sinusoidal Jitter Frequency (only for 5.0Gbit/s):
Frequency of the HF sinusoidal jitter component.

SSC Residual (only for 5.0Gbit/s):
The SSC Residual emulates the residual which is caused by
path length differences in the clock distribution and SSC
modulation in real world systems. The residual SSC is triangular.
It should be >= 75 ps.
Eye Height

Eye Height:
The eye height used in this procedure.

Confidence Level:
The confidence level for the BER measurement in this test.

Target BER:
The Target BER for the BER measurement used in the test. For
a Target BER of 1E-12 and a Confidence Level of 95% the test
will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Used Calibrations:

Random Jitter Calibration

ISI Calibration

CM Sinusoidal Interference Calibration

Eye Height Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
81
4
Computer PCI Express Bus Test Application
Figure 4-39: Result description
82
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:
Pass/Fail Result: Max passed jitter should be bigger than min spec.
Column 2:
Sinusoidal Jitter Frequency: Tested frequency points.
Column 3:
Max Passed LF Deterministic Jitter: Max jitter that passed the test.
Column 4:
Jitter Capability Test Setup: Max jitter that the hardware can generate.
Column 5:
Min Spec: Min jitter that has to pass the test to meet the specification.
Column 6:
Margin: Margin between the max passed jitter and the specification.
4.2.2.2.4
Receiver Sensitivity
Purpose and Method:
This test searches the minimum eye height at which the DUT passes the BER test.
The method starts with “Start Eye Height” and decreases with steps of “Step Size”.
The minimum passed value is the last test point that did not return an error.
Connection Diagram: The same as for “Receiver Compliance”
Parameters in Expert Mode:
1
Sensitivity Test

ISI: The amount of ISI introduced by the selected trace.

CMSI Frequency: Frequency of the Common Mode Sinusoidal
Interference.

CMSI Amplitude: Common Mode Sinusoidal Interference amplitude
added to the signal.
2
Generator Jitter

Use Jitter: If this is enabled, jitter is added to the test signal.

Random Jitter: The amount of random jitter (rms) added to the test
signal.

Swept sinusoidal Jitter: The amplitude of the sinusoidal jitter
component that is swept continuously from 1.5 to 100MHz during the
test.
HF Sinusoidal Jitter (only for 5.0Gbit/s): The amount of high frequency
sinusoidal jitter added to the test signal. It should be >= 27ps.
HF Sinusoidal Jitter Frequency (only for 5.0Gbit/s): Frequency of the
HF sinusoidal jitter component.
SSC Residual (only for 5.0Gbit/s): The SSC Residual emulates the
residual which is caused by path length differences in the clock
distribution and SSC modulation in real world systems. The residual
SSC is triangular. It should be >= 75 ps.



Keysight N5990A Test Automation Software Platform for PCIe User Guide
83
4
Computer PCI Express Bus Test Application
3
Eye Height




4
Loopback Training Eye Height: The eye height used for loopback
training.
Start Eye Height: The eye height (transitions bits) where the test
starts.
Stop Eye Height: The eye height (transitions bits) where the test stops.
Step Size: The amount the eye height is decreased each step to search
the “Min Passed Eye Height”.
BER Measurement





BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if
no bit error occurs.
Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Random Jitter Calibration

ISI Calibration

CM Sinusoidal Interference Calibration

Eye Height Calibration
84
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-40: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
85
4
Computer PCI Express Bus Test Application
Column1:
Pass/Fail Result: The Min Passed Eye Height measured should be smaller than the
Min Spec.
Column 2:
Min Passed Eye Height: The smallest eye height at which the DUT passes the BER
test.
Column 3:
Min Spec: The smallest eye height at which the DUT has to pass the BER test to
meet the specification.
Column 4:
Margin: Margin between the Min Passed Eye Height and the Min Spec.
86
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.2.3 Gen3 ASIC Tests
4.2.3.1 Calibration
4.2.3.1.1
Common Calibration Parameters
Scope Connection:
All calibrations can be done with either a differential probe or a single ended direct
connection on two channels, with one exception: The Common Mode Optimization
can only be done with a differential probe.
Transfer Function File for Package Model on Scope:
This is the path to the package model which is located on the oscilloscope. The
package model has to be embedded for some calibrations. The file cannot be copied
to the scope by the automation software. If the file is not on the oscilloscope the
automation software will instruct the user to copy the package model file manually.
Transfer Funtion File for Replica on Scope:
This is only available if “Emded Replica Channel” was selected in “Configure DUT>Show Parameters->Rx 8.0GT/s”.
This is the path to the transfer function of the replica channel which is located on the
oscilloscope.
Transfer Funtion File for Replica and Package Model on Scope:
This is only available if “Emded Replica Channel” was selected in “Configure DUT>Show Parameters->Rx 8.0GT/s” .
This is the path to the transfer function which includes replica channel and package
model which is located on the oscilloscope.
Step Response Low Time:
The length before the low to high transition in UIs.
Step Response High Time:
The length after the low to high transition in UIs.
Number of Averages for Step Response:
The number of averages for the step response.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
87
4
Computer PCI Express Bus Test Application
4.2.3.1.2
Generator Launch Voltage Calibration
Purpose and Method:
This procedure calibrates the DC amplitude right behind the couplers which add the
DM and CM interference to the test signal (TP1).
This calibration has to be done with a test pattern consisting of 128 ones followed by
128 zeros at 8Gbps. The calibration is done with 3 de-emphasis settings, 0, 1 and
2 dB. The Launch Voltage should be initially calibrated to 800mV for all other
calibrations. It is allowed to change the Launch Voltage later to adjust the eye height
or eye width.
For the short and long channel the voltage cannot be measured directly at TP1
because it is inside a box and thus not accessible. In that case the voltage is
measured at the end of the channel and the voltage at TP1 is calculated over the DC
resistance of the channel.
The calibration data is stored in a caltable (Pcie3Asic_<channel>_LaunchVoltage.txt).
For the measurements these calibration tables will be used to adjust the differential
voltage amplitude to the desired output value.
88
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-41: Connection diagram M8020A no channel
Figure 4-42: Connection diagram M8020A short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
89
4
Computer PCI Express Bus Test Application
Figure 4-43: Connection diagram M8020A long channel
90
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-44: Connection diagram J-BERT-B no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
91
4
Computer PCI Express Bus Test Application
Figure 4-45: Connection diagram J-BERT-B short channel
92
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-46: Connection diagram J-BERT-B long channel
Parameters in Expert Mode:
DC Channel Resistance: The DC resistance of the channel. It is used to calculate the
launch voltage on the virtual test point TP1 which is after the DM/CM Coupler and
before the short or long channel.
Used Calibrations: none
Keysight N5990A Test Automation Software Platform for PCIe User Guide
93
4
Computer PCI Express Bus Test Application
Figure 4-47: Result description
94
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:
Set Generator Voltage: The generator voltage set at the N4916B de-emphasis signal
converter output.
Column2:
Measured Voltage with 0dB DE: The “DC” amplitude measured with the scope at 0dB
de-emphasis.
Column3:
Measured Voltage with 1dB DE: The “DC” amplitude measured with the scope at 1dB
de-emphasis.
Column4:
Measured Voltage with 2dB DE: The “DC” amplitude measured with the scope at 2dB
de-emphasis.
4.2.3.1.3
Insertion Loss Calibration
Purpose and Method:
The Insertion Loss of the calibration channels + the replica channel has to be in a
well defined range. This calibration calculates the Insertion Loss from the step
response at 3 different de-emphasis levels. By adding de-emphasis IL can be
reduced to a certain degree. This is used to compensate IL during the Rx tests.
For every de-emphasis level the insertion loss is measured from 1GHz to 4GHz with
steps of 250MHz. The IL is measured using the Seasim software.
The calibration data is stored in a caltable (Pcie3Asic_<channel>_IL). This cal data is
used to evaluate the optimum amount of de-emphasis for the Rx tests.
Available for following hardware configurations: all
Keysight N5990A Test Automation Software Platform for PCIe User Guide
95
4
Computer PCI Express Bus Test Application
Figure 4-48: Connection diagram M8020A no channel
96
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-49: Connection diagram M8020A short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
97
4
Computer PCI Express Bus Test Application
Figure 4-50: Connection diagram M8020A long channel
98
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-51: Connection diagram J-BERT-B no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
99
4
Computer PCI Express Bus Test Application
Figure 4-52: Connection diagram J-BERT-B short channel
100
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-53: Connection diagram J-BERT-B long channel
Parameters in Expert Mode:
Allowed violations: Percentage of allowed violations.
Used Calibrations:
Generator Launch Voltage Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
101
4
Computer PCI Express Bus Test Application
Figure 4-54: Result description
102
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1: Frequency
Column 2: IL at 0 dB: insertion loss at 0 dB de-emphasis
Column 3: IL at 1 dB: insertion loss at 1 dB de-emphasis
Column 4: IL at 2 dB: insertion loss at 2 dB de-emphasis
4.2.3.1.4
CM Optimization
Purpose and Method:
Common Mode Sinusoidal Interference is generated by coupling two sine waves with
the same phase and amplitude to Normal and Complement of the test signal. If there
is a skew between the two channels or the amplitudes are not equal some amount of
the CM signal will transfer to Differential Mode which is not wanted. This calibration
minimizes this effect by adjusting the delay and the amplitude correction factor
between ch1 and ch2 of the 81150A function generator.
The differential amplitude between the channels of the 81150A is measured with a
DSO for several different delay and amplitude correction factor values. The channels
are perfectly adjusted if the differential amplitude between them is 0. There are
several steps to find the optimal delay and amplitude corrector factor.
The first step sets delays from -12 ns to 12 ns and measures the differential
amplitudes. It uses a linear search algorithm to look for the delay that generates the
minimum differential amplitude, “delay1”. The second step uses a binary search
algorithm to find the optimum delay in the range “delay1” ±100 ps, “delay2”. In the
next step delay is set to “delay2” and a binary search algorithm is used to find the
optimum amplitude corrector factor in the range 0.9 to 1.1. Then this optimal
amplitude corrector factor is set and a binary search algorithm is used again to find
the optimum delay in the range “delay2”±10ps. The last step is to verify that the
differential amplitude between the channels is below the defined maximum (3 mV)
with the optimal delay and amplitude corrector factor values.
The calibration data is stored in a caltable (Pcie3Asic_<channel>_CMResidual.txt).
For the measurements these calibration tables will be used for the amplitude
correction and skew adjustment of the 81150A function generator.
Available for following configuration:
Data Generator: J-BERT N4903B
Connection Diagram:
The same as for “Insertion Loss”.
Parameters in Expert Mode:

Enable DM IF For Optimization: Turn on Differential Mode Interference for
this Calibration. In some cases this leads to better results. If the calibration
fails DM IF can be turned on in a second attempt.

Number of Averages: To reduce the noise floor the differential signal has to
be averaged when measured with the scope.
Used Calibrations: none
Keysight N5990A Test Automation Software Platform for PCIe User Guide
103
4
Computer PCI Express Bus Test Application
Figure 4-55: Result description
104
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1:
Delay Ch1: The optimum delay between channel 1 and channel 2 of the 81150A
Column 2:
Amplitude Correction Factor Ch1: Amplitude Correction Factor for channel 1.
4.2.3.1.5
CM Sinusoidal Interference Calibration
Purpose and Method:
Common Mode Sinusoidal Interference (CMSI) is generated with a 81150A function
generator (s. CM Optimization). It is coupled into the data signal with pick off tees so
the resulting amplitude at the input of the Rx is attenuated and needs to be
calibrated.
First the output amplitude of the 81150A is set to 1V. The resulting amplitude is
measured and used to calculate a scale factor (set amplitude / measured
amplitude).This scale factor is used to calculate the necessary input value to have an
output signal of 300 mV (the level used in Rx tests), EndVoltage. Then CMSI is
calibrated from 0 to EndValue + 500mV with steps of 500mV. In every step the CMSI
is generated with the 81150A and measured with a DSO. When the scope connection
is differential the amplitude is only set in one of the channels of the 81150A. When
the scope connection is single ended the amplitude is set in both channels.
The calibration data is stored in a caltable (Pcie3Asic_<channel>_CmInterference.txt).
For the measurements these calibration tables will be used to adjust the amplitude of
the function generator to the desired output CMSI.
Connection Diagram:
The same as for “Insertion Loss”.
Parameters in Expert Mode:
Number of Averages: To reduce the noise floor the differential signal has to be
averaged when measured with the scope.
Used Calibrations:
CM Optimization
Keysight N5990A Test Automation Software Platform for PCIe User Guide
105
4
Computer PCI Express Bus Test Application
Figure 4-56: Result description
106
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1: Set CM SI Amplitude: CM SI set at the 81150A
Column 2: Measured CM SI Amplitude: The measured CM SI amplitude
Keysight N5990A Test Automation Software Platform for PCIe User Guide
107
4
Computer PCI Express Bus Test Application
4.2.3.1.6
DM Sinusoidal Interference Calibration
Purpose and Method:
When the J-BERT N4903B is used as the data generator Differential Mode Sinusoidal
Interference (DMSI) is generated with the Interference Module (Option J20). It is
coupled into to the data signal with pick off tees.
When the J-BERT M8020A is used as the data generator DMSI is generated with an
internal source.
DMSI is generated in a range from 0 to 400 mV in steps of 50 mV. It is measured with
a real time oscilloscope. For this calibration CMSI is set to 150mV.
The calibration data is stored in a caltable
(Pcie3Asic_<channel>_DmInterference.txt). For the measurements these calibration
tables will be used to adjust the DMSI amplitude to the desired value in the RX input.
Connection Diagram: The same as for “Insertion Loss”.
Parameters in Expert Mode:
Number of Averages: To reduce the noise floor the differential signal has to be
averaged when measured with the scope.
Used Calibrations:

CM Optimization

CM Sinusoidal Interference Calibration
108
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-57: Result description
Column 1: Set DM SI Amplitude: The jitter amplitude set in the instrument
Column 2: Measured DM SI Amplitude: The measured DM SI amplitude
Keysight N5990A Test Automation Software Platform for PCIe User Guide
109
4
Computer PCI Express Bus Test Application
4.2.3.1.7
Random Jitter Calibration
Purpose and Method:
In the Rx tests the input signal will be stressed with a combination of jitter sources to
simulate the possible impairments expected at the Rx input when operating in a
target system. Random jitter is added to simulate the effects of thermal noise. Due to
system intrinsic jitter the effective jitter level is different from the value set in the data
generator, so the jitter amplitude is calibrated.
The test automation starts with small RJ amplitudes and increases them in steps
defined by the “Jitter Step Size” parameter. It measures the corresponding amplitude
for every set amplitude. For this purpose the EasyJit+ application running on the real
time scope is used.
The calibration data is stored in a caltable (Pcie3Asic_<channel>_RandomJitter.txt).
For the measurements these calibration tables will be used to adjust the RJ
amplitude to the desired output RJ amplitudes.
Connection Diagram: The same as for “Insertion Loss.”
Parameters in Expert Mode:
Stop Jitter: The maximum RJ amplitude that is calibrated.
Jitter Step Size: The step size the jitter is increased for every step.
Used Calibrations: none
110
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-58: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
111
4
Computer PCI Express Bus Test Application
Column 1 Set Jitter: The jitter amplitude set in the instrument
Column 2 Actual Jitter: The measured jitter amplitude
4.2.3.1.8
Stressed Voltage Calibration
Purpose and Method:
This procedure calibrates the Eye-Height by adding Differential Mode Sinusoidal
Interference (DMSI) at different Launch Voltage levels.
Due to the low signal levels at the end of the channel there is a lot of noise on the
measured signal. The eye is nearly closed at that point. Just capturing a waveform at
this point with an 8 Gbps Compliance Pattern would be too inaccurate to calibrate
the Eye Height.
For this reason another approach is used. A step (128 zeros followed by 128 ones) is
applied at the input of the calibration channel. The step response is captured at the
output of the replica channel. The oscilloscope averages the step response which
minimizes noise. With the step response the complete electrical behaviour of the
channel is defined. With this data a statistical eye can by calculated. This is done
with the Seasim software. Additionally different impairments like random jitter,
sinusoidal jitter and differential mode sinusoidal inference (DMSI) are simulated by
Seasim. For the Stressed Voltage calibration RJ and SJ are always fixed. DMSI is
increased at every step.
The calibration data is stored in a caltable (Pcie3Asic_<channel>_StrVoltEyeHeight).
This cal data is used to evaluate the optimum amount of DMSI and Launch voltage to
get the desired Eye Height.
Connection Diagram: The same as for “Insertion Loss”.
Parameters in Expert Mode:



Random Jitter: The amount of RJ added to the simulation of the Stressed
Eye.
Sinusoidal Jitter: The amount of SJ added to the simulation of the Stressed
Eye.
DM Interference Step Size: The amount of DMSI added to the simulation at
every step
Used Calibrations:

Launch Voltage Calibration

Insertion Loss Calibration

DM Sinusoidal Interference
112
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-59: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
113
4
Computer PCI Express Bus Test Application
Column 1:
Set DMSI Amplitude: The DMSI amplitude set in the simulation
Column 2:
Eye Height at 500mV Launch Voltage: The simulated Eye-Height with 600 mV LV
Column 3:
Eye Height at 800mV Launch Voltage: The simulated Eye-Height with 800 mV LV
Column 4:
Eye Height at 1100mV Launch Voltage: The simulated Eye-Height with 1000 mV LV
114
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-60: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
115
4
Computer PCI Express Bus Test Application
Column 1:
Set DMSI Amplitude: The DMSI amplitude set in the simulation
Column 2:
Eye Width at 600mV Launch Voltage: The simulated Eye- Width with 600mV LV
Column 3:
Eye Width at 800mV Launch Voltage: The simulated Eye- Width with 800mV LV
Column 4:
Eye Width at 1000mV Launch Voltage: The simulated Eye- Width with 1000mV LV
4.2.3.1.9
Stressed Jitter Calibration
Purpose and Method:
This procedure calibrates the eye-width by adding random jitter at different launch
voltage levels.
The calibration is done for 600, 800 and 1000 mV of launch voltage. For each
amplitude random jitter is increased from 0 with equally spaced steps. The eye width
is measured capturing a step response and using Seasim software. Sinusoidal Jitter
is always fixed to 12.5 ps.
The calibration data is stored in a caltable (Pcie3Asic_<channel>_StrVoltEyeWidth).
This cal data is used to evaluate the optimum amount of random jitter and launch
voltage to get the desired eye width.
Available for following hardware configurations: Only for Long Channel
Connection Diagram: The same as for “Insertion Loss”.
Parameters in Expert Mode:



Sinusoidal Jitter: The amount of SJ added to the simulation of the Stressed
Eye.
End Random Jitter: The maximum amount of RJ added.
Random Jitter Step Size: The amount that RJ is increased in each step.
Used Calibrations:

Launch Voltage Calibration

Insertion Loss Calibration

Random Jitter Calibration
116
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-61: Result description
Column 1:
Set RJ Amplitude: The RJ amplitude set in the simulation
Column 2:
Eye Width at 600mV Launch Voltage: The simulated Eye- Width with 600mV LV
Column 3:
Eye Width at 800mV Launch Voltage: The simulated Eye- Width with 800mV LV
Column 4:
Eye Width at 1000mV Launch Voltage: The simulated Eye- Width with 1000mV LV
Keysight N5990A Test Automation Software Platform for PCIe User Guide
117
4
Computer PCI Express Bus Test Application
Figure 4-62: Result description
118
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1:
Set RJ Amplitude: The RJ amplitude set in the simulation
Column 2:
Eye Width at 600 mV Launch Voltage: The simulated Eye- Width with 600 mV LV
Column 3:
Eye Width at 800 mV Launch Voltage: The simulated Eye- Width with 800 mV LV
Column 4:
Eye Width at 1000 mV Launch Voltage: The simulated Eye- Width with 1000 mV LV
4.2.3.1.10
Stressed Voltage Verification
Purpose and Method:
For the Stressed Jitter Rx test eye-width has to be calibrated to 300 mUI. A second
requirement is that eye-height is in the range of 25 to 35 mV. This procedure checks
if that is possible. If it is outside this band the eye-height can be adjusted by
changing the launch voltage while keeping the eye-width at 300 mUI
Available for following hardware configurations: all
Connection Diagram: The same as for “Common Mode Optimization. ”
Parameters in Expert Mode:

Target Eye Height: The desired eye height.

Target Eye Width: The desired eye width.







Random Jitter: The amount of RJ added to the simulation of the Stressed
Eye. It cannot be changed; it is fixed to the value used in the Stressed
Voltage Calibration. To change it it is necessary to repeat that calibration.
Sinusoidal Jitter: The amount of SJ added to the simulation of the Stressed
Eye. It cannot be changed; it is fixed to the value used in the Stressed
Voltage Calibration. To change it it is necessary to repeat that calibration.
DM Sinusoidal Interference: The amount of DMSI added to the simulation to
obtain the desired eye.
Generator Launch Voltage: Launch Voltage used to obtain the desired eye.
De Emphasis: De-Emphasis level of the signal.
Pre Shoot: Pre shoot level of the signal.
Maximum Retries: Maximum number of times that the procedure tries to
get the target eye height and eye width.
Used Calibrations:

Launch Voltage Calibration

Insertion Loss Calibration

DMSI Calibration

Stressed Voltage Calibration

Eye-Width for Stressed Voltage Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
119
4
Computer PCI Express Bus Test Application
Figure 4-63: Result description
120
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1: Result: Pass/Fail
Column 2: Measured Eye-Height: Simulated Eye Height
Column 3: Min Spec Eye-Height
Column 4: Max Spec Eye-Height
Column 5: Measured Eye-Width: Simulated Eye Height
Column 6: Min Spec Eye-Width
Column 7: Max Spec Eye-Width
Keysight N5990A Test Automation Software Platform for PCIe User Guide
121
4
Computer PCI Express Bus Test Application
4.2.3.1.11
Stressed Jitter Verification
Purpose and Method:
For the Stressed Voltage Rx test eye-height has to be calibrated to a certain value
(e.g. 25 mV for the long channel). A second requirement is that eye-width is in the
range from 300 to 350 mUI. This procedure checks if that is possible. The eye-width
can be adjusted by changing the launch voltage while keeping the eye height
constant (by changing DMSI).
Available for following hardware configurations: Only for long channel
Connection Diagram: The same as for “Insertion Loss. ”
Parameters in Expert Mode:

Target Eye Height: The desired eye height.

Target Eye Width: The desired eye width.







Random Jitter: The amount of RJ added to the simulation of the Stressed
Eye. It cannot be changed; it is fixed to the value used in the Stressed
Voltage Calibration. To change it it is necessary to repeat that calibration.
Sinusoidal Jitter: The amount of SJ added to the simulation of the Stressed
Eye. It cannot be changed; it is fixed to the value used in the Stressed
Voltage Calibration. To change it it is necessary to repeat that calibration.
DM Sinusoidal Interference: The amount of DMSI added to the simulation to
obtain the desired eye.
Generator Launch Voltage: Launch Voltage used to obtain the desired eye.
De Emphasis: De-Emphasis level of the signal.
Pre Shoot: Pre shoot level of the signal.
Maximum Retries: Maximum number of times that the procedure tries to
get the target eye height and eye width.
Used Calibrations:

Launch Voltage Calibration

Insertion Loss Calibration

DMSI Calibration

Random Jitter

Stressed Voltage Calibration
122
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-64: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
123
4
Computer PCI Express Bus Test Application
Column 1: Result: Pass/Fail
Column 2: Measured Eye-Width: Simulated Eye Height
Column 3: Min Spec Eye-Width
Column 4: Max Spec Eye-Width
Column 5: Measured Eye-Height: Simulated Eye Height
Column 5: Min Spec Eye-Height
Column 6: Max Spec Eye-Height
4.2.3.2 Receiver
4.2.3.2.1
Common Receiver Parameters
Data Rate Specific:
1
Data rate deviation

Deviation in ppm that will be applied to the generator data rate.
2
Link Training

Link Training Mode:
o
JBERT _Link_Training: J-BERT sends a loopback training
sequence to the DUT.
o
Vendor_Specific: The vendor can place his DUT in loopback with
his own tools like JTEC, I2C…

Link Training Suite Setting File: The Link Training Suite settings file
(script file) which will be used for loopback training.

Default Link Training Lane Number for every Lane: The lane number
which will be encoded in the TS1s/TS2s sent to the DUT during
loopback training. This value will be set by default for all lanes. If Auto
each lane will be encoded with its own number.

Suppress Loopback Training Messages: When set to true this hides all
popup messages related to loopback training.
3
Error Detector




4
124
Use CDR: If true, CDR is used to generate a clock signal for the JBERT error detector. If false, the clock is supplied by the J-BERT data
generator. With CDR the J-BERT error detector performance is better.
Loop Bandwidth: The loop bandwidth of the J-BERT error detector
CDR.
Enable SSC tracking: Enable SSC tracking if DUT transmits with SSC.
Only for J-BERT N4903B.
Filter Gen1/Gen2 SKPOs for BER test: Ignore SKP symbol in the error
detector. Only for J-BERT N4903B.
BER Measurement

Relax Time: Time span between the point the stress signal is changed
and the BER measurement begins.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Channel Specific:
Pre Shoot: Pre shoot used for receiver tests.
De Emphasis: De Emphasis used for receiver tests.
Lane Specific
Link Training Lane Number: The lane number which will be encoded in the TS1/TS2s
sent to the DUT during loopback
training.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
125
4
Computer PCI Express Bus Test Application
4.2.3.2.2
EQ Coefficient Matrix Scan
Purpose and Method:
This procedure measures the BER with the combination of coefficients C+1 (Precursor) and C-1 (Post-cursor) to create a co-efficient matrix with the BER results. At
each step, the BER value is measured for different values of C+1 coefficient while the
C-1 coefficient value is kept constant. The result values are mapped on to a
triangular matrix where each element contains four entries (measured BER, preshoot, de-emphasis, and boost). Elements on a diagonal line from bottom left to top
right have the same maximum boost value. The elements of the matrix are displayed
in different colors depending on the measured BER value. If the element appears in
green color, the entries value are valid and they can be used for testing. As the color
reaches to red they are invalid for testing.
If the parameter “Allow user to enter optimum equalization for remaining tests” is set
to “True” a window appears to let the user select the values of pre-shoot and deemphasis from the result graph. These selected values are used as the initial values
of pre-shoot and de-emphasis for the EQ Pre-Shoot De-Emphasis Scan.
126
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-65: Connection diagram M8020A common clock architecture no external ref clock no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
127
4
Computer PCI Express Bus Test Application
Figure 4-66: Connection diagram M8020A common clock architecture no external ref clock short channel
128
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-67: Connection diagram M8020A common clock architecture no external ref clock long channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
129
4
Computer PCI Express Bus Test Application
Figure 4-68: Connection diagram M8020A common clock architecture external ref clock no channel
130
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-69: Connection diagram M8020A common clock architecture external ref clock short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
131
4
Computer PCI Express Bus Test Application
Figure 4-70: Connection diagram M8020A common clock architecture external ref clock, long channel
132
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-71: Connection diagram M8020A separated clock architecture no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
133
4
Computer PCI Express Bus Test Application
Figure 4-72: Connection diagram M8020A separated clock architecture short channel
134
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-73: Connection diagram M8020A separated clock architecture long channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
135
4
Computer PCI Express Bus Test Application
Figure 4-74: Connection diagram J-BERT-B common clock architecture no external ref clock no channel
136
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-75: Connection diagram J-BERT-B common clock architecture no external ref clock short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
137
4
Computer PCI Express Bus Test Application
Figure 4-76: Connection diagram J-BERT-B common clock architecture no external ref clock long channel
138
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-77: Connection diagram J-BERT-B common clock architecture external ref clock internal clock multiplier no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
139
4
Computer PCI Express Bus Test Application
Figure 4-78: Connection diagram J-BERT-B common clock architecture external ref clock internal clock multiplier short channel
140
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-79: Connection diagram J-BERT-B common clock architecture external ref clock internal clock multiplier long channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
141
4
Computer PCI Express Bus Test Application
Figure 4-80: Connection diagram J-BERT-B common clock architecture external ref clock N4880A clock multiplier no channel
142
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-81: Connection diagram J-BERT-B common clock architecture external ref clock N4880A clock multiplier short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
143
4
Computer PCI Express Bus Test Application
Figure 4-82: Connection diagram J-BERT-B common clock architecture external ref clock N4880A clock multiplier long channel
144
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-83: Connection diagram J-BERT-B separate clock architecture no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
145
4
Computer PCI Express Bus Test Application
Figure 4-84: Connection diagram J-BERT-B separate clock architecture short channel
146
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-85: Connection diagram J-BERT-B separate clock architecture long channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
147
4
Computer PCI Express Bus Test Application
Parameters in Expert Mode:
1
Loopback Training

Force training at each BER measurement: Force retraining at each
BER measurement for different PS/DE combination.
2
Eye Parameter

Eye-Height: The eye height of the signal when it is stressed with all the
impairments.







Eye-Width: The eye width of the signal when it is stressed with all the
impairments.
Random Jitter: The amount of RJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage
Calibration. To change it it is necessary to repeat that calibration.
Sinusoidal Jitter: The amount of SJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage
Calibration. To change it it is necessary to repeat that calibration.
Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.
Differential Mode Sinusoidal Interference: The amount of DMSI
added to the signal to obtain the desired eye.
Generator Launch Voltage: Launch Voltage used to obtain the desired
eye.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude.
3
Coefficient Variation

Coefficient Divider: The coefficient divider for C-1 and C+1.

Maximum Boost: Coefficient c+1 is increased until this Boots level is
exceeded.

Start Pre-Shoot: Start Pre-Shoot value in dB.

Start De-Emphasis: Start De-Emphasis value in dB.
4
BER Measurement

BER Mode: Read-only. The BER measurement is executed in
TargetBER mode for this test (measure until a target BER is achieved).


5
Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.
Confidence Level: The confidence level value used for the BER
measurement.
Equalization for remaining Rx Tests
Allow user to enter optimum equalization for remaining Rx tests:
Controls if a window appears at the end of the test to let the user set
the pre-shoot and de-emphasis values to use for the following tests.
Used Calibrations:

Preset Calibration
148
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application





4
Random Jitter Calibration
Sinusoidal Jitter Calibration
DM Sinusoidal Interference Calibration
Eye Height Calibration
Eye Width Calibration
Figure 4-86: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
149
4
Computer PCI Express Bus Test Application
4.2.3.2.3
Equalization Scan
Purpose and Method:
The purpose of this test is to find the optimum combination of de-emphasis and preshoot amplitude. As a first step, the procedure sets initial de-emphasis and preshoot values and adjusts the eye height to obtain the desired BER (slightly above 1e9). Then it keeps the initial pre-shoot and makes a de-emphasis scan, measuring the
BER for every de-emphasis value. After that it keeps the initial de-emphasis
amplitude and makes a pre-shoot scan. Finally the test shows the result tables, one
for the de-emphasis scan and one for the pre-shoot scan. Those let the user see the
best combination with the selected initial values.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:
1
Parameter

Scan Order: Select if De-Emphasis or Pre-Shoot is tested first.

Initial De Emphasis: The initial de-emphasis used for BER adjustment
and pre-shoot scan.

Initial Pre Shoot: The initial pre-shoot used for BER adjustment and
de-emphasis scan.
Force training at each Preset: If true, every time the de-emphasis or
the pre-shoot is changed the DUT is trained into loopback again.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude.


2
De-Emphasis Variation

Start De Emphasis: Start amplitude for the de-emphasis scan.

Stop De Emphasis: Stop amplitude for the de-emphasis scan.

De Emphasis Step Size: Step size for the de-emphasis scan.
3
Pre-Shoot Variation

Start Pre Shoot: Start amplitude for the pre-shoot scan.

Stop Pre Shoot: Stop amplitude for the pre-shoot scan.

De Pre Shoot Size: Step size for the pre-shoot scan.
4
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.

Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.

Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.


150
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
5
4
Equalization for remaining Rx Tests
Allow user to enter optimum equalization for remaining Rx tests.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height Calibration

Eye Width Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
151
4
Computer PCI Express Bus Test Application
Figure 4-87: Result description
152
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-88: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
153
4
Computer PCI Express Bus Test Application
4.2.3.2.4
Stressed Voltage Rx Test
Purpose and Method:
This test verifies that the DUT properly functions in presence of the minimum eye
height that the specification allows. The minimum value of eye height depends on the
calibration channel. The eye width must be between 0.3 and 0.35 UI. Eye height is
generated by adding the combination of Differential Mode Sinusoidal Interference
and Launch Voltage that also gets as close as possible to the desired eye width.
Random Jitter and Sinusoidal Jitter are fixed to the values used in Stressed Voltage
Calibration.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:

Enable Impairments for Loopback Training.

Eye-Height: The eye height of the signal when it is stressed with all
impairments.

Eye-Width: The eye width of the signal when it is stressed with all
impairments.
Random Jitter: The amount of RJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage Calibration. To
change it it is necessary to repeat that calibration.
Sinusoidal Jitter: The amount of SJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage Calibration. To
change it it is necessary to repeat that calibration.









Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.
Differential Mode Sinusoidal Interference: The amount of DMSI added to the
signal to obtain the desired eye.
Generator Launch Voltage: Launch Voltage used to obtain the desired eye.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude added to the signal.
BER Mode: The BER measurement can be executed for a fixed time or until
a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the test. For
a Target BER of 1E-12 and a Confidence Level of 95% the test will run 10
minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no bit error occurs.
Confidence Level: Confidence level value when BER mode is TargetBer.

BER Measurement Duration: the duration of the BER measurement when
mode is FixedTime.

Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Launch Voltage Calibration

CM Optimization

CM Sinusoidal Interference Calibration

DM Sinusoidal Interference Calibration

Random Jitter Calibration
154
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application


4
Insertion Loss Calibration
Stressed Voltage Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
155
4
Computer PCI Express Bus Test Application
Figure 4-89: Result description
156
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1:
Pass/Fail Result: Measured BER should be less than the target BER
Column 2:
Target BER.
Column 3:
BER: Measured BER.
Figure 4-90: Connection diagram M8000 separated clock architecture short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
157
4
Computer PCI Express Bus Test Application
Figure 4-91: Connection diagram M8000 separated clock architecture long channel
Figure 4-92: Connection diagram M8000 common clock architecture no external ref clock no channel
158
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-93: Connection diagram M8000 common clock architecture no external ref clock short channel
Figure 4-94: Connection diagram M8000 common clock architecture no external ref clock long channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
159
4
Computer PCI Express Bus Test Application
Figure 4-95: Connection diagram M8000 common clock architecture external ref clock no channel
160
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-96: Connection diagram M8000 common clock architecture external ref clock short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
161
4
Computer PCI Express Bus Test Application
Figure 4-97: Connection diagram M8000 common clock architecture external ref clock long channel
162
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-98: Connection diagram J-BERT-B separated clock architecture no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
163
4
Computer PCI Express Bus Test Application
Figure 4-99: Connection diagram J-BERT-B separated clock architecture short channel
164
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-100: Connection diagram J-BERT-B separated clock architecture long channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
165
4
Computer PCI Express Bus Test Application
Figure 4-101: Connection diagram J-BERT-B common clock architecture no external ref clock no channel
166
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-102: Connection diagram J-BERT-B common clock architecture no external ref clock short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
167
4
Computer PCI Express Bus Test Application
Figure 4-103: Connection diagram J-BERT-B common clock architecture no external ref clock long channel
168
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-104: Connection diagram J-BERT-B common clock architecture external ref clock internal clock multiplier no channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
169
4
Computer PCI Express Bus Test Application
Figure 4-105: Connection diagram J-BERT-B common clock architecture external ref clock internal clock multiplier short channel
170
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-106: Connection diagram J-BERT-B common clock architecture external ref clock internal clock multiplier long channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
171
4
Computer PCI Express Bus Test Application
Figure 4-107: Connection diagram J-BERT-B common clock architecture external ref clock N4880A clock multiplier no channel
172
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-108: Connection diagram J-BERT-B common clock architecture external ref clock N4880A clock multiplier short channel
Keysight N5990A Test Automation Software Platform for PCIe User Guide
173
4
Computer PCI Express Bus Test Application
Figure 4-109: Connection diagram J-BERT-B common clock architecture external ref clock N4880A clock multiplier long channel
174
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Parameters in Expert Mode:

Eye-Height: The eye height used in this procedure.

Eye-Width: The eye width used in this procedure.

Random Jitter: The amount of random jitter (rms) added to the test signal.

Sinusoidal Jitter: The amount of sinusoidal jitter added to the test signal.




Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude added to the signal.
BER Measurement duration: Test duration.
Allowed Bit Error: Max number of error bits to pass the test.
Used Calibrations:

Launch Voltage Calibration

CM Optimization

CM Sinusoidal Interference Calibration

DM Sinusoidal Interference Calibration

Random Jitter Calibration

Insertion Loss Calibration

Stressed Voltage Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
175
4
Computer PCI Express Bus Test Application
Figure 4-110: Result description
Column 1:
Pass/Fail Result: The number of errors should be smaller than “Allowed Bit Error”.
Column 2:
Received Bits: Number of bits received during the test.
Column 3:
Bit Errors: Number of errors.
176
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.3.2.5
4
Stressed Jitter Rx Test
Purpose and Method:
This test verifies that the receiver meets the eye width specification. Eye width is set
to the minimum of the specification, which is 0.3UI, and the eye height must be
between 25 and 35 mVpp. Eye width is generated by adding the combination of
Random Jitter and Launch Voltage that also gets as close as possible to the desired
eye height. Sinusoidal Jitter is fixed to the value used in Stressed Jitter Calibration.
Available for following hardware configurations: Only for Long Channel.
Connection Diagram:
The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:

Enable Impairments for Loopback Training.










Eye-Height: The eye height of the signal when it is stressed with all
impairments.
Eye-Width: The eye width of the signal when it is stressed with all
impairments.
Random Jitter: The amount of RJ added to the signal that, in combination
with the Launch Voltage, creates an eye with the desired eye width.
Sinusoidal Jitter: The amount of SJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Jitter Calibration. To
change it it is necessary to repeat that calibration.
Differential Mode Sinusoidal Interference: The amount of DMSI added to the
signal. It is 0 for this test and only included as a read-only parameter for
reference.
Generator Launch Voltage: Launch Voltage added to the signal that, in
combination with the RJ, creates an eye with the desired eye width.
Common Mode Sinusoidal Interference: The amount of CMSI added to the
signal. It is 0V for this test and only included as a read-only parameter for
reference.
BER Mode: The BER measurement can be executed for a fixed time or until
a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the test. For
a Target BER of 1E-12 and a Confidence Level of 95% the test will run 10
minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no bit error occurs.
Confidence Level: Confidence level value when BER mode is TargetBer.

BER Measurement Duration: The duration of the BER measurement when
mode is FixedTime.

Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Launch Voltage Calibration

CM Optimization

CM Sinusoidal Interference Calibration

DM Sinusoidal Interference Calibration

Random Jitter Calibration

Insertion Loss Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
177
4
Computer PCI Express Bus Test Application

178
Stressed Voltage Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-111: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
179
4
Computer PCI Express Bus Test Application
Column 1:
Pass/Fail Result: The number of bit errors should be smaller than “Allowed Bit
Error”.
Column 2:
Allowed Bit Errors
Column 3:
Measured Bit Errors.
180
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.3.2.6
4
Sensitivity with Generator Launch Voltage Rx Test
Purpose and Method:
This test characterizes the minimum eye height at which the DUT passes the BER
test. The method starts with “Start Eye Height” and decreases with steps of “Eye
Height Step Size” until the number of bit errors is bigger than 0. The Eye height is
generated by adjusting the Generator Launch Voltage. Differential Mode Sinusoidal
Interference is fixed.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:
1
Sensitivity Variation

Start Eye Height: The eye height where the test starts.

Stop Eye Height: The eye height where the test stops.

2
Parameter

Random Jitter: The amount of RJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage
Calibration. To change it it is necessary to repeat that calibration.

Sinusoidal Jitter: The amount of SJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage
Calibration. To change it it is necessary to repeat that calibration.

Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.

Differential Mode Sinusoidal Interference: The amount of DMSI added
to the signal to achieve the desired eye height.

3
Eye Height Step Size: The amount the eye height is decreased by one
step to search the “Min passed Eye Height”.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude added to the signal.
BER Measurement





BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.
Confidence Level: Confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Launch Voltage Calibration

CM Optimization
Keysight N5990A Test Automation Software Platform for PCIe User Guide
181
4
Computer PCI Express Bus Test Application





CM Sinusoidal Interference Calibration
DM Sinusoidal Interference Calibration
Random Jitter Calibration
Insertion Loss Calibration
Stressed Voltage Calibration
Figure 4-112: Result description
182
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1:
Pass/Fail Result: Max Passed Eye Height should be smaller than the Spec Limit.
Column 2:
Max Passed Eye Height: The smallest eye height at which the DUT passes the error
test.
Column 3:
Min Spec: The smallest eye height at which the DUT has to pass the error test to
meet the specifications.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
183
4
Computer PCI Express Bus Test Application
4.2.3.2.7
Sensitivity with Differential Interference Mode Rx Test
Purpose and Method:
This test characterizes the minimum eye height at which the DUT passes the BER
test. The method starts with “Start Eye Height” and decreases with steps of “Eye
Height Step Size” until the number of bit errors is bigger than 0. The Eye height is
generated by adjusting the Differential Mode Sinusoidal Interference. Generator
Launch Voltage is fixed.
Connection Diagram:
The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:
1
Sensitivity Variation

Start Eye Height: The eye height where the test starts.

Stop Eye Height: The eye height where the test stops.

2
Parameter

Random Jitter: The amount of RJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage
Calibration. To change it it is necessary to repeat that calibration.

Sinusoidal Jitter: The amount of SJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage
Calibration. To change it it is necessary to repeat that calibration.



3
Eye Height Step Size: The amount the eye height is decreased by one
step to search the “Min passed Eye Height”.
Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.
Generator Launch Voltage: Launch Voltage used to obtain the desired
eye height.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude added the signal.
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.

Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.

Confidence Level: Confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.


Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Launch Voltage Calibration

CM Optimization
184
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application





4
CM Sinusoidal Interference Calibration
DM Sinusoidal Interference Calibration
Random Jitter Calibration
Insertion Loss Calibration
Stressed Voltage Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
185
4
Computer PCI Express Bus Test Application
Figure 4-113: Result description
186
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1:
Pass/Fail Result: Max Passed Eye Height should be smaller than the Spec Limit.
Column 2:
Max Passed Eye Height: The smallest eye height at which the DUT passes the error
test.
Column 3:
Min Spec: The smallest eye height at which the DUT has to pass the error test to
meet the specifications.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
187
4
Computer PCI Express Bus Test Application
4.2.3.2.8
Receiver Jitter Tolerance
Purpose and Method:
This test characterizes how much jitter a DUT can tolerate at different frequencies of
sinusoidal jitter. Starting with the “Min Frequency” the jitter is increased with equally
spaced steps until the BER test fails. The BER test fails when the number of error bits
is bigger than 0. This procedure is repeated for all SJ frequencies.
Connection Diagram:
The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:
1
Sinusoidal Jitter Variation









Frequency Mode: Specifies the distribution of the frequency points to
test. It can be:
Compliance Frequencies: The frequencies defined in the specification
for compliance testing.
Equally Spaced Frequencies.
User Defined Frequencies.
Single Frequency.
Frequency Scale: The results can be represented in logarithmic or
linear scaling. For “Equally Spaced Frequencies” the frequency points
depend on the chosen frequency scale.
Start Frequency: Start frequency for “Equally Spaced Frequencies”.
Stop Frequency: Stop frequency for “Equally Spaced Frequencies”.
Number of Frequency Steps: Number of frequency points for “Equally
Spaced Frequencies”.
Search Algorithm (HyteresisUp or Binary):
The Binary search is not recommended for
devices with a long recovery time.

Frequency Points: Frequency points for “User Defined Frequencies”.

Start Jitter Points: Start jitter amplitudes for “User Defined
Frequencies”. In the other frequency modes start jitter amplitude is
always 0.
Jitter Frequency: Single frequency point for “Single Frequency”.

2
188

Jitter Steps: Number of jitter values tested to search the “Max passed
jitter” at one frequency.

Show Min Failed Points: The results graph can show the minimum
failed jitter in addition to the maximum passed jitter for each tested
frequency.
Parameter

Default Impairments:

Use Stressed Voltage Impairments: The impairments are set to values
that match the stressed voltage specification.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application




3
4
Use Stressed Jitter Impairments: The impairments are set to values
that match the stressed jitter specification.
Generator Launch Voltage.
Differential Mode Sinusoidal Interference: The amount of DMSI
added to the signal.
Random Jitter: The amount of random jitter (rms) added to the signal.

CMSI Amplitude: Common Mode Sinusoidal Interference amplitude
added to the signal.

Force Retraining on Each Frequency: Re-train the DUT at each tested
jitter frequency.
BER Measurement





BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.
Confidence Level: Confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Launch Voltage Calibration

CM Optimization

CM Sinusoidal Interference Calibration

DM Sinusoidal Interference Calibration

Random Jitter Calibration

Insertion Loss Calibration

Stressed Voltage Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
189
4
Computer PCI Express Bus Test Application
Figure 4-114: Result description
190
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:
Pass/Fail Result: Max Passed Jitter should be bigger than Min Spec.
Column 2:
Sinusoidal Jitter Frequency: The tested frequency points.
Column 3:
Max Passed Jitter: Max jitter that passed the Target BER.
Column 4:
Jitter Capability Test Setup: Max jitter that the hardware can generate.
Column 5:
Min Spec: Min jitter that has to pass the test to meet the specification.
Column 6:
Margin: Margin between the max passed jitter and the specification.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
191
4
Computer PCI Express Bus Test Application
4.2.3.3 Receiver Setup (Only for Debugging)
4.2.3.3.1
Common Parameters
Channel Specific:

Pre Shoot: Pre-shoot level of the signal. It is 0 by default.

192
De Emphasis: De Emphasis level of the signal. It is set to the optimum deemphasis using the Insertion Loss Calibration.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.3.3.2
4
Stressed Voltage Rx Setup
Purpose and Method:
The purpose of this procedure is to configure the data generator with the parameters
needed in the Stressed Voltage Rx test using the calibration data saved on the PC
where ValiFrame is running. The method begins like Rx Stressed Voltage test but
then the test will not run, it only leaves the setup prepared. The set parameters are
the differential amplitude, random jitter, sinusoidal jitter, common mode sinusoidal
interference jitter and differential mode sinusoidal interference. The differential
amplitude and the DMSI are defined by the eye height and width desired using the
stressed voltage calibration.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:

Eye-Height: The desired eye height.

Eye-Width: The desired eye width.

Random Jitter: The amount of RJ added to the signal. It cannot be
changed; it is fixed to the value used in the Stressed Voltage Calibration. To
change it it is necessary to repeat that calibration.





Sinusoidal Jitter: The amount of SJ added to the signal. It cannot be
changed; it is fixed to the value use in the Stressed Voltage Calibration. To
change it is necessary to repeat that calibration.
Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.
Differential Mode Sinusoidal Interference: The amount of DMSI added to the
signal to obtain the desired eye.
Generator Launch Voltage: Launch Voltage used to obtain the desired eye.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude added to the signal.
Used Calibrations:

Launch Voltage Calibration

CM Optimization

CM Sinusoidal Interference Calibration

DM Sinusoidal Interference Calibration

Random Jitter Calibration

Insertion Loss Calibration

Stressed Voltage Calibration
Result Description: None.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
193
4
Computer PCI Express Bus Test Application
4.2.3.3.3
Stressed Jitter Rx Setup
Purpose and Method:
The purpose of this procedure is to configure the data generator with the parameters
needed in the Stressed Jitter Rx. This procedure is only available for Long Channel
setup. The set parameters are the differential amplitude, random jitter, common
mode sinusoidal interference jitter and differential mode sinusoidal interference.
Connection Diagram: Same as for “Receiver Stressed Jitter Rx test”
Parameters in Expert Mode:

Eye-Height: The desired eye height.

Eye-Width: The desired eye width.

Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude added to the signal.
Used Calibrations:

Launch Voltage Calibration

CM Optimization

CM Sinusoidal Interference Calibration

DM Sinusoidal Interference Calibration

Random Jitter Calibration

Insertion Loss Calibration

Stressed Jitter Calibration
Result Description: None.
194
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.3.3.4
4
Stressed Voltage Characterization Rx Setup
Purpose and Method:
The purpose of this procedure is the same as the Stressed Voltage Rx Setup. The
only difference is that in this case the user can directly select the generator launch
voltage and the differential mode sinusoidal interference instead of the eye height
and width.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:

Generator Voltage: Differential amplitude of the signal.

Random Jitter: The amount of random jitter (rms) added to the signal.

DMSI: Differential mode sinusoidal interference amplitude added to the
signal.

Sinusoidal Jitter: The amount of sinusoidal jitter added to the signal.


Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.
Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude added to the signal.
Used Calibrations:

Launch Voltage Calibration

CM Optimization

CM Sinusoidal Interference Calibration

DM Sinusoidal Interference Calibration

Random Jitter Calibration

Insertion Loss Calibration

Stressed Voltage Calibration
Result Description: None.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
195
4
Computer PCI Express Bus Test Application
4.2.4 Gen1 & Gen2 CEM Tests
4.2.4.1 Calibration
4.2.4.1.1
Common Calibration Parameters
Scope Connection:
All calibrations can be done with either a differential probe or a single ended direct
connection on two channels, with one exception: The Common Mode Optimization
can only be done with a differential probe.
Jitter Unit:
All the jitter parameters can be displayed in time or in unit interval. The graphics
results will be represented in the chosen unit.
4.2.4.1.2
Random Jitter Calibration
Purpose and Method:
In the Rx tests the input signal will be stressed with a combination of jitter sources to
simulate the possible impairments expected at the Rx input when operating in a
target system. Random jitter is added to simulate the effects of the thermal noise.
Due to system intrinsic jitter the effective jitter level is different from the value set in
the data generator, so the jitter amplitude is calibrated.
The test automation calibrates six equally spaced RJ values (from 0 to 10ps). The JBERT sends a clock pattern during this calibration procedure. The actual jitter it is
measured on a DSO using the RJ/DJ-separation software EzJIT.
The calibration data is stored in a caltable
(Pcie<generation>_<DutType>_RandomJitter.txt). The calibration must be done for
generation 1 and 2 for add-in-cards and system. For the measurements these
calibration tables will be used to calculate the RJ amplitude that needs to be set on
the generator to get the desired RJ amplitude at the test point.
196
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-115: Connection diagram M8020A addincard
Keysight N5990A Test Automation Software Platform for PCIe User Guide
197
4
Computer PCI Express Bus Test Application
Figure 4-116: Connection diagram M8020A system
198
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-117: Connection diagram J-BERT-B addincard
Keysight N5990A Test Automation Software Platform for PCIe User Guide
199
4
Computer PCI Express Bus Test Application
Figure 4-118: Connection diagram J-BERT-B system card
Parameters in Expert Mode: None
Used Calibrations: none
200
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-119: Result description
Column 1 Set Jitter: The jitter amplitude set in the instrument
Column 2 Actual Jitter: The measured jitter amplitude
Keysight N5990A Test Automation Software Platform for PCIe User Guide
201
4
Computer PCI Express Bus Test Application
4.2.4.1.3
De-Emphasis Calibration
Purpose and Method:
This procedure calibrates the de-emphasis. The test automation starts with -6 dB of
de-emphasis (by default), increases it with a step size of 0.5 dB and measures for
every set value the corresponding de-emphasis. The calibration ends when the set
de-emphasis is 0 or the measured de-emphasis is bigger than -0.22 dB.
The calibration data is stored in a caltable
(Pcie<generation>_<DutType>__DeEmphasis.txt). The calibration must be done for
generation 1 and 2 for add-in-cards and systems. For the measurements these
calibration tables will be used to calculate the de-emphasis level that needs to be set
on the generator to get the desired de-emphasis level at the test point.
Connection Diagram:
The same as for “Random Jitter Calibration”
Parameters in Expert Mode:

Eye-Height: The eye height used in this procedure.

Start De-Emphasis: The minimum de-emphasis that is calibrated.
Used Calibrations: none
202
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-120: Result description
Column 1 Set De-Emphasis: The de-emphasis set in the instrument.
Column 2 Actual De-Emphasis: The measured de-emphasis.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
203
4
Computer PCI Express Bus Test Application
4.2.4.1.4
Eye Height Calibration
Purpose and Method:
The test fixtures attenuate the data signal. To compensate for this, the data signal
differential swing is calibrated.
The test automation calibrates five equally spaced differential voltage amplitudes.
The minimum amplitude is 100mV and the maximum is the maximum value that the
data generator can generate.
For this calibration the data generator sends the compliance pattern. It also adds
random jitter, ISI, swept sinusoidal jitter and for generation 2 high frequency
sinusoidal jitter and SSC residual to the signal. The eye height is measured in the
scope using horizontal histograms. The actual eye height depends on the ISI trace so
the calibration procedure repeats this method for each trace.
The calibration data is stored in a caltable
(Pcie<generation>_<DutType>__EyeHeight.txt). The calibration must be done for
generation 1 and 2 for add-in-cards and systems. For the measurements these
calibration tables will be used to adjust the differential voltage amplitude to the
desired eye height.
Connection Diagram: The same as for “Random Jitter Calibration”
Parameters in Expert Mode: None
Used Calibrations:

Random Jitter Calibration

De-Emphasis Calibration
204
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-121: Result description
Column 1
Set Diff Voltage: The differential voltage amplitude set in the instrument
Column 2
Measured Eye Height: The measured eye height amplitude
Keysight N5990A Test Automation Software Platform for PCIe User Guide
205
4
Computer PCI Express Bus Test Application
4.2.4.1.5
Eye Height Verification
Purpose and Method:
This procedure checks if it is possible to generate a signal with an eye height value
that is inside a range of +-10% tolerance with respect to the eye height target value.
The target eye height is the minimum value specified in the CTS.
For this verification the data generator sends the compliance pattern. It also adds
CMSI, random jitter, ISI, HF sinusoidal jitter and SSC residual to the signal. The eye
height is measured in the scope using horizontal histograms.
Connection Diagram: The same as for “Random Jitter Calibration”
Parameters in Expert Mode: None
Used Calibrations:

Random Jitter Calibration

Isi Calibration

CM Sinusoidal Interference

De-Emphasis Calibration
Figure 4-122: Result description
Column 1
Set Diff Voltage: The differential voltage amplitude set in the instrument
Column 2
Measured Eye Height: The measured eye height amplitude
206
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.2.4.2 Receiver
4.2.4.2.1
Common Receiver Parameters
Data Rate Specific:
1
Data rate deviation

Deviation in ppm that will be applied to the generator data rate (for
Add-in Cards only).
2
Link Training

Link Training Mode:
o
JBERT _Link_Training: J-BERT sends a loopback training
sequence to the DUT.
o
Vendor_Specific: The vendor can place his DUT in loopback with
his own tools like JTEC, I2C…

Link Training Suite Setting File: The Link Training Suite setting file
(script file) which will be used for loopback training.

Default Link Training Lane Number for every Lane: The lane number
which will be encoded in the TS1s/TS2s sent to the DUT during
loopback training. This value will be set by default for all lanes. If Auto
each lane will be encoded with its own number.

Suppress Loopback Training Messages: When set to true this hides all
popup messages related to loopback training.
3
Error Detector

Use CDR: If true, CDR is used to generate a clock signal for the JBERT error detector. If false, the clock is supplied by the J-BERT data
generator. With CDR the J-BERT error detector performance is better.

Loop Bandwidth: The loop bandwidth of the J-BERT error detector
CDR.

Enable SSC tracking: Enable SSC tracking if DUT transmit with SSC.
Only for J-BERT N4903B.

Filter Gen1/Gen2 SKPOs for BER test: Ignore SKP symbol in the error
detector. Only for J-BERT N4903B.
3) BER Measurement

Relax Time: Time span between the point the stress signal is changed
and the BER measurement begins.
Lane Specific

4.2.4.2.2
Link Training Lane Number: The lane number which will be encoded in the
TS1/TS2s sent to the DUT during loopback training.
Receiver Compliance
Keysight N5990A Test Automation Software Platform for PCIe User Guide
207
4
Computer PCI Express Bus Test Application
Purpose and Method:
This test determines if the DUT meets the receiver specifications. The procedure
measures the BER when all jitter types and the eye height are set to their spec limit
values (maximum for jitter, minimum for eye height). In expert mode these values can
be changed.
Figure 4-123: Connection diagram J-BERT-B M8020 addincard common clock architecture no external ref clock
208
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-124: Connection diagram J-BERT-B M8020A addincard common clock architecture external ref clock internal clock multiplier
Keysight N5990A Test Automation Software Platform for PCIe User Guide
209
4
Computer PCI Express Bus Test Application
Figure 4-125: Connection diagram J-BERT-B M8020A addincard separate clock architecture
210
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-126: Connection diagram J-BERT-B M8020A system common clock architecture internal clock multiplier
Keysight N5990A Test Automation Software Platform for PCIe User Guide
211
4
Computer PCI Express Bus Test Application
Figure 4-127: Connection diagram J-BERT-B M8020A system separate clock architecture
212
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-128: Connection diagram J-BERT-B addincard common clock architecture no external ref clock
Keysight N5990A Test Automation Software Platform for PCIe User Guide
213
4
Computer PCI Express Bus Test Application
Figure 4-129: Connection diagram J-BERT-B addincard common clock architecture external ref clock internal clock multiplier
214
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-130: Connection diagram J-BERT-B addincard common clock architecture external ref clock N4880A clock multiplier
Keysight N5990A Test Automation Software Platform for PCIe User Guide
215
4
Computer PCI Express Bus Test Application
Figure 4-131: Connection diagram J-BERT-B addincard separate clock architecture
216
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-132: Connection diagram J-BERT-B system common clock architecture internal clock multiplier
Keysight N5990A Test Automation Software Platform for PCIe User Guide
217
4
Computer PCI Express Bus Test Application
Figure 4-133: Connection diagram J-BERT-B system common clock architecture N4880A clock multiplier
218
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-134: Connection diagram J-BERT-B system separate clock architecture
Keysight N5990A Test Automation Software Platform for PCIe User Guide
219
4
Computer PCI Express Bus Test Application
Parameters in Expert Mode:
1
Generator Jitter

Random Jitter: The amount of random jitter (rms) added to the test
signal.

Swept sinusoidal Jitter: The amplitude of the sinusoidal jitter
component that is swept continuously from 1.5 to 100MHz during the
test.

SSC Residual (only for 5.0Gbit/s): The SSC Residual emulates the
residual which is caused by path length differences in the clock
distribution and SSC modulation in real world systems. The residual
SSC is triangular. It should be >= 75 ps.
2
Eye Height

Eye Height: The eye height used in this procedure.
3
BER Measurement





BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.
Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Random Jitter Calibration

De-Emphasis Calibration

Eye Height Calibration
220
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-135: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
221
4
Computer PCI Express Bus Test Application
Column 1: Pass/Fail Result: The BER measured should be smaller than the Target
BER.
Column 2: Target BER.
Column 3: BER: Measured BER.
222
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.4.2.3
4
Receiver Jitter Tolerance
Purpose and Method:
This test procedure searches the maximum sinusoidal jitter at which the DUT passes
the BER test. The test first uses relatively large steps to go linearly from “Start Jitter”
up. When BER test fails it goes back down with mid-sized steps until it passes again.
From there it steps up again with small steps until an error is found again. The
maximum passed value is the last test point that did not return an error. All of this
happens separately for each frequency.
Connection Diagram: The same as for “Receiver Compliance”
Parameters in Expert Mode:
1
Sinusoidal Jitter Variation









Frequency Mode: Specifies the distribution of the frequency points to
test. It can be:
Compliance Frequencies: The frequencies defined in the specification
for compliance testing.
Equally Spaced Frequencies.
User Defined Frequencies.
Single Frequency.
Frequency Scale: The results can be represented in logarithmic or
linear scaling. For “Equally Spaced Frequencies” the frequency points
depend on the chosen frequency scale.
Start Frequency: Start frequency for “Equally Spaced Frequencies”.
Stop Frequency: Stop frequency for “Equally Spaced Frequencies”.
Number of Frequency Steps: Number of frequency points for “Equally
Spaced Frequencies”.
Search Algorithm (HyteresisUp or Binary):
The Binary search is not recommended for
devices with a long recovery time.





2
Frequency Points: Frequency points for “User Defined Frequencies”.
Start Jitter Points: Start jitter amplitudes for “User Defined
Frequencies”. In the other frequency modes start jitter amplitude is
always 0.
Jitter Frequency: Single frequency point for “Single Frequency”.
Jitter Step Size: The size of the smallest sinusoidal jitter amplitude
step used to search the “Max passed jitter” at each frequency.
Show Min Failed Points: The results graph can show the minimum
failed jitter in addition to the maximum passed jitter for each tested
frequency.
Generator Jitter


Random Jitter: The amount of random jitter (rms) added to the test
signal.
HF Sinusoidal Jitter (only for 5.0Gbit/s): The amount of high frequency
Keysight N5990A Test Automation Software Platform for PCIe User Guide
223
4
Computer PCI Express Bus Test Application
sinusoidal jitter added to the test signal. It should be >= 27ps.


3
HF Sinusoidal Jitter Frequency (only for 5.0Gbit/s): Frequency of the
HF sinusoidal jitter component.
SSC Residual (only for 5.0Gbit/s): The SSC Residual emulates the
residual which is caused by path length differences in the clock
distribution and SSC modulation in real world system. The residual
SSC is triangular. It should be >= 75 ps.
Parameter

Force retraining on each frequency: Re-train the DUT at each tested
jitter frequency.
4
Eye Height

Eye Height: The eye height used in this procedure.
5
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.

Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.

Confidence Level: The confidence level value when BER mode
is
TargetBer.

BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.

Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Random Jitter Calibration

De-Emphasis Calibration

Eye Height Calibration
224
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-136: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
225
4
Computer PCI Express Bus Test Application
Column1:
Pass/Fail Result: Max passed jitter should be bigger than min spec.
Column 2:
Sinusoidal Jitter Frequency: Tested frequency points.
Column 3:
Max Passed LF Deterministic Jitter: Max jitter that passed the Target BER.
Column 4:
Jitter Capability Test Setup: Max jitter that the hardware can generate.
Column 5:
Min Spec: Min jitter that has to pass the test to meet the specification.
Column 6:
Margin: Margin between the max passed jitter and the specification.
4.2.4.2.4
Receiver Sensitivity
Purpose and Method:
This test searches the minimum eye height at which the DUT passes the BER test.
The method starts with “Start Eye Height” and decreases with steps of “Step Size”.
The minimum passed value is the last test point that did not return an error.
Connection Diagram:
The same as for “Receiver Compliance”
Parameters in Expert Mode:
1
Generator Jitter

Use Jitter: If this is enabled, jitter is added to the test signal.





2
HF Sinusoidal Jitter (only for 5.0Gbit/s): The amount of high frequency
sinusoidal jitter added to the test signal. It should be >= 27ps.
HF Sinusoidal Jitter Frequency (only for 5.0Gbit/s): Frequency of the
HF sinusoidal jitter component.
SSC Residual (only for 5.0Gbit/s): The SSC Residual emulates the
residual which is caused by path length differences in the clock
distribution and SSC modulation in real world systems. The residual
SSC is triangular. It should be >= 75 ps.
Eye Height


226
Random Jitter: The amount of random jitter (rms) added to the test
signal.
Swept sinusoidal Jitter: The amplitude of the sinusoidal jitter
component that is swept continuously from 1.5 to 100 MHz during the
test.
Loopback Training Eye Height: The eye height used for loopback
training.
Start Eye Height: The eye height (transitions bits) where the test
starts.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
3
4

Stop Eye Height: The eye height (transitions bits) where the test stops.

Step Size: The amount the eye height is decreased each step to search
the “Min Passed Eye Height”.
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.

Target BER: The Target BER for the BER measurement used in the
test. For a Target BER of 1E-12 and a Confidence Level of 95% the
test will run 10 minutes for 5 Gbit/s and 20 minutes for 2.5 Gbit/s if no
bit error occurs.

Confidence Level: The confidence level value when BER mode is
TargetBer.

BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.

Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Random Jitter Calibration

De-Emphasis Calibration

Eye Height Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
227
4
Computer PCI Express Bus Test Application
Figure 4-137: Result description
228
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:
Pass/Fail Result: The Min Passed Eye Height measured should be smaller than the
Min Spec.
Column 2:
Min Passed Eye Height: The smallest eye height at which the DUT passes the BER
test.
Column 3:
Min Spec: The smallest eye height at which the DUT has to pass the BER test to
meet the specification.
Column 4:
Margin: Margin between the Min Passed Eye Height and the Min Spec.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
229
4
Computer PCI Express Bus Test Application
4.2.5 Gen3 CEM Tests
4.2.5.1 Calibration
4.2.5.1.1
Common Calibration Parameters
Scope Connection:
All calibrations can be done with either a differential probe or a single ended direct
connection on two channels, with one exception: The Common Mode Optimization
can only be done with a differential probe.
4.2.5.1.2
Equalization Preset Calibration
Purpose and Method:
The PCIe3 Base Specification defines several combinations of pre-shoots and deemphasis values. In this calibration presets P0 to P9 are calibrated at the end of the
cables which are later plugged into the SMP connectors of Rx lane 1 of the
compliance board riser card or the compliance base board. After calibrating one
combination of pre-shoot and de-emphasis the differential voltage is adjusted to
800 mV at this calibration point. The calibration for each pre-shoot and de-emphasis
combination is done in several iterations.
Figure 4-138: Connection diagram M8020A
230
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-139: Connection diagram J-BERT-B
Parameters in Expert Mode:
Differential Voltage: The target differential voltage for all presets. Should be 800mV
Used Calibrations: none
Keysight N5990A Test Automation Software Platform for PCIe User Guide
231
4
Computer PCI Express Bus Test Application
Figure 4-140: Result description
Column1:
Preset Number:
Column2:
Pre-Shoot: The target pre-shoot.
Column3:
De-Emphasis: The target de-emphasis.
Column4:
Differential Voltage: The target differential voltage.
Column5:
Set Pre-Shoot: The Pre-shoot which needs to be set to obtain the target value.
Column6:
Set De-Emphasis: The de-emphasis which needs to be set to obtain the target.
Column6:
Set Differential Voltage: The differential voltage which needs to be set to obtain the
target.
232
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.1.3
4
Equalization Custom Preset Calibration
Purpose and Method:
The calibration is nearly the same as the Equalization Preset Calibration. The
difference is that the user can specify custom combinations of pre-shoot and deemphasis.
Connection Diagram:
The same as for “Equalization Preset Calibration”
Parameters in Expert Mode:

Differential Voltage: The target differential voltage for all presets. Should be
800mV.

Custom Presets: A list of custom defined presets.
Used Calibrations: none
Figure 4-141: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
233
4
Computer PCI Express Bus Test Application
Column1:
Custom Preset Number:
Column2:
Pre-Shoot: The target pre-shoot.
Column3:
De-Emphasis: The target de-emphasis.
Column4:
Differential Voltage: The target differential voltage.
Column5:
Set Pre-Shoot: The Pre-shoot which needs to be set to obtain the target value.
Column6:
Set De-Emphasis: The de-emphasis which needs to be set to obtain the target.
Column7:
Set Differential Voltage: The differential voltage which needs to be set to obtain the
target.
234
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.1.4
4
Random Jitter Calibration
Purpose and Method:
In the Rx tests the input signal will be stressed with a combination of jitter sources to
simulate the possible impairments expected at the Rx input when operating in a
target system. Random jitter is added to simulate the effects of thermal noise. Due to
system intrinsic jitter the effective jitter level is different from the value set in the data
generator, so the jitter amplitude is calibrated.
The test automation starts with 0 UI of Random Jitter and increases it by “Jitter Step
Size” until “Stop Jitter”. It measures the resulting random jitter amplitude for every
set value.
Connection Diagram:
The same as for “Equalization Preset Calibration”
Parameters in Expert Mode:

Number of averages for jitter measurement: To reduce the noise floor the
differential signal has to be averaged when measured with the scope.

Stop Random Jitter: The maximum jitter that is calibrated.

Random Jitter Step Size: The amount the random jitter is increased in each
step.
Used Calibrations: None
Keysight N5990A Test Automation Software Platform for PCIe User Guide
235
4
Computer PCI Express Bus Test Application
Figure 4-142: Result description
236
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1: Set Random Jitter: The jitter amplitude set in the instrument
Column2: Measured Random Jitter.
4.2.5.1.5
Sinusoidal Jitter Calibration
Purpose and Method:
This procedure calibrates the sinusoidal jitter amplitude at two different frequencies
(5 MHz and 100 MHz). The test automation starts with 0 ps of SJ and increases that
value in leanear steps. For each step the procedure measures the actual amplitude
for both frequencies.
Available for following hardware configurations: all
Connection Diagram: The same as for “Equalization Preset Calibration”
Parameters in Expert Mode:

CDR loop-bandwidth.

Number of averages for jitter measurement: To reduce the noise floor the
differential signal has to be averaged when measured with the scope.
Used Calibrations: none
Keysight N5990A Test Automation Software Platform for PCIe User Guide
237
4
Computer PCI Express Bus Test Application
Figure 4-143: Result description
238
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1: Set Jitter: The jitter amplitude set in the instrument.
Column 2: SJ (x frequency): The measured jitter amplitude at x frequency.
4.2.5.1.6
DM Sinusoidal Interference Calibration
Purpose and Method:
When the J-BERT N4903B is used as the data generator Differential Mode Sinusoidal
Interference (DMSI) is generated with the Interference Module (Option J20). It is
coupled into to the data signal with pick off tees.
When the J-BERT M8020A is used as the data generator DMSI is generated with an
internal source.
The resulting DMSI amplitude at the test point is measured with a real-time
oscilloscope.
The data output of the de-emphasis signal generator is set to 0V for this calibration.
Available for following hardware configurations:
Keysight N5990A Test Automation Software Platform for PCIe User Guide
239
4
Computer PCI Express Bus Test Application
Figure 4-144: Connection diagram J-BERT-B M8020A
240
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-145: Connection diagram J-BERT-B
Parameters in Expert Mode: None.
Used Calibrations: None
Keysight N5990A Test Automation Software Platform for PCIe User Guide
241
4
Computer PCI Express Bus Test Application
Figure 4-146: Result description
Column 1: Set DM SI Amplitude: The DMSI amplitude set in the instrument
Column 2: Measured DM SI Amplitude: The measured DMSI amplitude
242
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.1.7
4
Eye Height and Width Calibration
Purpose and Method:
This procedure calibrates eye-height and eye-width by adding random jitter and
differential mode sinusoidal interference (DMSI). Starting with “Start DMSI” the
Jitter is increased with equally spaced steps from “Start RJ” to “Stop RJ” and the
eye-height and eye-width are measured. This procedure is now repeated for all
remaining DMSI amplitudes.
Available for following hardware configurations: All

Connection Diagram: The same as for “DM Sinusoidal Interference
Calibration”
Figure 4-147: Connection diagram J-BERT-B M8020A System
Keysight N5990A Test Automation Software Platform for PCIe User Guide
243
4
Computer PCI Express Bus Test Application
Figure 4-148: Connection diagram J-BERT-B system
Parameters in Expert Mode:

Optimize CTLE: Select if CTLE optimization is used. Default is false.

RJ for CTLE optimization: Random jitter set if CTLE optimization is used.

DMSI for CTLE optimization: DMSI set if CTLE optimization is used.

CTLE Index: CTLE index if CTLE optimization is not used.

Start DMSI: The first set DMSI value.

Stop DMSI: The last set DMSI value.

Number of DMSI Steps: The number of DMSI steps.

Start RJ: The first set RJ value.

Stop RJ: The last set RJ value.

Number of RJ Steps: The number of RJ steps.

Number of Averages: To reduce the noise floor the differential signal has to
be averaged when measured with the scope.
Used Calibrations:

Equalization Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration
244
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-149: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
245
4
Computer PCI Express Bus Test Application
Column 1: Set DM Interference: The DMSI set in the instrument
Column x+1: Eye Height (x RJ): The measured eye height for x RJ.
Figure 4-150: Result description
246
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column 1: Set DM Interference: The DMSI set in the instrument
Column x + 1: Eye Width (x RJ): The measured eye width for x RJ.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
247
4
Computer PCI Express Bus Test Application
4.2.5.1.8
Compliance Eye Calibration
Purpose and Method:
This procedure checks if it is possible to generate an eye-height and an eye- width
that meet the specification by adding Random Jitter and Differential Mode Sinusoidal
Interference. The method starts with default RJ and DMSI values and checks if the
obtained eye-height and eye-width are the target values. If they are not, RJ and
DMSI are recalculated with an algorithm that uses the difference between the
measured and the target values of the eye amplitudes. The procedure is repeated
until the target values are found or until the “Max Number of Search Steps” is
reached. If the “Max Number of Search Steps” is reached it checks if the optimum
combination of RJ and DMSI tested meets the specification.
Connection Diagram: The same as for “Eye Height and Width Calibration.”
Parameters in Expert Mode:

Target Eye-Height: Read-only. Max spec value - 1mV.

Target Eye-Width: Read-only. Max spec value - 5ps.

Max Number of Search Steps: Max number of times that the RJ and DMSI
can be recalculated to get the target values.

Number of Averages: To reduce the noise floor the differential signal has to
be averaged when measured with the scope.
Used Calibrations:

Equalization Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height and Width Calibration
248
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-151: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
249
4
Computer PCI Express Bus Test Application
Column 1: DMSI: Optimum DMSI added to the signal.
Column 2: RJ: Optimum RJ added to the signal.
Column 3: Eye Width: Eye width measured.
Column 4: Eye Height: Eye height measured.
250
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.1.9
4
Compliance Eye Verification
Purpose and Method:
This test verifies that it is possible to generate a signal with the eye-width and eyeheight according the CTS for RX compliance tests. The procedure uses the
Compliance Eye Calibration data to find the Random Jitter and Differential Mode
Sinusoidal Interference combination that should generate that eye. Then the eye is
measured using TestSig software.
It is also possible to verify non-compliance values by changing the parameter
“Compliance Mode” to false. In this case the Eye Height and Eye Width Calibration
data is used to find the necessary RJ and DSMI combination.
Connection Diagram: The same as for “Eye Height and Width Calibration.”
Parameters in Expert Mode:

Compliance Mode: If true all other parameter are set according to the CTS
for RX compliance tests.

Target Eye Height: Eye height desired.

Target Eye Width: Eye height desired.

Differential Mode Sinusoidal Interference: DMSI necessary to generate the
target eye height and eye width according the calibrations. Read Only.

Random Jitter: RJ necessary to generate the target eye according the
calibrations. Read Only.

Number of Averages: Maximum number of times that the procedure tries to
get the target eye height and eye width.
Used Calibrations:

Equalization Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height and Width Calibration

Compliance Eye Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
251
4
Computer PCI Express Bus Test Application
Figure 4-152: Result description
Column 1: Eye-Height: Simulated Eye Height
Column 2: Eye-Width: Simulated Eye Height
252
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.2.5.2 Receiver Tests
4.2.5.2.1
Common Rx Tests Parameters
Data Rate Specific:
1
Data rate deviation

Deviation in ppm that will be applied to the generator data rate. For
Add-in Cards only.
2
Link Training

Use Custom Training Voltage:
Set a custom differential voltage for link training only. The remaining
part of the test procedure will use the defined training voltage.

Custom Voltage: Differential voltage to use for the link training only.

Link Training Mode:
o JBERT Link Training: J-BERT sends a loopback training
sequence to the DUT.
o JBERT Interactive Link Training: The link training is
performed using the internal feature of the M8020A JBERT, skipping the link equalization phase.
o Vendor_Specific: The vendor can place his DUT in
loopback with his own tools like JTEC, I2C…
o DUT Target Preset: Preset that the M8020A J-BERT will
request the DUT to use. Only available if the link training
mode is set to JBERT Interactive Link Training.

Link Training Suite Setting File: The Link Training Suite settings file
(script file) which will be used for loopback training.


3
Default Link Training Lane Number for every Lane: The lane number
which will be encoded in the TS1s/TS2s sent to the DUT during
loopback training. This value will be set by default for all lanes. If Auto
each lane will be encoded with its own number. This parameter is
available only if JBERT_Link_Training is selected.
Suppress Loopback Training Messages: All loopback related message
boxes, e.g. “Please power cycle DUT” are not displayed. The user has
to assure that the DUT is in loopback when starting Rx tests.
Error Detector

Use CDR: If true, CDR is used to generate a clock signal for the JBERT error detector. If false, the clock is supplied by the J-BERT data
generator. With CDR the J-BERT error detector performance is better.





Loop Bandwidth: The loop bandwidth of the J-BERT error detector
CDR.
Enable SSC tracking: Enable SSC tracking if DUT transmits with SSC.
Only for J-BERT N4903B.
Filter Gen1/Gen2 SKPOs for BER test: Ignore SKP symbol in the error
detector. Only for J-BERT N4903B.
Peaking: CDR peaking. Only for J-BERT M8020A.
Analyzer Equalization: Apply PCI-Express Equalization settings to the
received signal. Only for J-BERT M8020A.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
253
4
Computer PCI Express Bus Test Application


4
Capture and Compare Mode: If enabled, the received data is captured
and saved in a pattern. A new analyzer sequence is generated with a
single block containing the captured pattern. This mode is only
available with a common reference clock architecture.
Pause before Auto-Align: Pause before the BER measurement so that
the user can perform manual optimization of the DUT receiver.
BER Measurement

Relax Time: Timespan between the point the stress signal is changed
and the BER measurement begins.
Lane Specific

Use Calibrated Preset.

Preset: Calibrated preset added to the signal.

Pre-Shoot: Pre-shoot added to the signal the calibrated preset is not used.
254

De-Emphasis: De-emphasis added to the signal if the calibrated preset is
not used.

Link Training Lane Number: The lane number which will be encoded in the
TS1/TS2s sent to the DUT during loopback training. This parameter is
available only if JBERT_Link_Training is selected.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.2.2
4
EQ Coefficient Matrix Scan
Purpose and Method:
This procedure measures the BER with the combination of coefficients C+1 (Precursor) and C-1 (Post-cursor) to create a co-efficient matrix with the BER results. At
each step, the BER value is measured for different values of C+1 coefficient while the
C-1 coefficient value is kept constant. The result values are mapped on to a
triangular matrix where each element contains four entries (measured BER, preshoot, de-emphasis, and boost). Elements on a diagonal line from bottom left to top
right have the same maximum boost value. The elements of the matrix are displayed
in different colors depending on the measured BER values. If the element appears in
green color, the entries value are valid and they can be used for testing. As the color
reaches to red they are invalid for testing.
If the parameter “Allow user to enter optimum equalization for remaining tests” is set
to “True” a window appears to let the user select the values of pre-shoot and deemphasis from the result graph. These selected values are used as the initial values
of pre-shoot and de-emphasis for the EQ Pre-Shoot De-Emphasis Scan.
Figure 4-153: Connection diagram M8020A addincard common clock architecture no external ref clock
Keysight N5990A Test Automation Software Platform for PCIe User Guide
255
4
Computer PCI Express Bus Test Application
Figure 4-154: Connection diagram M8020A addincard common clock architecture external ref clock internal clock multiplier
256
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-155: Connection diagram M8020A addincard separate clock architecture
Keysight N5990A Test Automation Software Platform for PCIe User Guide
257
4
Computer PCI Express Bus Test Application
Figure 4-156: Connection diagram M8020A system common clock architecture
258
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-157: Connection diagram M8020A system separate clock architecture
Keysight N5990A Test Automation Software Platform for PCIe User Guide
259
4
Computer PCI Express Bus Test Application
Figure 4-158: Connection diagram J-BERT-B addincard common clock architecture no external ref clock
260
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-159: Connection diagram J-BERT-B addincard common clock architecture external ref clock internal clock multiplier
Keysight N5990A Test Automation Software Platform for PCIe User Guide
261
4
Computer PCI Express Bus Test Application
Figure 4-160: Connection diagram J-BERT-B addincard common clock architecture external ref clock N4880 clock multiplier
262
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-161: Connection diagram J-BERT-B addincard separate clock architecture
Keysight N5990A Test Automation Software Platform for PCIe User Guide
263
4
Computer PCI Express Bus Test Application
Figure 4-162: Connection diagram J-BERT-B system common clock architecture internal clock multiplier
264
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-163: Connection diagram J-BERT-B system common clock architecture N4880 clock multiplier
Keysight N5990A Test Automation Software Platform for PCIe User Guide
265
4
Computer PCI Express Bus Test Application
Figure 4-164: Connection diagram J-BERT-B system separate clock architecture
266
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Parameters in Expert Mode:
1
Loopback Training

Force Retraining at each BER measurement: Force retraining at each
BER measurement for different PS/DE combination.

Pre-Shoot used for LB Training.

De-Emphasis used for LB Training.
2
Eye Parameter

Eye-Height: The eye height for the compliance test.

Eye-Width: The eye width for the compliance test.

Differential Mode Sinusoidal Interference: The amount of DMSI
which needs to be added to the signal to achieve the desired eyeheight and eye width in combination with the RJ.

Random Jitter: The amount of RJ which needs to be added to the
signal to achieve the desired eye-height and eye-width in combination
with the DMSI.
Sinusoidal Jitter: 12.5 ps of SJ has to be added for this procedure.


3
Sinusoidal Jitter Frequency: The frequency of the sinusoidal jitter
component.
Coefficient Variation

Test Calibrated Presets only: If true tests only calibrated presets (P0P9). If false scans the whole coefficient matrix with uncalibrated preshoot and de-emphasis values.

Coefficient Divider: The coefficient divider for C-1 and C+1.



Maximum Boost: Coefficient c+1 is increased until this boost level is
exceeded.
Start Pre-Shoot: Start pre-shoot value in dB.
Start De-Emphasis: Start de-emphasis value in dB.
4
BER Measurement

BER Mode: The BER measurements can be executed for a fixed time or
until a target BER is achieved.

Target BER: The Target BER for the BER measurements used in the
procedure.

Confidence Level: The confidence level value when BER mode is
TargetBer.
5
Equalization for remaining Rx Tests
Allow user to enter optimum equalization for remaining Rx tests.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
267
4
Computer PCI Express Bus Test Application

Eye Height and Width Calibration
Figure 4-165: Result description
268
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.2.3
4
Equalization Scan
Purpose and Method:
The purpose of this test is to find the optimum combination of de-emphasis and preshoot amplitude. As a first step, the procedure sets initial de-emphasis and preshoot values and adjusts the eye height to obtain the desired BER (slightly above 1e9). Then it keeps the initial pre-shoot and makes a de-emphasis scan, measuring the
BER for every de-emphasis value. After that it keeps the initial de-emphasis
amplitude and makes a pre-shoot scan. Finally the test shows the result tables, one
for the de-emphasis scan and one for the pre-shoot scan. Those let the user see the
best combination with the selected initial values.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:
1
Parameter

Scan Order: Select if de-emphasis or pre-shoot is tested first.

Initial De Emphasis: The initial de-emphasis used for BER adjustment
and pre-shoot scan.

Initial Pre Shoot: The initial pre-shoot used for BER adjustment and
de-emphasis scan.
Force training at each Preset: If true, every time the de-emphasis or
the pre-shoot is changed the DUT is trained into loopback again.


Common Mode Sinusoidal Interference: Common Mode Sinusoidal
Interference amplitude.
2
De-Emphasis Variation

Start De Emphasis: Start amplitude for the de-emphasis scan.

Stop De Emphasis: Stop amplitude for the de-emphasis scan.

De Emphasis Step Size: Step size for the de-emphasis scan.
3
Pre-Shoot Variation

Start Pre Shoot: Start amplitude for the pre-shoot scan.

Stop Pre Shoot: Stop amplitude for the pre-shoot scan.

De Pre Shoot Size: Step size for the pre-shoot scan.
4
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.




5
Target BER: The Target BER for the BER measurement used in the
test.
Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Equalization for remaining Rx Tests
Keysight N5990A Test Automation Software Platform for PCIe User Guide
269
4
Computer PCI Express Bus Test Application
Allow user to enter optimum equalization for remaining Rx tests.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height and Width Calibration
270
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-166: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
271
4
Computer PCI Express Bus Test Application
De Emphasis: Set De-emphasis amplitude.
BER for De Emphasis Scan: measured BER.
272
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-167: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
273
4
Computer PCI Express Bus Test Application
Pre Shoot: Set Pre Shoot amplitude.
BER for Pre Shoot Scan: measured BER.
274
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.2.4
4
Preset Compliance Test 2.8
Purpose and Method:
This test determines if the DUT meets the receiver specifications for different presets.
Eye height, eye width and sinusoidal jitter are set to the specified values. Eye height
and eye width are generated adding the adequate amount of random jitter and
DMSI. The procedure measure the number of errors during “BER Measurement
duration” and checks if the “Target BER” is satisfied. In this procedure presets P7 and
P8 are tested.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:

Enable Impairments for Loopback Training: Jitter and DMSI are enabled or
disabled during loopback training.

Eye Height: Eye height for the compliance test.

Eye Width: Eye width for the compliance test.

Differential Mode Sinusoidal Interference: The amount of DMSI which needs
to be added to achieve the desired eye height and eye width in combination
with RJ.





Random Jitter: The amount of RJ which needs to be added to achieve the
desired eye height and eye width in combination with DMSI.
Sinusoidal Jitter: 12.5ps of SJ has to be added for this test.
Sinusoidal Jitter Frequency: Frequency of the sinusoidal jitter component.
BER Measurement Duration: The duration of the BER measurement.
Target BER: The Target BER for the BER measurement used in the test.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height and Width Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
275
4
Computer PCI Express Bus Test Application
Figure 4-168: Result description
Column1:
Pass/Fail Result: The BER measured should be smaller than the Target BER.
Column 2:
Preset: Preset tested.
Column 3:
Target BER: Max BER allowed to pass the test.
Column4:
Measured BER.
276
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Parameters in Expert Mode:

Compliance Mode: If true all other parameters are set according to the CTS
for Rx compliance tests.

Enable Impairments for Loopback Training: Jitter and DMSI are enabled or
disabled during loopback training.

Eye Height: Eye height for the compliance test.

Eye Width: Eye width for the compliance test.






Differential Mode Sinusoidal Interference: The amount of DMSI which needs
to be added to achieve the desired eye height and eye width in combination
with RJ.
Random Jitter: The amount of RJ which needs to be added to achieve the
desired eye height and eye width in combination with DMSI.
Sinusoidal Jitter: 12.5 ps of SJ has to be added for this test.
Sinusoidal Jitter Frequency: Frequency of the sinusoidal jitter component.
BER Measurement Duration: Test duration.
Target BER: The Target BER for the BER measurement used in the test.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height and Width Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
277
4
Computer PCI Express Bus Test Application
Figure 4-169: Result description
Column1:
Pass/Fail Result: The BER measured should be smaller than the Target BER.
Column 2:
Preset: Preset tested.
Column 3:
Target BER: Max BER allowed to pass the test.
Column4:
Measured BER.
278
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.5.2.5
4
Compliance Tests 2.8 and 2.9
Purpose and Method:
The test is nearly the same as the Preset Compliance Test but only for preset P7. The
Compliance Test 2.8 corresponds to Add-in Cards and the Compliance Test 2.9
corresponds to System Boards.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:

Enable Impairments for Loopback Training: Jitter and DMSI are enabled or
disabled during loopback training.

Eye Height: Eye height for the compliance test.

Eye Width: Eye width for the compliance test.

Differential Mode Sinusoidal Interference: The amount of DMSI which needs
to be added to achieve the desired eye height and eye width in combination
with RJ.

Random Jitter: The amount of RJ which needs to be added to achieve the
desired eye height and eye width in combination with DMSI.
Sinusoidal Jitter: 12.5 ps of SJ has to be added for this test.
Sinusoidal Jitter Frequency: Frequency of the sinusoidal jitter component.





BER Mode: The BER measurement can be executed for a fixed time or until
a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the test.
Confidence Level: The confidence level value when BER mode is TargetBer.

BER Measurement Duration: The duration of the BER measurement when
mode is FixedTime.

Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height and Width Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
279
4
Computer PCI Express Bus Test Application
Figure 4-170: Result description
280
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:
Pass/Fail Result: The BER measured should be smaller than the Target BER.
Column 2:
Allowed Bit Errors: The allowed number of bit errors with the BER measurement.
Column 3:
Received Bit: Number of received bits during BER measurement.
Column4:
Measured BER.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
281
4
Computer PCI Express Bus Test Application
4.2.5.2.6
Receiver Jitter Tolerance
Purpose and Method:
This test characterizes how much jitter a DUT can tolerate at different frequencies of
sinusoidal jitter. Starting with the “Min Frequency” the jitter is increased with equally
spaced steps until the BER test fails. The BER test fails when the number of error bits
is bigger than 0. This procedure is repeated for all SJ frequencies.
Connection Diagram:
The same as for “EQ Coefficient Matrix Scan.”
Parameters in Expert Mode:
1
Sinusoidal Jitter Variation









Frequency Mode: Specifies the distribution of the frequency points to
test. It can be:
Compliance Frequencies: The frequencies defined in the specification
for compliance testing.
Equally Spaced Frequencies.
User Defined Frequencies.
Single Frequency.
Frequency Scale: The results can be represented in logarithmic or
linear scaling. For “Equally Spaced Frequencies” the frequency points
depends on the chosen frequency scale.
Start Frequency: Start frequency for “Equally Spaced Frequencies”.
Stop Frequency: Stop frequency for “Equally Spaced Frequencies”.
Number of Frequency Steps: Number of frequency points for “Equally
Spaced Frequencies”.
Search Algorithm (HyteresisUp or Binary):
The Binary search is not recommended for
devices with a long recovery time.

Frequency Points: Frequency points for “User Defined Frequencies”.

Start Jitter Points: Start jitter amplitudes for “User Defined
Frequencies”. In the other frequency modes start jitter amplitude is
always 0.
Jitter Frequency: Single frequency point for “Single Frequency”.

2

Jitter Step Size: The size of the smallest sinusoidal jitter amplitude
step used to search the “Max passed jitter” at each frequency.

Show Min Failed Points: The results graph can show the minimum
jitter failed jitter in addition to the maximum passed jitter for each
tested frequency.
Parameter

282
Use Compliance RJ and DMSI Values: If true, the amount of RJ and
DMSI used during the compliance test is applied. Together with
12.5 ps SJ this results in the eye-height and eye-width specified for
the compliance test.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application

Random Jitter: The amount of RJ which needs to be added
to achieve the compliance eye height and eye width in
combination with DMSI.

Differential Mode Sinusoidal Interference: The amount of DMSI
which needs to be added to achieve the compliance eye height and
eye width in combination with RJ.
Force Retraining on Each Frequency: Re-train the DUT at each tested
jitter frequency

3
4
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.




Target BER: The Target BER for the BER measurement used in the
test..
Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
283
4
Computer PCI Express Bus Test Application
Figure 4-171: Result description
284
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:
Pass/Fail Result: Max passed Jitter should be bigger than Min Spec.
Column 2:
Sinusoidal Jitter Frequency: Test frequency points.
Column 3:
Max Passed Jitter: Max jitter that passed the Target BER.
Column 4:
Jitter Capability Test Setup: Max jitter that the hardware can generate.
Column 5:
Min Spec:
Min jitter that has to pass the test to meet the specification.
Column 6:
Margin: Margin between the max passed jitter and the specification.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
285
4
Computer PCI Express Bus Test Application
4.2.5.2.7
Receiver Sensitivity
Purpose and Method:
This test searches the minimum eye height at which the DUT passes the BER test.
The method starts with “Start Eye Height” and decreases with steps of “Step Size”.
The minimum passed value is the last test point that did not return an error. Eye
height is generated changing the Differential Mode Sinusoidal Interference, the
random jitter is fixed to the compliance value.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”
Parameters in Expert Mode:
1
Sensitivity Variation

Start Eye Height: The start eye height. This is the largest eye-height in
the tests.


2
Stop Eye Height: The stop eye height. This is the lowest eye-height in
the tests unless the DUT already fails at a higher eye-height.
Eye-Height Step Size: The amount the eye height is decreased each
step to search the “Min Passed Eye Height”.
Parameter

Random Jitter: The amount of RJ which needs to be added to achieve
the compliance eye height and eye width in combination with DMSI.

Sinusoidal Jitter: 12.5 ps of SJ has to be added for this test.






Sinusoidal Jitter Frequency: Frequency of the sinusoidal jitter
component.
BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the
test.
Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height Calibration
286
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-172: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
287
4
Computer PCI Express Bus Test Application
Column1:
Pass/Fail Result: The Min Passed Eye Height measured should be smaller than the
Min Spec.
Column 2:
Min Passed Eye Height: The minimum eye height at which the DUT passes the BER
test.
Column 3:
Min Spec: The minimum eye height at which the DUT has to pass the BER test to
meet the specification.
288
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
4.2.5.3 Link Equalization Receiver Tests
4.2.5.3.1
Common Link Equalization Rx Test Parameters
Data Rate Specific:
1
Data rate deviation

Deviation in ppm that will be applied to the generator data rate. For
Add-in Cards only.
2
Link Training

Link Training Suite Setting File: The Link Training Suite settings file
(script file) which will be used for loopback training.

3
Error Detector






3.
Suppress Loopback Training Messages: All loopback related message
boxes, e.g. “Please power cycle DUT” are not displayed. The user has
to assure that the DUT is in loopback when starting Rx tests.
Use CDR: If true, CDR is used to generate a clock signal for the JBERT error detector. If false, the clock is supplied by the J-BERT data
generator. With CDR the J-BERT error detector performance is better.
Loop Bandwidth: The loop bandwidth of the J-BERT error detector
CDR.
Peaking: CDR peaking. Only for J-BERT M8020A.
Analyzer Equalization: Apply PCI-Express Equalization settings to the
received signal. Only for J-BERT M8020A.
Capture and Compare Mode: If enabled, the received data is captured
and saved in a pattern. A new analyzer sequence is generated with a
single block containing the captured pattern. This mode is only
available with a common reference clock architecture.
Pause before Auto-Align: Pause before the BER measurement so that
the user can perform manual optimization of the DUT receiver.
BER Measurement

Relax Time: Time span between the point the stress signal is changed
and the BER measurement begins.
Lane Specific:

Initial Preset: The starting generator transmitter preset (for Add-in Cards
only),

DUT Tx Preset: the transmitter preset that the generator will request the
DUT to use.
Note: Lane 0 is the only available lane for Link Equalization Tests.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
289
4
Computer PCI Express Bus Test Application
4.2.5.3.2
Compliance Tests 2.10 and 2.11
Purpose and Method:
These tests use the interactive link training feature of the M8020A J-BERT to let the
DUT negotiate the generator transmitter preset that will be used.
The Compliance Test 2.10 (for Add-in Cards) can be divided in two phases. In the
first phase, the CBB rev.3 is used and the starting generator transmitter presets are
P7 and P8. In the second phase, the CBB rev.2 is used and the starting generator
transmitter presets are P1, P7 and P8.
The Compliance Test 2.11 (for System Boards) consists of one phase, using the CLB
rev.3.
Connection Diagram: The same as for “EQ Coefficient Matrix Scan”, depending on
the DUT and the CLB/CBB rev. required.
Parameters in Expert Mode:












Enable Impairments for Loopback Training: Jitter and DMSI are enabled or
disabled during loopback training.
Eye Height: Eye height for the compliance test.
Eye Width: Eye width for the compliance test.
Differential Mode Sinusoidal Interference: The amount of DMSI which needs
to be added to achieve the desired eye height and eye width in combination
with RJ.
Random Jitter: The amount of RJ which needs to be added to achieve the
desired eye height and eye width in combination with DMSI.
Sinusoidal Jitter: 12.5 ps of SJ has to be added for this test.
Sinusoidal Jitter Frequency: Frequency of the sinusoidal jitter component.
BER Mode: The BER measurement can be executed for a fixed time or until
a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the test.
Confidence Level: The confidence level value when BER mode is TargetBer.
BER Measurement Duration: The duration of the BER measurement when
mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height and Width Calibration
290
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-173: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
291
4
Computer PCI Express Bus Test Application
Column1:
Pass/Fail Result: The BER measured should be smaller than the Target BER.
Column 2:
Initial generator transmitter preset (used for Add-in Card only).
Column 3:
Final generator transmitter preset (requested by the DUT).
Column 4:
Final generator Pre-Shoot.
Column 5:
Final generator De-Emphasis.
Column 6:
Allowed bit errors.
Column 7:
Measured bit errors.
4.2.5.3.3
Receiver Jitter Tolerance
Purpose and Method:
This test characterizes how much jitter a DUT can tolerate at different frequencies of
sinusoidal jitter. After training the DUT into loopback mode, using the M8020A JBERT interactive link training feature, a BER test is performed. Starting with the “Min
Frequency” the jitter is increased with equally spaced steps until the BER test fails.
The BER test fails when the number of error bits is bigger than 0. This procedure is
repeated for all SJ frequencies.
Connection Diagram:
The same as for “EQ Coefficient Matrix Scan”, depending on the DUT and the
CLB/CBB revision required
Parameters in Expert Mode:
1
Sinusoidal Jitter Variation









292
Frequency Mode: Specifies the distribution of the frequency points to
test. It can be:
Compliance Frequencies: The frequencies defined in the specification
for compliance testing.
Equally Spaced Frequencies.
User Defined Frequencies.
Single Frequency.
Frequency Scale: The results can be represented in logarithmic or
linear scaling. For “Equally Spaced Frequencies” the frequency points
depends on the chosen frequency scale.
Start Frequency: Start frequency for “Equally Spaced Frequencies”.
Stop Frequency: Stop frequency for “Equally Spaced Frequencies”.
Number of Frequency Steps: Number of frequency points for “Equally
Spaced Frequencies”.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Search Algorithm (HyteresisUp or Binary):
The Binary search is not recommended for
devices with a long recovery time.





Frequency Points: Frequency points for “User Defined Frequencies”.
Start Jitter Points: Start jitter amplitudes for “User Defined
Frequencies”. In the other frequency modes start jitter amplitude is
always 0.
Jitter Frequency: Single frequency point for “Single Frequency”.
Jitter Step Size: The size of the smallest sinusoidal jitter amplitude
step used to search the “Max passed jitter” at each frequency.
Show Min Failed Points: The results graph can show the minimum
jitter failed jitter in addition to the maximum passed jitter for each
tested frequency.
2
Parameter

Use Compliance RJ and DMSI Values: If true, the amount of RJ and
DMSI used during the compliance test is applied. Together with
12.5 ps SJ this results in the eye-height and eye-width specified for
the compliance test.

Random Jitter: The amount of RJ which needs to be added
to achieve the compliance eye height and eye width in
combination with DMSI.

Differential Mode Sinusoidal Interference: The amount of DMSI
which needs to be added to achieve the compliance eye height and
eye width in combination with RJ.

Force Retraining on Each Frequency: Re-train the DUT at each tested
jitter frequency.
3
BER Measurement

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.

Target BER: The Target BER for the BER measurement used in the
test..

Confidence Level: The confidence level value when BER mode is
TargetBer.


BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.
Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
293
4
Computer PCI Express Bus Test Application
Figure 4-174: Result description
294
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Column1:
Pass/Fail Result: Max passed Jitter should be bigger than Min Spec.
Column 2:
Sinusoidal Jitter Frequency: Test frequency points.
Column 3:
Max Passed Jitter: Max jitter that passed the Target BER.
Column 4:
Jitter Capability Test Setup: Max jitter that the hardware can generate.
Column 5:
Min Spec:
Min jitter that has to pass the test to meet the specification.
Column 6:
Margin: Margin between the max passed jitter and the specification.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
295
4
Computer PCI Express Bus Test Application
4.2.5.3.4
Receiver Sensitivity
Purpose and Method:
This test searches the minimum eye height at which the DUT passes the BER test.
The DUT is trained into loopback mode using the M8020A J-BERT interactive link
training feature and a BER measurement is performed, starting with an eye height of
“Start Eye Height” and decreases with steps of “Step Size”. The minimum passed
value is the last test point that did not return an error. Eye height is generated
changing the Differential Mode Sinusoidal Interference, the random jitter is fixed to
the compliance value.
The same as for “EQ Coefficient Matrix Scan”, depending on the DUT and the
CLB/CBB revision required.
Parameters in Expert Mode:
1
Sensitivity Variation

Start Eye Height: The start eye height. This is the largest eye-height in
the tests.

Stop Eye Height: The stop eye height. This is the lowest eye-height in
the tests unless the DUT already fails at a higher eye-height.

Eye-Height Step Size: The amount the eye height is decreased each
step to search the “Min Passed Eye Height”.
2
Parameter


Random Jitter: The amount of RJ which needs to be added to achieve
the compliance eye height and eye width in combination with DMSI.
Sinusoidal Jitter: 12.5 ps of SJ has to be added for this test.

Sinusoidal Jitter Frequency: Frequency of the sinusoidal jitter
component.

BER Mode: The BER measurement can be executed for a fixed time or
until a target BER is achieved.
Target BER: The Target BER for the BER measurement used in the
test.



Confidence Level: The confidence level value when BER mode is
TargetBer.
BER Measurement Duration: The duration of the BER measurement
when mode is FixedTime.

Allowed Bit Error: The allowed number of bit errors during the BER
measurement when mode is FixedTime.
Used Calibrations:

Preset Calibration

Random Jitter Calibration

Sinusoidal Jitter Calibration

DM Sinusoidal Interference Calibration

Eye Height Calibration
296
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-175: Result description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
297
4
Computer PCI Express Bus Test Application
Column1:
Pass/Fail Result: The Min Passed Eye Height measured should be smaller than the
Min Spec.
Column 2:
Min Passed Eye Height: The minimum eye height at which the DUT passes the BER
test.
Column 3:
Min Spec: The minimum eye height at which the DUT has to pass the BER test to
meet the specification.
4.2.5.4 Link Equalization Transmitter Tests
4.2.5.4.1
Common Link Equalization Tx Test Parameters
Data Rate Specific:
1
2
Parameter

Deviation in ppm that will be applied to the generator data rate. For
Add-in Cards only.

EQ Preset Measurement: The user can select to run the preset
measurement of the captured waveforms locally using SigTest or the
N5393C PCI-Express scope application if a connection to it is
available.

Scope sampling rate: Sampling rate that will be used to capture the
waveforms.
Link Training



3
Suppress Loopback Training Messages: All loopback related message
boxes, e.g. “Please power cycle DUT” are not displayed. The user has
to assure that the DUT is in loopback when starting Rx tests.
Error Detector

Use CDR: If true, CDR is used to generate a clock signal for the JBERT error detector. If false, the clock is supplied by the J-BERT data
generator. With CDR the J-BERT error detector performance is better.

Loop Bandwidth: The loop bandwidth of the J-BERT error detector
CDR.
Peaking: CDR peaking. Only for J-BERT M8020A.

298
Generator Start Preset: The preset the generator will use at the
beginning of the link training. Available for Add-in Cards only.
EQ Transmitter Test Script File: The Link Training Suite settings file
(script file) which will be used for loopback training.

Analyzer Equalization: Apply PCI-Express Equalization settings to the
received signal. Only for J-BERT M8020A.

Capture and Compare Mode: If enabled, the received data is captured
and saved in a pattern. A new analyzer sequence is generated with a
single block containing the captured pattern. This mode is only
available with a common reference clock architecture.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application

4
Pause before Auto-Align: Pause before the BER measurement so that
the user can perform manual optimization of the DUT receiver.
Lane 0 is the only available lane for Link Equalization Tests
Keysight N5990A Test Automation Software Platform for PCIe User Guide
299
4
Computer PCI Express Bus Test Application
4.2.5.4.2
Compliance Test 2.3
Purpose and Method:
This test, for Add-in Cards only, uses the interactive link training feature of the
M8020A J-BERT. The M8020A J-BERT runs the link training, setting several initial
equalization transmitter presets on the DUT and skipping the link equalization phase.
Once the DUT is in loopback, the DUT signal is captured and analyzed to check
whether the DUT is using the preset requested by the M8020A J-BERT or not.
Connection Diagram:
Figure 4-176: Connection diagram Compliance Test 2.3
Used Calibrations:
1.
300
None.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-177: Result Description
Column1:
Pass/Fail Result: The measured Pre-Shoot and De-Emphasis must be within the
specification limits.
Column 2:
Initial DUT transmitter preset (set by the M8020A J-BERT).
Column 3:
Measured Pre-Shoot on the DUT waveform.
Column 4:
Pre-Shoot lower specification limit.
Column 5:
Pre-Shoot upper specification limit.
Column 6:
Measured De-Emphasis on the DUT waveform.
Column 7:
De-Emphasis lower specification limit.
Column 8:
De-Emphasis upper specification limit.
Column 9:
A comment is added to each test step if fails, explaining the reason.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
301
4
Computer PCI Express Bus Test Application
4.2.5.4.3
Compliance Tests 2.4 and 2.7
Purpose and Method:
These tests use the interactive link training feature of the M8020A J-BERT to train
the DUT into loopback mode, running the link equalization phase completely.
A certain initial transmitter preset is set to the DUT. A successful link training rises an
event which is used to capture the waveforms of the M8020A J-BERT and the DUT.
At that moment, the captured waveform from the M8020A J-BERT contains the
preset change request and the waveform from the DUT contains the
acknowledgement of that request plus the physical transition from the initial
transmitter preset to the requested preset.
The captured data is decoded and two timespans are calculated: one between the
request and the acknowledgement, and other between the request and the electrical
transition.
Finally, once the DUT is in loopback mode, a similar preset measurement as for the
Compliance Test 2.3 is performed.
The test is divided in two parts. In the first part, the M8020A-JBERT requests
transmitter presets. In the second part, the M8020A-JBERT requests the pre-cursor,
cursor and post-cursor reported by the DUT.
In the Compliance Test 2.4 (for Add-in Cards), the initial transmitter preset is set by
the M8020A J-BERT. In the Compliance Test 2.7 (for System Boards), the user has to
manually set the DUT initial transmitter preset.
Connection Diagram:
The same as for “Link Equalization Compliance Test 2.3”. For the Compliance Test
2.7, the Add-in Card, CBB and Riser Card are replaced with the System Board and
the CLB. The reference clock connection is similar to the previously described
connection setups involving a System Board.
Parameters in Expert Mode:



Max Number of Retries: in case the acquired data cannot be decoded,
the link training can be repeated in order to get new data.
Scope visible range: Waveform range which will be used at the
moment when the link equalization phase is performed.
Use CTLE: Apply a CTLE filter as defined in the PCI-Express
specification to the DUT’s waveform. If true, several DC gains can be
selected.
Used Calibrations:
1
302
None.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-178: Result Description
Keysight N5990A Test Automation Software Platform for PCIe User Guide
303
4
Computer PCI Express Bus Test Application
Column1:
Pass/Fail Result: Max passed Jitter should be bigger than Min Spec.
Column 2:
DUT Target Preset: the transmitter preset the M8020A-JBERT requests the DUT to
use.
Column 3:
Electrical response time: calculated timespan between the request from the M8020A
J-BERT and the physical preset transition on the DUT waveform.
Column 4:
Protocol response time: calculared timespan between the request from the M8020A
J-BERT and the acknowledgement.
Column 5:
Measured Pre-Shoot on the DUT waveform.
Column 6:
Pre-Shoot lower specification limit.
Column 7:
Pre-Shoot upper specification limit.
Column 8:
Measured De-Emphasis on the DUT waveform.
Column 9:
De-Emphasis lower specification limit.
Column 10:
De-Emphasis upper specification limit.
Column 11:
A comment is added to each test step.
304
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4.2.6
4
Rx Switch Matrix
When more than one lane needs to be tested it is useful to use the switch. That
avoids needing to change the connections a lot of times.
To use the switch for receiver tests check the option “Use Switch for Rx tests” in the
Station Configuration. Then the Bitifeye BIT-2100 will appear in the instruments list.
Pressing the button “Rx Switch configuration” bring up a window where the module
type (SP4T or SP6T) and the slot position for each Rx and Tx lane can be chosen.
Figure 4-179: Rx switch configuration
Keysight N5990A Test Automation Software Platform for PCIe User Guide
305
4
Computer PCI Express Bus Test Application
The switch should not affect the test results. Therefore when the switch is used in the
receiver tests it is also included in the calibrations. It will be connected between the
input signal and the Rx lanes:
Figure 4-180: Connection setup for random jitter calibration
306
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
The output of the Tx lanes is the point that needs to be calibrated, so the switch is
not used here. It is necessary to manually change the cables from each Tx lane to the
oscilloscope.
In calibrations where the board is not included, the switch output will be connected
directly to the oscilloscope.
Figure 4-181: Connection setup for sinusoidal jitter configuration
In receiver tests the switch is always connected between the input signal and the Rx
lanes of the board, and between the Tx lanes and the error detector input.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
307
4
Computer PCI Express Bus Test Application
Figure 4-182: Connection setup for compliance test
308
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Computer PCI Express Bus Test Application
4
Figure 4-183: Detailed connection of the switch
Keysight N5990A Test Automation Software Platform for PCIe User Guide
309
Keysight N5990A Test Automation Software Platform for PCIe
User Guide
5
Troubleshooting and
Support
5.1
5.1
Log List and File / 311
Log List and File
In the case of problems the Log List can often help identifying the root cause. To
activate the Log List, click on the Log List button. The log file can be accessed by
right-clicking within the Log List section as shown in Figure 5-1. Note that all log
information will be lost when the N5990A application is terminated unless the log file
is saved.
Figure 5-1: ValiFrame N5990A log list and file
5
Troubleshooting and Support
In case of persisting problems with an application, send the Log File with a problem
description to:

312
[email protected]
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Keysight N5990A Test Automation Software Platform for PCIe
User Guide
6
6.1
Appendix
6.1
Data Structure and Backup / 313
6.2
Remote Interface / 316
6.3
Controlling Loop Parameters and Looping Over Selected Tests / 324
6.4
IBerReader / 328
6.5
Main Power Switch Control / 331
Data Structure and Backup
6.1.1 ValiFrame Data Straucture
All ValiFrame internal data is saved in the application data folder:
"Documents and Settings\All Users\Application Data\BitifEye \ValiFrame" for
Windows XP or "ProgramData\BitifEye\ValiFrame" for Windows 7.
Windows hides the system folders by default. To make the application data folder
visible, the "Hidden Files and Folders" setting needs to be set to "Show hidden files
and folders" in the Windows file explorer > View settings.
The ValiFrame application data folder contains the following folders:

Images

Settings

Pattern

Calibrations

Tmp
6.1.1.1
Images
The "Images" folder contains the connection diagram images.
6
Appendix
6.1.1.2 Settings
The “Settings” folder contains the default setting file for the instrument and .vset files
which contains the changes to the default registry entries. For each application, a
sub folder is created and a ValiFrame.vset file is created in this sub folder as soon as
any ValiFrame setting is changed from its default. The settings files contain for
example the instrument connection setup.
6.1.1.3 Pattern
The Pattern folder contains the test pattern files. These are text files which contain
the pattern in hexadecimal format.
6.1.1.4
Calibrations
The calibration data is stored in the “Calibrations” folder. For each calibration
procedure at least one calibration file is stored. These files are text files and can be
imported into MS Excel.
6.1.1.5 Tmp
All temporary files are created in the Tmp folder. The sub folder "Results" contains
the Excel file of the final result of each calibration and test procedure. This is a safety
feature and these files are used for recovery in case the user forgot to save them.
314
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
6.1.2 ValiFrame Backup
Use the ValiFrame application data folder to save calibration data, modified test
pattern or settings for backup or transfer to another PC.
The files in the folders, “Images” and “Pattern” will be generated or if they already
exist, be overwritten during a ValiFrame installation. In the “Settings” folder, all
instrument settings are overwritten by the installation except the .vset files. In the
“Calibration” folder, all files are generated by the calibration procedures and will not
be overwritten by the installation. To compare or archive the calibration data, backup
the “Calibration” folder.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
315
6
Appendix
6.2
Remote Interface
6.2.1 Introduction
The N5990A ValiFrame remote interface allows ValiFrame functionality (such as test
setup information, calibration and test procedures, and results) to be accessed from
external programming environments, for example MS.NET/C#, VEE, LabView,
TestExec SL, or TestStand. The remote interface can thus be used to control N5990A
by external software. In typical use, a top-level external test sequencer takes
advantage of ValiFrame functionality.
If ValiFrame is to be used as a top-level test sequencer, the control of external
software is achieved with N5990A opt. 500, User Programming. Refer to the User
Programming Manual for details.
316
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
6.2.2 Interface Description
The ValiFrame functionality is accessible via ValiFrameRemote.dll. It contains a class
ValiFrameRemote in the BitifEye.ValiFrame.ValiFrameRemote namespace (see Figure
6-1). Its use is illustrated by the ValiFrameRemoteTester application. The source
code and the Visual Studio solution of this example are available on the BitifEye
support webpage. Using this interface requires that the ValiFrame dlls are either in
the same folder or the Windows Path variable contains the folder in which these dlls
are located.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
317
6
Appendix
Figure 6-1: Members of the valiframeremote class
318
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
6.2.3 Using the Remote Interface
1.
2.
3.
4.
5.
6.
Add the ValiFrameRemote.dll as a reference to the project.
Create an instance of the ValiFrameRemote class.
Call SetConfigurationFile(string filename), if it is needed. It is required only
when the station configuration file generated by the station configurator is
not to be used. This file format is same as the files generated by the station
configurator, which can be found in the Valiframe Application data folder
(Windows XP: C:\documents and settings\all users\application
data\bitifeye\valiframe\settings\<application name>\ValiFrame.vset,
or Windows 7:
c:\programdata\bitifeye\valiframe\settings\<application
name>\ValiFrame.vset). The station configuration files contain just
the differences to the registry. Refer to Figure 6-2 for more details.
By calling InitApplication(string applicationName), the instruments of the
selected Test Station (see section 3.1) are connected and initialized. The
applicationName is PciExpress.
Call either ConfigureApplication() or LoadProject(string filename) to
initialize the DUT properties and test procedures. The project file can be
generated with the ValiFrame User Interface and it contains the DUT
properties, the selected test procedures and the properties of each test
procedure.
Calling Configure Application() prompts a dialog for setting the DUT
properties.
The number and type of available test procedures can depend on the DUT properties!
7.
Get the list of available procedures with GetProcedures(out int[]
procedureIds, out string[] procedureNames[]).
8. Select procedures individually with SelectProcedures(int[] procedureIds) or
combined with Run(int[] procedureIds, out stringxmlResult).
9. Execute selected procedures by calling any of the Run functions given
below:
10. The Run(out string[]xmlResults) executes all selected procedures. The
results of all procedures executed are returned at the end of the execution
of all selected procedures.
The RunProcedure(int id, out string xmlResult) executes a single procedure
and returns the result in an xml formatted string.
The RunProcedures(int[] procedureIds, out string[] xmlResults) executes
the list of procedures given in the procedureIds array.
The StartRun() function returns immediately. It is mainly used for
event-driven programming. In this case the events StatusChanged()
and ProcedureCompleted() can be used to determine the actual status of
the ValiFrame sequencer and read the results. The
ProcedureCompleted() event
provides the ID and the xmlResult of the procedure completed. After the run the
xmlResults are also
available via the Result property.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
319
6
Appendix
<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<Folder name="ValiFrame">
<Folder name="Stations">
<Folder name="PCI Express 3 Station">
<Folder name="Instruments">
<Folder name="Instrument2">
<Property name="Address">TCPIP0::192.168.0.102::inst0::INSTR</Property>
<Property name="Timeout">00:01:00</Property>
<Property name="Description">CMSI (Common Mode Sinusoidal Interference)
Source</Property>
</Folder>
<Folder name="Instrument1">
<Property name="Offline">True</Property>
<Property name="Address">TCPIP0::192.168.0.133::inst0::INSTR</Property>
<Property name="Timeout">00:01:00</Property>
<Property name="Description">M8020A J-BERT with integrated jitter
sources for BER tests</Property>
<Property name="Dll">VFAgM8000.dll</Property>
</Folder>
<Folder name="Instrument3">
<Property name="Offline">True</Property>
<Property name="Address">TCPIP0::192.168.0.103::inst0::INSTR</Property>
<Property name="Timeout">00:05:00</Property>
<Property name="Description">Realtime scope for stress signal
calibration</Property>
<Property name="Dll">VFAgDso.dll</Property>
</Folder>
<Folder name="Instrument4">
<Property name="Offline">True</Property>
<Property name="Address">192.168.0.103</Property>
<Property name="Timeout">00:03:00</Property>
<Property name="Description">PciExpress Tx application running on
realtime scope</Property>
<Property name="Dll">VFAgN5393C.dll</Property>
</Folder>
<Folder name="Instrument5">
<Property name="Offline">True</Property>
<Property name="Address">192.168.0.104;username;password</Property>
<Property name="Timeout">00:01:00</Property>
<Property name="Description">Main power switch</Property>
<Property name="Dll">VFNetIo230B.dll</Property>
</Folder>
<Folder name="Instrument6">
<Property name="Offline">True</Property>
<Property name="Address">TCPIP0::192.168.0.5::5025::SOCKET</Property>
<Property name="Timeout">00:00:30</Property>
<Property name="Description">Switch Matrix for Rx Tests</Property>
<Property name="Dll">VFBit2100.dll</Property>
</Folder>
</Folder>
<Folder name="Properties">
<Property name="Show All Instruments">False</Property>
<Property name="System Configuration">Unknown</Property>
<Property name="Generator Type">JBERT M8020A</Property>
<Property name="Common Mode Interference Source"> 81150A</Property>
<Property name="Reference Clock Source">None</Property>
<Property name="Seasim Access">Executable</Property>
320
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
<Property name="Seasim Script Path">C:\Seasim</Property>
<Property name="Python 2.6 Path">C:\Python26</Property>
<Property name="Power Switch Type">NetIo230B</Property>
<Property name="Use switch for Tx tests">False</Property>
<Property name="Use switch for Rx tests">True</Property>
<Property name="Use External Reference Clock">False</Property>
<Property name="Switch Type for Rx Modules in Rx tests">SP4T</Property>
<Property name="Switch Type for Tx Modules in Rx tests">SP4T</Property>
<Property name="Switch Type for Modules in Tx tests">SP4T</Property>
<Property name="Rx + slot in the Switch">Slot 1</Property>
<Property name="Rx - slot in the Switch">Slot 2</Property>
<Property name="Tx + slot in the Switch">Slot 3</Property>
<Property name="Tx - slot in the Switch">Slot 4</Property>
</Folder>
<Folder name="Children" />
<Property name="Software Version">ValiFrame 1.0</Property>
</Folder>
</Folder>
<Folder name="Database">
<Folder name="Properties">
<Property name="Offline">True</Property>
<Property name="ApplicationServerHostname">127.0.0.1:8082</Property>
</Folder>
</Folder>
</Folder>
Figure 6-2: Example of a station configuration file
If the ValiFrame sequencer is called via a .NET GUI (System.Windows.Forms.Form),
the current status, the available procedures, and the procedure selection can be
shown and modified by passing a TreeView control via the ProductPreTreeView
property to the ValiFrame sequencer prior to the InitApplication() call. In this case,
the TreeView control directly shows which procedures were selected as well as the
procedure currently being processed during the run. At the end of each run, the
pass/fail result is given. Refer to the ValiFrameRemoteTester source code for more
details.
The log entries generated by the ValiFrame sequencer can be accessed via the
LogChanged() event. Each time the sequencer generates a log entry this event will
be broadcast. It is recommended that the user monitors this event and tracks the log
changes to identify problems during execution.
The procedures requiring interaction with the user will pop up dialog panels. For
example, each time a new connection between an instrument and the DUT is
necessary, the procedure will start to display pop-up windows with the required
connections. The dialog can be suppressed by attaching to the
ConnectionChangeRequired() event. In some cases, internal dialogs or message
boxes are also shown. For full automation without any user interaction, events must
be defined and implemented such that the controlling environment can react to all
dialog and message boxes without user input. Currently, how to handle these dialogs
has to be decided case by case.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
321
6
Appendix
6.2.4 Results Format
Each Procedure Run will produce an xml-formatted result string, which can be
accessed via the out parameters of the Run() functions or the Results property of the
ValiFrameRemote class. The result string starts with a summary, which contains the
procedure name, ID, result, and the time stamp of the procedure run (Figure 6-3):
<?xml version="1.0" encoding="utf-16"?>
<Test Results>
<Summary>
<ProcedureName>Jitter Tolerance Test 2 MHz SJ RBR Lane 0</ProcedureName>
<ProcedureID>400008</ProcedureID>
<Result>Passed</Result>
<DateTime>4/30/2009 11:29:14 AM</DateTime>
</Summary>
<DocumentElement>
<Parameters>
<Name>Number of Lanes</Name>
<Value>1</Value>
</Parameters>
<Parameters>
<Name>Spec. Version</Name>
<Value>1.1</Value>
</Parameters>
<Parameters>
<Name>ISI Amplitude</Name>
<Value>570 mUI</Value>
</Parameters>
<Parameters>
<Name>Step Mode</Name>
<Value>False</Value>
</Parameters>
<Parameters>
<Name>Parade DP621 Device</Name>
<Value>False</Value>
</Parameters>
</DocumentElement>
<Data>
<ColumnHeader>|Result|Jitter Freq.|Sin.-Jitter Amp.|Number of Errors|Min
Spec|Max Spec|Details|</ColumnHeader>
<Values>|pass|2000000|0.981|2|0|1000||</Values>
</Data>
</Test Results>
Figure 6-3: Result string format
The following part contains the list of parameters. These parameters may be changed
via the project file or the remote interface. The last part contains the test data. It
starts with the column header, followed by one or more data rows. The format is
similar to what is obtained in the Excel output if the same procedure is run via the
ValiFrame user interface. Each column name/value is separated by the pipe symbol
'|'.
322
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
Keysight N5990A Test Automation Software Platform for PCIe User Guide
6
323
6
Appendix
6.3
Controlling Loop Parameters and Looping Over Selected Tests
Often parameters such as temperatures or supply voltages need to be varied
systematically. A simple example would be repeating tests over a temperature range
from –10 to 30 °C to verify an operating temperature range. In this case, after the
tests have been run at –10 °C, the temperature of the climate chamber is increased
by the selected temperature step width, for example, 1°C. The tests are then
repeated at –9 °C. After the test execution, the temperature is incremented again and
the tests are rerun repeatedly until they are finally run at 30 °C. This repetitive
process is called looping. In this example, the temperature within a climate chamber
is the loop parameter. While the loop is executed, the test results have to be
documented for each loop parameter value. In practice, multiple loop levels might be
required, as shown in Figure 6-4.
Figure 6-4: Temperature and voltage sweeps using N5990A sequencer
324
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
As the loop parameters are typically customer-specific, N5990A permits a list of loop
parameters to be specified. N5990A supports:
1.
Looping over user-specified parameters or run tests with a single parameter
value.
2.
Defining a set of loop parameters and for each parameter a range of test
points.
3.
Using custom drivers to control instruments that are not part of the
ValiFrame Test Station (see section 3.1, Test Station Selection and
Configuration), e.g. climate chambers, ovens, and power supplies.
4.
Saving the results of each test together with the actual loop parameter
value independently of the results from the other runs.
5.
An overview of each run after the end of the test execution.
These features are provided by an interface called IVFEnvironmentalControl.
The definition of this interface is:
namespace BitifEye.ValiFrame.Instruments
{
public interface IVFEnvironmentalControl
{
string UserLabel { get; }
void Connect();
void Disconnect();
string[] GetParameterList();
string[] GetParameterValues();
void Init();
bool SetNextValue();
void SetToDefault();
}
}
The interface has to be implemented by a class EnvironmentalControl in a .NET dll
named EnvironmentalControl.dll, which then needs to be copied into the ValiFrame
Program Files Folder. ValiFrame will load this dll and call the function of the Interface
in the following order:
Keysight N5990A Test Automation Software Platform for PCIe User Guide
325
6
Appendix
6.3.1 Connect()
At startup of ValiFrame allows the implementation to load the instrument drivers and
connect to them.
6.3.2 SetToDefault()
After the Connect() call, the implementation should set all instruments with initial
values to set default values. It is recommended that the sequence is stated with
nominal values to ensure that the test setup is done properly. With this setting, the
first run will be done and the Init() call will not overwrite the values.
6.3.3 Init()
The function is used to initialize the instruments with start values at the beginning of
test sequence.
6.3.4 GetParameterList() and GetParameterValues()
These functions are used to get the parameter names and values lists and put them
into the result output of each test procedure.
6.3.5 SetNextValue()
If this function returns true at the end of each run over the selected test procedures,
ValiFrame will run the selected tests again. This function should get the next
parameter set, set the controlling instruments, and return true if a new set of
parameters is available.
326
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
6.3.5.1 Example
For a sweep over temperature starts at 20 °C, increasing the temperature by 2 °C at
each run, and ending at 40 °C, the function should increase the temperature of the
chamber and return true if 40 °C is not reached. If the next step is greater than 40 °C,
this function should return false. ValiFrame will end the test sequence in this case.
6.3.6 Disconnect()
It is called at the closing of ValiFrame. The driver should set the instruments to
default values and disconnect from the instruments. An example project is available
on the BitifEye webpage.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
327
6
Appendix
6.4
IBerReader
ValiFrame cannot integrate all possible instruments and custom interfaces to
communicate with the DUT. To overcome this problem, the customer can provide
a .NET DLL which implements the IBerReader interface. This DLL is used by
ValiFrame, and invoked during the test; the DLL then takes care of the instrument or
DUT communication.
To use this feature in ValiFrame PCIe, go to Configure DUT > Show Parameters
dialog, and change the property BER Reader to "Custom BER Reader." This option
will be only be available if the dll with name PcieCustomBerReader.dll is there in its
installation folder.
PCIe-specific calling conventions:

Connect(string)
The string parameter is an empty string by default. It can be changed by
setting the "Address" property in Confugure DUT > Show Parameters dialog.
This is used to do general initialization or start external programs if it is
required.

Disconnect()
This method will be called every time a test run is finished (after all selected
tests are done, not after each individual test).
It is used to clean up or shut down external programs, if applicable.

Init(string)
This will be called when the DUT needs to be put into a specific state.
The possible values are "8000000000 T/s", "5000000000 T/s," or
"2500000000 T/s" depending on the data rate.
If Clock arquitecture is “Separate Reference Clock No Ssc” add the
string “, SRNS” and if clock arquitecture is “Separate Reference Clock
Independent Ssc” add “, SRIS”
328
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
6.4.1 IBerReader Interface
using System;
using System.Collections.Generic;
using System.Text;
namespace BerReader
{
public interface IberReader
{
/// <summary>
/// This method is called to connect to your error reader.
/// </summary>
/// <param name="address">The address string can be used by your
implementation
/// to configure the connection to the BerReader interface</param>
void Connect(string address);
/// <summary>
/// This method is called to close the connection
/// </summary>
void Disconnect();
/// <summary>
/// This method will be called prior the individual tests to tell
the device
/// what mode is tested. This can be used to load appropriate
/// setups.
/// </summary>
///
<param name="mode"> configuration mode in which the DUT will be
tested. The possible valueas are 8000000000 T/s, 5000000000
T/s or 2500000000 T/s. If Clock arquitecture is Separate Ref
Clock No Ssc add the string “, SRNS” and if clock
arquitecture is Separate Ref Clock Independent Ssc add “,SRIS”
</param>
void Init(string mode);
/// <summary>
/// Is called at the beginning of the error measurement and allows
/// a reset for the DUT to be implemented.
/// </summary>
void ResetDut();
/// <summary>
/// Starts the counters. This method MUST reset all counters!
/// </summary>
void Start();
/// <summary>
/// Stop the DUT to read out the counters (see
/// GetReadCounterWithoutStopSupported()).
/// </summary>
void Stop();
Keysight N5990A Test Automation Software Platform for PCIe User Guide
329
6
Appendix
/// <summary>
/// This method returns counters, the 1 st counting the bits/frames/lines
/// or bursts and the 2nd one counting the errors detected by the
BerReader.
/// The automation software will compute the BER using the following
/// equation BER=errorCounter/bitCounter. In the case bitCounter = 0 even
when
/// the stimulus is sending data, this is also interpreted as fail.
/// </summary>
/// <param name="bitCounter">
received
Contains the number of bits
which
are
/// by the DUT. If it is not possible to count bits the value can also
contain
/// frames, or bursts. It is just a matter of the value defined as target
BER.
/// If it is not possible to get the number of bits/frames/bursts then
the
/// method can return a value of -1 and the automation software can
compute
/// the number of bits from the data rate and the runtime.</param>
/// <param name="errorCounter"> Total number of errors since the last
start.
/// </param>
void GetCounter(out double bitCounter, out double errorCounter);
/// <summary>
/// This method returns a Boolean value indicating whether the device
/// supports reading the counters while it is running. If this method
/// returns false, the device needs to be stopped to read the counters.
/// In this case the automation software will stop data transmission
/// before calling the GetCounter() function, and re-start data
transmission
/// again after reading the counter values.
/// </summary>
/// <returns> false if device needs to be stopped before reading the
counters,
/// true if the counters can be read on the fly.</returns>
bool GetReadCounterWithoutStopSupported();
/// <summary>
/// This property returns a number to multiply the value delivered by the
/// bitCounter in the GetCounter() function.
/// </summary>
Double NumberOfBitsPerFrame {set; get;};
/// This property returns the number of payload
/// bits in a frame used for the detection of the BER.
/// If i.e. the errorCounter in the GetCounter() function is just the
/// checksum error then this parameter is the number of the payload.
/// </summary>
double NumberOfCountedBitsPerFrame {set; get;};
}
}
330
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6.5
6
Main Power Switch Control
Intended to power ON/OFF automatically the DUT and run the loopback training
without user interaction.
The Main Power Switch Control can be selected as:

Manual


Netlo 230B. It is a PDU (Power Distribution Unit) that integrates one 230 V
input and four 230 V outlets which allow to connect virtually any 230 V
powered device)
SynaccessNP
If it is selected as Manual, the DUT has to be power cycle manually. A dialog asking
for power cycling the DUT, pops-up in the initialisation of each receiver test
procedure (See Figure 6-5).
Figure 6-5: Manual Power Cycle Dialog
The number of user interactions for Manual option is equal to the number of times
that the DUT need to be trained into loopback.
When it is selected as Netlo230B or SynaccsessNP, the DUT is power cycle
automatically. A dialog asking to check the connection between the power supply
and the power switch, pops-up in the first receiver test procedure executed
(See Figure 6-6).
Figure 6-6: Automatic Power Cycle Dialog
In this case, the number of user interactions (related with the power cycle) is one,
independently of the number of Rx tests and the number of times that a retraining is
required.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
331
6
Appendix
Some properties related with the remote controllable power switch can be selected in
the Parameters Dialog (See Figure 6-7).
Figure 6-7: Power Switch Parameters (I)
332
Keysight N5990A Test Automation Software Platform for PCIe User Guide
Appendix
6
The same properties can be selected in the Parameters Panel of the Main Windows
(See Figure 6-8).
Figure 6-8: Power Switch Parameters (II)
These configurable properties are:

Channel: This sets the channel number of the power switch channel which
is connected to the DUT.

On-Off Duration: This is the duration between turning the DUT off and then
turning it on again.

Setting Time: This is the wait time after the DUT is turned on and before the
test continues with loopback training.

Max Retries for LB Training: Maximum number of times that ValiFrame will
try to train the DUT into loopback mode. If it is not possible within these
tries the test will be aborted automatically. When Power Switch Automation
is unselected, ValiFrame asks the user to retry every time loopback fails.
Keysight N5990A Test Automation Software Platform for PCIe User Guide
333
This information is subject to change without notice.
© Keysight Technologies 2015
Edition 5.0, September 2015
www.keysight.com
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement