Agilent Technologies 89410A/89441A Operator’s Guide

Agilent Technologies Part Number 89400-90038

For instruments with firmware version A.08.00
Printed in U.S.A.

Print Date: May 2000

© Hewlett-Packard Company, 1993 to 2000. All rights reserved.
8600 Soper Hill Road Everett, Washington 98205-1209 U.S.A.

This software and documentation is based in part on the Fourth

Berkeley Software Distribution under license from The Regents of the
University of California. We acknowledge the following individuals and
institutions for their role in the development: The Regents of the
University of California .

Portions of the TCP/IP software are copyright Phil Karn, KA9Q.

i
 The Analyzer at a Glance

2 1 10 11

13

12

17
14 15
16

18

4 5 7 8

6 9

This illustration shows the Agilent 89441A Vector Signal Analyzer, which consists of
two components: the IF section on top and the RF section on bottom. The IF section is
the Agilent 89410A; the RF section is the Agilent 89431A. Note that you can order the
89431A to convert an 89410A into an 89441A (see Options and Accessories later in this
manual).

ii
Front Panel
1-A softkey’s function changes as different 11-Use the SYSTEM hardkeys and their
menus are displayed. Its current function is menus to control various system functions
determined by the video label to its left, on the (online help, plotting, presetting, and so on).
analyzer’s screen.
12-Use the MEASUREMENT hardkeys and
2-The analyzer’s screen is divided into two their menus to control the analyzer’s receiver
main areas. The menu area, a narrow column and source, and to specify other measurement
at the screen’s right edge, displays softkey parameters.
labels. The data area, the remaining portion of
the screen, displays traces and other data. 13-The REMOTE OPERATION hardkey and
LED indicators allow you to set up and monitor
3-The POWER switch turns the analyzer on the activity of remote devices.
and off. If you have an 89441A, you must turn
on the power switches on both the top and 14-Use the MARKER hardkeys and their
bottom box. menus to control marker positioning and
marker functions.
4-Use a 3.5 inch flexible disk (DS,HD) in this
disk drive to save your work. 15-The knob’s primary purpose is to move a
marker along the trace. But you can also use it
5-The KEYBOARD connector allows you to to change values during numeric entry, move a
attach an optional keyboard to the analyzer. cursor during text entry, or select a hypertext
The keyboard is most useful for writing and link in help topics
editing Agilent Instrument BASIC programs.
16-Use the Marker/Entry key to determine the
6- The SOURCE connector routes the knob’s function. With the Marker indicator
analyzer’s source output to your DUT. If you illuminated the knob moves a marker along the
have an 89441A, the Source output from the IF trace. With the Entry indicator illuminated the
section (top box) is connected to the RF knob changes numeric entry values.
section (bottom box). If the RF section has
option AY8 (internal RF source), the SOURCE 17-Use the ENTRY hardkeys to change the
output connector on the RFsection is a type-N ; value of numeric parameters or to enter
otherwise it is a BNC. numeric characters in text strings.

7-The EXT TRIGGER connector lets you 18-The optional CHANNEL 2 input connector
provide an external trigger for the analyzer. routes your test signal or DUT output to the
analyzer’s receiver. For ease of upgrading, the
8-The PROBE POWER connectors provides CHANNEL 2 BNC connector is installed even if
power for various Agilent active probes. option AY7 (second input channel) is not
installed.
9-The INPUT connector routes your test signal
or DUT output to the analyzer’s receiver. If For more details on the front panel,
you have an 89441A, the INPUT on the RF display the online help topic “Front
section (bottom box) is connected to the Panel”. See the chapter “Using
Online Help” if you are not familiar
CHANNEL 1 input on the IF section (top box).
with using the online help index.
10-Use the DISPLAY hardkeys and their
menus to select and manipulate trace data and
to select display options for that data.

iii
This page left intentionally blank

iv
Saftey Summary
The following general safety precautions must be observed during all phases of
operation of this instrument. Failure to comply with these precautions or with
specific warnings elsewhere in this manual violates safety standards of design,
manufacture, and intended use of the instrument. Agilent Technologies, Inc.
assumes no liability for the customer’s failure to comply with these requirements.

GENERAL
This product is a Safety Class 1 instrument (provided with a protective earth
terminal). The protective features of this product may be impaired if it is used in
a manner not specified in the operation instructions.

All Light Emitting Diodes (LEDs) used in this product are Class 1 LEDs as per
IEC 60825-1.

ENVIRONMENTAL CONDITIONS
This instrument is intended for indoor use in an installation category II,
pollution degree 2 environment. It is designed to operate at a maximum relative
humidity of 95% and at altitudes of up to 2000 meters. Refer to the
specifications tables for the ac mains voltage requirements and ambient
operating temperature range.

BEFORE APPLYING POWER

Verify that the product is set to match the available line voltage, the correct fuse
is installed, and all safety precautions are taken. Note the instrument’s external
markings described under Safety Symbols.

GROUND THE INSTRUMENT

To minimize shock hazard, the instrument chassis and cover must be connected
to an electrical protective earth ground. The instrument must be connected to
the ac power mains through a grounded power cable, with the ground wire
firmly connected to an electrical ground (safety ground) at the power outlet.
Any interruption of the protective (grounding) conductor or disconnection of
the protective earth terminal will cause a potential shock hazard that could
result in personal injury.

v
 FUSES
Only fuses with the required rated current, voltage, and specified type (normal
blow, time delay, etc.) should be used. Do not use repaired fuses or
short-circuited fuse holders. To do so could cause a shock or fire hazard.

DO NOT OPERATE IN AN EXPLOSIVE ATMOSPHERE

Do not operate the instrument in the presence of flammable gases or fumes.

DO NOT REMOVE THE INSTRUMENT COVER

Operating personnel must not remove instrument covers. Component
replacement and internal adjustments must be made only by qualified service
personnel.

Instruments that appear damaged or defective should be made inoperative and

secured against unintended operation until they can be repaired by qualified
service personnel.
WARNING The WARNING sign denotes a hazard. It calls attention to a procedure,
practice, or the like, which, if not correctly performed or adhered to,
could result in personal injury. Do not proceed beyond a WARNING
sign until the indicated conditions are fully understood and met.

Caution The CAUTION sign denotes a hazard. It calls attention to an operating

procedure, or the like, which, if not correctly performed or adhered to, could
result in damage to or destruction of part or all of the product. Do not proceed
beyond a CAUTION sign until the indicated conditions are fully understood and
met.

vi
Safety Symbols

Warning, risk of electric shock

Caution, refer to accompanying documents

Alternating current
Both direct and alternating current

Earth (ground) terminal

Protective earth (ground) terminal

Frame or chassis terminal

Terminal is at earth potential.

Standby (supply). Units with this symbol are not completely disconnected from ac mains when
this switch is off

vii
This page left intentally blank.

viii
 Options and Accessories: Agilent 89410A
To determine if an option is installed, press [System Utility] [option setup]. Installed
options are also listed on the analyzer’s rear panel. To order an option to
upgrade your 89410A, order 89410U followed by the option number.

To convert your 89410A DC-10 MHz Vector Signal Analyzer to an

89441A DC-2650 MHz Vector Signal Analyzer, order an 89431A. To order an
option when converting your 89410A to an 89441A, order 89431A followed by
the option number.

IMPORTANT To convert older HP 89410A analyzers (serial numbers below 3416A00617),

contact your nearest Agilent Technologies sales and service office.

Agilent 89410U Agilent 89430/

Option Description Opt 89431A Opt
Add Precision Frequency Reference AY5 —
Add Vector Modulation Analysis and Adaptive Equalization AYA AYA
Add Waterfall and Spectrogram AYB AYB
Add Digital Video Modulation Analysis and Adaptive Equalization AYH AYH
(requires option AYA and UFG or UTH)
Add Enhanced Data rates for GSM Evolution (EDGE) (requires B7A B7A
option AYA)
Add Digital Wideband CDMA Analysis B73 B73
(requires options AYA and UTH)
Add Digital ARIB rev 1.0-1.2 W-CDMA Analysis (requires option B79 B79
B73)
Add 3GPP version 3.1 W-CDMA Analysis 080 080
(requires options AYA and UTH)
Add Second 10 MHz Input Channel AY7 AY7
Extend Time Capture to 1 megasample AY9 AY9
Add 4 Megabyte Extended RAM and Additional I/O UFG (obsolete: order option UTH)
Add 20 Megabyte Extended RAM and Additional I/O UTH UTH
Add Advanced LAN Support (requires option UFG or UTH) UG7 UG7
Add Agilent Instrument BASIC 1C2 1C2
Add PC-Style Keyboard and Cable U.S. version 1F0 1F0
Add PC-Style Keyboard and Cable German version 1F1 1F1
Add PC-Style Keyboard and Cable Spanish version 1F2 1F2
Add PC-Style Keyboard and Cable French version 1F3 1F3
Add PC-Style Keyboard and Cable U.K. version 1F4 1F4
Add PC-Style Keyboard and Cable Italian version 1F5 1F5
Add PC-Style Keyboard and Cable Swedish version 1F6 1F6

ix
 continued on next page...

Option Description 89410U Opt 89431A Opt

Add Front Handle Kit AX3 AX3
Add Rack Flange Kit AX4 AX4
Add Flange and Handle Kit AX5 AX5
Add Extra Manual Set OB1 OB1
Add Extra Instrument BASIC Manuals OBU OBU
Add Service Manual OB3 OB3
Add Internal RF Source — AY8
Delete High Precision Frequency Reference — AY4
Add 50 - 75 Ohm Minimum Loss Pads — 1D7
Firmware Update Kit UE2 UE2

The accessories listed in the following table are supplied with the
Agilent 89410A.

Supplied Accessories Part Number

Line Power Cable See Installation and
Verification Guide
Standard Data Format Utilities 5061-8056
Agilent Technologies 89410A/89441A (see title page in manual)
Operator’s Guide
Agilent Technologies 89410A Getting (see title page in manual)
Started Guide
Agilent Technologies 89410A Installation and Verification Guide (see title page in manual)
Agilent Technologies 89400-Series GPIB Command Reference (see title page in manual)
GPIB Programmer’s Guide (see title page in manual)
Agilent Technologies 89400-Series GPIB Quick Reference (see title page in manual))
Coax BNC(m)-to-coax BNC(m) connector (with option AY5) 1250-1499

x
The accessories listed in the following table are available for the Agilent 89410A.

Available Accessories Part Number

Agilent 89411A 21.4 MHz Down Converter Agilent 89411A
89400-Series Using Instrument BASIC Agilent 89441-90013
Instrument BASIC User’s Handbook Agilent E2083-90005
Spectrum and Network Measurements Agilent 5960-5718
Box of ten 3.5-inch double-sided, double-density disks Agilent 92192A
Active Probe Agilent 41800A
Active Probe Agilent 54701A
Active Divider Probe Agilent 1124A
Resistor Divider Probe Agilent 10020A
Differential Probe (requires Agilent 1142A) Agilent 1141A
Probe Control and Power Module Agilent 1142A
50 Ohm RF Bridge Agilent 86205A
Switch/Control Unit Agilent 3488A
High-Performance Switch/Control Unit Agilent 3235A
GPIB Cable - 1 meter Agilent 10833A
GPIB Cable - 2 meter Agilent 10833B
GPIB Cable - 4 meter Agilent 10833C
GPIB Cable - 0.5 meter Agilent 10833D
HP Printer or Plotter (contact your local
Hewlett-Packard sales
representative)

xi
Options and Accessories: 89441A
To determine if an option is installed, press [System Utility] [option setup]. Installed
options are also listed on the analyzer’s rear panel. To order an option for an
Agilent 89441A analyzer, order Agilent 89441U followed by the option number.

Option Description Agilent 89441U Option

Add Internal RF Source AY8
Add High Precision Frequency Reference AYC
Add Vector Modulation Analysis and Adaptive Equalization AYA
Add Waterfall and Spectrogram AYB
Add Digital Video Modulation Analysis and Adaptive Equalization AYH
(requires options AYA and UFG or UTH)
Add Enh. Data rates for GSM Evol (EDGE) (requires option AYA) B7A
Add Digital Wideband CDMA Analysis B73
(requires options AYA & UTH)
Add Digital ARIB 1.0-1.2 W-CDMA Analysis (requires option B73) B79
Add 3GPP v 3.1 W-CDMA Analysis (requires opts AYA & UTH) 080
Add Second 10 MHz Input Channel AY7
Extend Time Capture to 1 megasample AY9
Add 4 Megabyte Extended RAM and Additional I/O UFG (obsolete: order option UTH)
Add 20 megabyte Extended RAM and Additional I/O UTH
Add Advanced LAN Support (requires option UFG or UTH) UG7
Add Agilent Instrument BASIC 1C2
Add 50 - 75 Ohm Minimum Loss Pads 1D7
Add PC-Style Keyboard and Cable U.S. version 1F0
Add PC-Style Keyboard and Cable German version 1F1
Add PC-Style Keyboard and Cable Spanish version 1F2
Add PC-Style Keyboard and Cable French version 1F3
Add PC-Style Keyboard and Cable U.K. version 1F4
Add PC-Style Keyboard and Cable Italian version 1F5
Add PC-Style Keyboard and Cable Swedish version 1F6
Add Front Handle Kit AX3
Add Flange and Handle Kit AX5
Add Extra Manual Set OB1
Add Extra Instrument BASIC Manuals OBU
Add Service Manual OB3

xii
 Firmware Update Kit UE2
The accessories listed in the following table are supplied with the
Agilent 89441A.

Supplied Accessories Part Number

Line Power Cable See Installation and
Verification Guide
Rear Panel Lock Foot Kit Agilent 5062-3999
BNC Cable - 12 inch Agilent 8120-1838
2 BNC Cables - 8.5 inch HP 8120-2682
Coax BNC(m)-to-coax BNC(m) Connector (deleted with option AY4) Agilent 1250-1499
Type N-to-BNC Adapter (2 with option AY8) Agilent 1250-0780
Serial Interface Interconnect Cable Agilent 8120-6230
Interconnect Cable EMI Suppressor Agilent 9170-1521
Standard Data Format Utilities Agilent 5061-8056
Agilent 89410/89441A Operator’s Guide (see title page in manual)
Agilent Technologies 89441A Getting Started Guide (see title page in manual)
Agilent Technologies 89441A Installation and Verification Guide (see title page in manual)
Agilent Technologies 89400-Series GPIB Command Reference (see title page in manual)
GPIB Programmer’s Guide (see title page in manual)
Agilent Technologies 89400-Series GPIB Quick Reference (see title page in manual)

xiii
The accessories listed in the following table are available for the Agilent 89441A.

Available Accessories Part Number

Agilent 89411A 21.4 MHz Down Converter Agilent 89411A
89400-Series Using Instrument BASIC Agilent 89441-90013
Instrument BASIC User’s Handbook Agilent E2083-90005
Spectrum and Network Measurements Agilent 5960-5718
Box of ten 3.5-inch double-sided, double-density disks Agilent 92192A
Active Probe Agilent 41800A
Active Probe Agilent 54701A
Active Divider Probe Agilent 1124A
Resistor Divider Probe Agilent 10020A
Differential Probe (requires Agilent 1142A) Agilent 1141A
Probe Control and Power Module Agilent 1142A
50 Ohm RF Bridge Agilent 86205A
Switch/Control Unit Agilent 3488A
High-Performance Switch/Control Unit Agilent 3235A
GPIB Cable - 1 meter Agilent 10833A
GPIB Cable - 2 meter Agilent 10833B
GPIB Cable - 4 meter Agilent 10833C
GPIB Cable - 0.5 meter Agilent 10833D
HP Plotters and Printers (contact your local
Hewlett-Packard sales
representative)

xiv
Notation Conventions
Before you use this book, it is important to understand the types of keys on
the front panel of the analyzer and how they are denoted in this book.

Hardkeys Hardkeys are front-panel buttons whose functions are always the same.
Hardkeys have a label printed directly on the key. In this book, they are printed like this:
[Hardkey].
Softkeys Softkeys are keys whose functions change with the analyzer’s current menu
selection. A softkey’s function is indicated by a video label to the left of the key (at the
edge of the analyzer’s screen). In this book, softkeys are printed like this: [softkey].
Toggle Softkeys Some softkeys toggle through multiple settings for a parameter.
Toggle softkeys have a word highlighted (of a different color) in their label. Repeated
presses of a toggle softkey changes which word is highlighted with each press of the
softkey. In this book, toggle softkey presses are shown with the requested toggle state
in bold type as follows:
“Press [key name on]” means “press the softkey [key name] until the selection on is active.”
Shift Functions In addition to their normal labels, keys with blue lettering also have a
shift function. This is similar to shift keys on an pocket calculator or the shift function
on a typewriter or computer keyboard. Using a shift function is a two-step process.
First, press the blue [Shift] key (at this point, the message “shift” appears on the
display). Then press the key with the shift function you want to enable.
Shift function are printed as two key presses, like this:
[Shift] [Shift Function]
Numeric Entries Numeric values may be entered by using the numeric keys in the
lower right hand ENTRY area of the analyzer front panel. In this book values which are
to be entered from these keys are indicted only as numerals in the text, like this:
Press 50, [enter]
Ghosted Softkeys A softkey label may be shown in the menu when it is inactive. This
occurs when a softkey function is not appropriate for a particular measurement or not
available with the current analyzer configuration. To show that a softkey function is not
available, the analyzer ‘’ghosts’’ the inactive softkey label. A ghosted softkey appears
less bright than a normal softkey. Settings/values may be changed while they are
inactive. If this occurs, the new settings are effective when the configuration changes
such that the softkey function becomes active.

xv
This page left intentionally blank.

xvi
In This Book
This book, “Agilent 89410A/Agilent89441A Operator’s Guide”, is designed to
advance your knowledge of the Agilent 89410A and Agilent 89441A Vector Signal
Analyzers. You should already feel somewhat comfortable with this analyzer, either
through previous use or through performing the tasks in either product’s Getting
Started Guide.” The book consists of both measurement tasks and concepts.

Measurement tasks
Measurement tasks provide step-by-step examples of how to perform specific tasks
with your Agilent 89410A or Agilent 89441A Vector Signal Analyzer. These tasks
may be similar to measurements you wish to make and you can modify them to meet
your own needs. Even if these tasks are not specifically related to your
measurement needs, you may find it helpful to perform the tasks anyway—they only
take a few minutes each—since they will help you become familiar with many of
your analyzer’s features.

Concepts
The concepts section provides you with a conceptual overview of the
Agilent 89410A and Agilent 89441A and their essential features. This section
assumes that you are already familiar with basic measurement concepts and is
helpful in understanding the similarities and differences between the Agilent 89400
series analyzers and other analyzers you may have used. The concepts are also
essential if you want to make the best use of the analyzer’s features.

To Learn More About the Agilent 89410A and Agilent 89441A

You may need to use other books in the Agilent 89400 series manual set. See the
“Documentation Roadmap” at the end of this book to learn what each book contains.

xvii
This page left intentionally blank.

xviii
 TABLE OF CONTENTS

1 Demodulating an Analog Signal

To perform AM demodulation 1-2
To perform PM demodulation 1-4
To perform FM demodulation 1-6

2 Measuring Phase Noise

To measure phase noise 2-2
Special Considerations for phase noise measurements: 2-3

3 Characterizing a Transient Signal

To set up transient analysis 3-2
To analyze a transient signal with time gating 3-4
To analyze a transient signal with demodulation 3-5

4 Making On/Off Ratio Measurements

To set up time gating 4-2
To measure the on/off ratio 4-4

5 Making Statistical Power Measurements

To display CCDF 5-2
To display peak, average, and peak/average statistics 5-4

6 Creating Arbitrary Waveforms

To create a waveform using a single, measured trace 6-2
To create a waveform using multiple, measured traces 6-4
To create a short waveform using ASCII data 6-6
To create a long waveform using ASCII data 6-8
To create a contiguous waterfall or spectrogram display 6-10
To create a fixed-length waterfall display 6-12
To determine number of samples and ∆t 6-14
To output the maximum number of samples 6-15

xix
7 Using Waterfall and Spectrogram Displays (Opt. AYB)
To create a test signal 7-2
To set up and scale a waterfall display 7-4
To select a trace in a waterfall display 7-6
To use markers with waterfall displays 7-8
To use buffer search in waterfall displays 7-10
To set up a spectrogram display 7-11
To enhance spectrogram displays 7-12
To use markers with spectrogram displays 7-14
To save waterfall and spectrogram displays 7-15
To recall waterfall and spectrogram displays 7-16

8 Using Digital Demodulation (Opt. AYA)

To prepare a digital demodulation measurement 8-2
To demodulate a standard-format signal 8-4
To select measurement and display features 8-5
To set up pulse search 8-6
To set up sync search 8-8
To select and create stored sync patterns 8-9
To demodulate and analyze an EDGE signal 8-10
To troubleshoot an EDGE signal 8-12
To demodulate and analyze an MSK signal 8-14
To demodulate a two-channel I/Q signal 8-16

9 Using Video Demodulation

(Opt. AYH)
To prepare a VSB measurement 9-2
To determine the center frequency for a VSB signal 9-4
To demodulate a VSB signal 9-6
To prepare a QAM or DVB QAM measurement 9-8
To demodulate a QAM or DVB QAM signal 9-10
To select measurement and display features 9-12
To set up sync search (QAM only) 9-13
To select and create stored sync patterns (QAM only) 9-14
To demodulate a two-channel I/Q signal 9-15

xx
10 Analyzing Digitally Demodulated Signals (Options
AYA and AYH)
To demodulate a non-standard-format signal 10-2
To use polar markers 10-4
To view a single constellation state 10-5
To locate a specific constellation point 10-6
To use X-axis scaling and markers 10-7
To examine symbol states and error summaries 10-8
To view and change display state definitions 10-10
To view error displays 10-12

11 Creating User-defined Signals (Options AYA and

AYH)
To create an ideal digitally modulated signal 11-2
To check a created signal 11-4
To create a user-defined filter 11-6

12 Using Adaptive Equalization (Options AYA and AYH)

To determine if your analyzer has Adaptive Equalization 12-2
To load the multi-path signal from the Signals Disk 12-3
To demodulate the multi-path signal 12-4
To apply adaptive equalization 12-6
To measure signal paths 12-8
To learn more about equalization 12-10

13 Using Wideband CDMA (Options B73, B79, and 080)

To view a W-CDMA signal 13-2
To demodulate a W-CDMA signal 13-4
To view data for a single code layer 13-6
To view data for a single code channel 13-8
To view data for one or more slots 13-10
To view the symbol table and error parameters 13-12
To use x-scale markers on code-domain power displays 13-14

14 Using the LAN (Options UTH & UG7)

To determine if you have options UTH and UG7 14-2
To connect the analyzer to a network 14-3
To set the analyzer’s network address 14-4
To activate the analyzer’s network interface 14-5
To send GPIB commands to the analyzer 14-6
To select the remote X-Windows server 14-7

xxi
 To initiate remote X-Windows operation 14-8
To use the remote X-Windows display 14-9
To transfer files via the network 14-10

15 Using the Agilent 89411A Downconverter

Connection and setup details for the Agilent 89411A 15-4
Calibration 15-8

16 Extending Analysis to 26.5 GHz with 20 MHz

Information Bandwidth
Overview 16-2
System Description 16-3
Agilent 89410A Operation 16-5
HP/Agilent 71910A Operation 16-5
Mirrored Spectrums 16-6
IBASIC Example Program 16-6
System Configuration 16-7
Agilent 89410A Configuration 16-7
HP/Agilent 71910A Configuration 16-8
System Connections 16-10
Operation 16-13
Controlling the Receiver 16-14
Changing Center Frequency 16-14
Setting the Mirror Frequency Key 16-14
Changing the Reference Level 16-14
Resolution Bandwidth 16-16
DC Offset and LO Feedthrough 16-16
Calibrating the System 16-17
Calibration Methods 16-18
DC Offset 16-19
Channel Match 16-20
IQ Gain, Delay Match 16-20
Quadrature 16-22

17 Choosing an Instrument Mode

Why Use Scalar Mode? 17-2
Why Use Vector Mode? 17-4
Why Use Analog Demodulation Mode? 17-6
The Advantage of Using Multiple Modes 17-7
Scalar—the big picture 17-7

xxii
 Vector—the important details 17-7
Analog Demodulation—another view of the details 17-7
Instrument Mode? Measurement Data? Data Format? 17-8
Instrument modes 17-8
Measurement data 17-8
Data format 17-8
Unique Capabilities of the Instrument Modes 17-9

18 What Makes this Analyzer Different?

Time Domain and Frequency Domain Measurements 18-2
The Y-axis (amplitude) 18-3
The X-axis (frequency) 18-3
What are the Different Types of Spectrum Analyzers? 18-4
Swept-tuned spectrum analyzers 18-4
Real-time spectrum analyzers 18-5
Parallel-filter analyzers 18-5
FFT analyzers 18-6
The Difference 18-8
Vector mode and zoom measurements 18-8
Stepped FFT measurements in Scalar mode 18-9

19 Fundamental Measurement Interactions

Measurement Resolution and Measurement Speed 19-2
Resolution bandwidth 19-2
Video filtering 19-3
Frequency span 19-3
Bandwidth coupling 19-4
Flexible bandwidth mode 19-4
Display resolution and frequency span 19-5
Windowing 19-6
General 19-6
Windows used with this analyzer 19-6
Enhancing the Measurement Speed 19-8
Digital storage 19-9
Zero response and DC measurements 19-9
Special Considerations in Scalar Mode 19-10
Sweep time limitations 19-10
Stepped measurements 19-10
The relationship between frequency resolution and display resolution
19-11

xxiii
 Resolution bandwidth limitations 19-11
What is a detector and why is one needed 19-12
Manual sweep 19-13
Special Considerations in Vector Mode 19-14
Time data 19-15
The time record 19-16
Why is a time record needed? 19-16
Time record, span and resolution bandwidth 19-17
Measurement speed and time record length 19-17
How do the parameters interact? 19-18
Time record length limitations 19-19
Time record processing 19-20

20 Analog Demodulation Concepts

What is Analog Demodulation? 20-2
Applications 20-2
Using analog demodulation for zero span measurements 20-2
How Does Analog Demodulation Work
in the Agilent 89400 Series Analyzer? 20-3
Special Considerations for Analog Demodulation 20-4
Time Correction and Analog Demodulation 20-5
The Importance of Span Selection 20-6
Including all important signal data 20-6
Checking for interfering signals 20-7
The Importance of Carrier Identification 20-8
Auto carrier with AM demodulation 20-8
Auto carrier with PM demodulation 20-8
Auto carrier with FM demodulation 20-8
Special considerations for auto carrier use 20-8
AM Demodulation Specifics 20-9
The algorithm 20-9
PM Demodulation Specifics 20-10
The algorithm 20-10
Auto carrier off 20-10
Auto carrier on 20-10
FM Demodulation Specifics 20-12

xxiv
 The algorithm 20-12
Interactions with other features 20-13
Choosing trigger type with analog demodulation 20-13
Using gating and averaging with analog demodulation 20-13
Two-channel measurements and analog demodulation 20-13

21 Gating Concepts
What is Time Gating? 21-2
How Does it Work? 21-4
Important Concepts 21-5
Parameter Interactions 21-6

22 Digital Demodulation Concepts (Opt. AYA)

Overview 22-2
What you learn in this chapter 22-2
If you need background references 22-2
What this analyzer does 22-2
Measurement Flow 22-4
General block diagram 22-4
Digital Demodulator Block diagram (except FSK) 22-5
Digital Demodulator Block diagram: FSK 22-6
Measurement management 22-8
Measurement and display choices 22-8
Carrier locking 22-9
I-Q measured signal 22-10
I-Q reference signal 22-10
Special considerations for FSK demodulation 22-10
Parameter interactions 22-11
Span considerations 22-11
Data size considerations 22-12
Resolution bandwidth 22-12
Display limitations 22-12
Feature Availability in Digital Demod 22-13
Special considerations for sync search 22-14
Special considerations for pulsed signals 22-15
Speed and resolution considerations 22-16
Maximizing speed - measurement and display 22-16
Maximizing resolution 22-16
Filtering 22-17
General information 22-17

xxv
 Filter choices for the measured and reference signals 22-17
Square-root raised cosine filters 22-18
Raised cosine filters 22-18
Gaussian filter 22-19
Low pass filter (for FSK) 22-19
User defined filters 22-19
IS-95 Filters 22-20
EDGE Filter 22-21
EDGE (winRC) Filter 22-21

23 Video Demodulation Concepts (Opt. AYH)

Overview 23-2
What you learn in this chapter 23-2
What option AYH does 23-2
Measurement Flow 23-3
General block diagram 23-3
Digital demodulator block diagram: QAM and DVB QAM 23-4
Digital demodulator block diagram: VSB 23-6
Measurement management 23-7
Measurement and display choices 23-7
Carrier locking (all except VSB) 23-8
Carrier locking and pilot search: VSB 23-9
Input Range 23-10
I-Q measured signal 23-10
I-Q reference signal 23-10
Parameter interactions 23-11
Data size considerations 23-11
Resolution bandwidth 23-11
Span considerations 23-12
Display limitations 23-13
Feature Availability in Video Demodulation 23-14
Special considerations for sync search 23-15
Special considerations for pulsed signals 23-16
Maximizing speed - measurement and display 23-16
Maximizing resolution 23-16
Filtering 23-17
General information 23-17

xxvi
24 Wideband CDMA Concepts (Options B73, B79, and
080)
Overview 24-2
What you learn in this chapter 24-2
What option B73 does 24-2
What option B79 does 24-3
What option 080 does 24-3
Measurement Flow 24-4
Setting up a W-CDMA Measurement 24-6
Signal Connections and Input Range 24-6
Frequency Span 24-7
Center Frequency 24-7
Scramble Code 24-7
Chip Rates, Code Layers, and Symbol Rates 24-8
Main Length 24-9
Filtering 24-9
Mirrored Spectrums 24-9
Time-Domain Corrections 24-9
Trigger Signal 24-10
Viewing Measurement Results 24-11
Code-Domain Power Displays 24-12
Time-Domain Displays 24-13
Time Gating 24-14
Parameter interactions 24-15
Data size considerations 24-15
Resolution bandwidth 24-15
Points Per Symbol 24-15
Feature Availability in W-CDMA 24-16
Troubleshooting W-CDMA Measurements 24-17
Index
Need Assistance?
Documentation Road Map

xxvii
1

Demodulating an
Analog Signal
This chapter shows how to demodulate AM, FM, and PM signals using the
Analog Demodulation instrument mode. In these examples the signals are
provided by the Signals Disk which accompanies this documentation.

1-1
 Demodulating an Analog Signal

To perform AM demodulation

The following procedure demonstrates demodulation using files on the

signals disk that you load into the analyzer’s data registers and use as
arbitrary source signals. The sample signal is a 5 MHz carrier that is
amplitude modulated with a sine wave.

1. Initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Load the source signal file AMSIG.DAT into data register D1:
Insert the Signals Disk in the analyzer’s disk drive.
Press [Save/Recall], [default disk], [internal disk] to select the internal disk drive.
Press [Return] (bottom softkey), [catalog on]to display the files on the disk.
Rotate the knob until the file AMSIG.DAT is highlighted.
Press [recall trace], [from file into D1], [enter].
3. Connect the SOURCE output to the channel 1 INPUT.
4. Turn on the source and select arbitrary signal D1 (25 kHz sine modulating
5 MHz):
Press [Source], [source on], [source type], [arbitrary]
5. Set the frequency span:
Press [Frequency], [span], 500, [kHz].
The display should now appear as shown below.

Spectrum of the AM signal.

1-2
 Demodulating an Analog Signal

6. Turn on AM demodulation and examine the recovered modulation signal:

Press [Instrument Mode], [Analog Demodulation ] (with option AYH, press
[Instrument Mode], [demod type], [Analog Demodulation], [Return]).
Press [demodulation setup], [ch1 result], [AM].
Press [Auto Scale] to scale the display information, as shown below:

The AM demodulated spectrum.

7. Examine the recovered time-domain information:
Press [Measurement Data], [main time] ([main time ch1] in a 2-channel analyzer) to
display the time data.
Press [Auto Scale] to scale the display information.
Press [Trigger], [trigger type], [internal source] to stabilize the display.

Recovered signal is volts as a function of time.

l The span value sets the effective sample rate (∆t) and range of allowed
RBW values.
l The RBW value determines the record length (T).
l The number of measured points = T/∆t.

1-3
 Demodulating an Analog Signal

To perform PM demodulation

The following procedure demonstrates PM demodulation using a file on the

signals disk that you load into the analyzer’s data registers and use as an
arbitrary source signal. The sample signal is a 5 MHz carrier that is phase
modulated with a triangle wave.

1. Initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Load the source signal file PMSIG.DAT into data register D2:
Insert the Signals Disk in the analyzer’s disk drive.
Press [Save/Recall], [default disk], [internal disk] to select the internal disk drive.
Press [Return] (bottom softkey), [catalog on]to display the files on the disk.
Rotate the knob until the file PMSIG.DAT is highlighted.
Press [recall trace], [from file into D2], [enter].
3. Connect the SOURCE output to the channel 1 INPUT.
4. Turn on the source and select arbitrary signal D2
(a 25 kHz triangle wave modulating a 5 MHz carrier):
Press [Source], [source on], [source type], [arb data reg], [D2], [Return], [arbitrary].
5. Set the frequency span:
Press [Frequency], [span], 500, [kHz].
The display should now appear as shown below.

Spectrum of the phase modulated signal.

1-4
 Demodulating an Analog Signal

6. Turn on demodulation (PM) and examine the recovered modulation signal:

Press [Instrument Mode], [Analog Demodulation ] (with option AYH, press
[Instrument Mode], [demod type], [Analog Demodulation], [Return]).
Press [demodulation setup], [ch1 result], [PM].
Press [Auto Scale] to scale the display information.
The display should now appear as shown below.

The PM demodulated spectrum.

7. Examine the recovered time-domain information:
Press [Measurement Data], [main time] ([main time ch1] in a 2-channel analyzer)
to display the time data.
Press [Trigger], [trigger type], [internal source] to stabilize the display.
Press [Auto Scale] to scale the display information.
Press [Display], [more display setup], [grids off ].
The display should now appear as shown below.

The recovered signal is in radians as a function of time.

1-5
 Demodulating an Analog Signal

To perform FM demodulation

The following procedure demonstrates FM demodulation using a file on the

signals disk that you load into the analyzer’s data registers and use as an
arbitrary source signal. The sample signal is a 5 MHz carrier that is phase
modulated with a triangle wave.
1. Initialize the analyzer:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Load the source signal file PMSIG.DAT into data register D2:
Insert the Signals Disk in the analyzer’s disk drive.
Press [Save/Recall], [default disk], [internal disk] to select the internal disk drive.
Press [Return] (bottom softkey), [catalog on]to display the files on the disk.
Rotate the knob until the file PMSIG.DAT is highlighted.
Press [recall trace], [from file into D2], [enter].
3. Connect the SOURCE output to the channel 1 INPUT.
4. Turn on the source and select arbitrary signal D2
(a 25 kHz triangle wave modulating a 5 MHz carrier):
Press [Source], [source on], [source type], [arb data reg], [D2], [Return], [arbitrary].
5. Set the frequency span:
6. Press [Frequency], [span], 500, [kHz].
The display should now appear as shown below.

Spectrum of the phase modulated signal.

7. Select FM demodulation:

1-6
 Demodulating an Analog Signal

Press [Instrument Mode], [Analog Demodulation ] (with option AYH, press

[Instrument Mode], [demod type], [Analog Demodulation], [Return]).
Press [demodulation setup], [ch1 result], [FM].
Press [Auto Scale] to scale the display information.
The display should now appear as shown below.

The FM demodulated spectrum.

8. Examine the recovered signal:
Press [Measurement Data], [main time] ([main time ch1] in a 2-channel analyzer)
to display the time data.
Press [Trigger], [trigger type], [internal source] to stabilize the display.
Press [Auto Scale] to scale the display information.
Press [Display], [more display setup], [grids off].
The display should now appear as shown below.

The recovered signal is in Hz as a function of time.

The recovered time data is displayed as a square wave because the relation
between phase modulation and frequency modulation is a derivative and
the derivative of a triangle wave is a square wave. Compare this recovered
FM signal with the recovered PM signal in the previous task.

1-7
2

Measuring Phase Noise

This chapter demonstrates how to perform a phase noise measurement
using a simulated input signal from a time capture signal.

2-1
 Measuring Phase Noise

To measure phase noise

The following procedure demonstrates phase noise measurements using

files on the signals disk that you load into the analyzer’s time-capture
buffer. Then, instead of measuring data on the input, we analyze the data
in the capture buffer. Reading data from the capture buffer is not
normally part of measuring phase noise, but is necessary to document the
procedure with simulated measurement data.

The settings that constitute a phase noise measurement:

l Instrument Mode: PM demodulation
l Measurement Data: PSD (power spectral density)
l Data Format: x axis is log scale
l Averaging, as required

1. Initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. If your analyzer has the optional second input channel installed, turn it off:
Press [Input], [channel 2], [ch2 state off].
3. Load the time-capture data into the capture buffer:
Insert the Signals Disk in the analyzer’s disk drive.
Press [Save/Recall], [default disk], [internal disk] to select the internal disk drive.
Press [Return] (bottom softkey), [catalog on]to display the files on the disk.
Rotate the knob until the file CAPT.DAT is highlighted.
Press [recall more], [recall capture buffer], [enter] (takes about 2 minutes).
Using time-captured data causes the analyzer to automatically select the
capture buffer data instead of the input channel and sets the center
frequency, span, and resolution bandwidth to those used when the data
was captured. In a normal phase noise measurement you will set these
parameters instead of loading captured data.

2-2
 Measuring Phase Noise

1. Select PM demodulation:
Press [Instrument Mode], [Analog Demodulation ] (with option AYH, press [Instrument
Mode], [demod type], [Analog Demodulation], [Return]).
Press [demodulation setup], [ch1 result], [PM].
2. Select PSD measurement data:
Press [Measurement Data], [PSD] ([PSD ch1] for a 2-channel analyzer).
3. Set the x-axis scale to log:
Press [Data Format], [x-axis log].
4. Turn on averaging:
Press [Average], [average on], [num averages], 100, [enter], [fast avg on].
5. Run (or start) the measurement and scale the results:
Press [Meas Restart] to make the measurement.
Press [Auto Scale] to scale the trace data.
The display should now appear as shown below.

Phase Noise Plot

Special Considerations for phase noise measurements:
l This is a measurement of S0 which is defined as the power in both sidebands of
the phase noise. L(f) is typically defined as the power in one sideband.
l RBW must be small enough so that the low frequency portion of the log X-axis is
valid. RBW should be less than the start frequency.
l Note that the span in demodulation mode is one-half the span of the instrument.

2-3
3

Characterizing a
Transient Signal
This chapter demonstrates two methods of characterizing a transient signal.
In this case you will characterize a simulated transmitter turn-on signal.

3-1
 Characterizing a Transient Signal

To set up transient analysis

This procedure demonstrates several transient signal characterization

methods. The signal is loaded from disk into a data register and selected
as an arbitrary source signal. The signal simulates a transmitter turning on.
1. Select the baseband mode and initialize the analyzer:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Load the source signal file XMITR.DAT into data register D1:
Insert the Signals Disk in the analyzer’s disk drive.
Press [Save/Recall], [default disk], [internal disk] to select the internal disk drive.
Press [Return] (bottom softkey), [catalog on] to display the files on the disk.
Rotate the knob until the file XMITR.DAT is highlighted.
Press [recall trace], [from file into D1], [enter].
3. Connect the SOURCE output to the channel 1 INPUT.
4. Turn on the source and select arbitrary signal D1:
Press [Source], [source on], [source type], [arbitrary].
5. Select a window and increase the display resolution by increasing the number
of frequency points:
Press [ResBW/Window], [main window], [gaussian top]
Press [Return], [num freq pts], 801, [enter] (or use the increment key).
6. Set the display to show spectrum and time information:
Press [Display], [2 grids].
Press [B],[Mea surement Data], [main time] (toggle to ch1 for 2-channel
analyzer).

3-2
 Characterizing a Transient Signal

7. Set the sweep and trigger:

Press [Trigger], [trigger type], [internal source].
Press [Sweep], [single], [Pause|Single] to simulate a “transient.”
Press [Auto Scale].
The display should now appear as shown below.

Spectrum (top) and time domain representation (bottom) of transient signal.

3-3
 Characterizing a Transient Signal

To analyze a transient signal with time gating

This procedure assumes that the steps in “To set up transient analysis”
have been performed. If not, do so before continuing.
1. Turn on time gating, set the gate length, and set up the knob to move the gate:
Press [Time], [gate on ], [gate length], 3, [us], [ch1 gate dly].
Press the [Marker|Entry] hardkey so that the knob’s “Entry” LED is on.
2. Rotate the knob to move the gate over the very first part of the transient signal
appearing in the lower trace; see the plot below.
3. Set up the marker to show the movement of the spectrum’s peak:
Press [A], [Marker Function], [peak track on].
Press [Shift], [Marker⇒] (turns offset marker on and zeros it).
4. Now, move the time gate across the transient signal’s time display (by turning
the knob) and note the movement of the spectrum peak:
Press [Time], and make sure [ch1 gate dly ] is selected (if not, press it).
Press [Marker|Entry] if the knob’s “Entry” LED is not on.

Rotate the knob, moving the gate further to the right into the transient
signal and stop long enough for the spectrum to update. Then move it
again and stop. The reference marker (square) remains at the location of
the “transient start up” making it easier to see the carrier movement as
the regular marker (diamond) tracks the peak. Marker readouts in the
display pictured below show that early in the transient there is as much
as 1.00 MHz variation in carrier frequency.

With time gating on, the spectrum shown (top) is that of the data inside
the gate markers (bottom). In this case, moving the time gate across the
time signal (bottom) shows that the carrier frequency varies with time
(the spectral peak moves). We can use FM demod to show this, too.
(grids were turned off in the illustration to highlight the gate markers.)

3-4
 Characterizing a Transient Signal

To analyze a transient signal with demodulation

This procedure analyzes the frequency and amplitude variations of the

transient signal with demodulation. It assumes that the steps in “To set up
transient analysis” have been performed. If not, do so before beginning.
1. If you just finished the setup procedure, go to step 2 (don’t perform this step).
If you just finished the time gating analysis, go back to main time:
Press [Time], [gate off]
2. Turn on FM demodulation:
Press [Instrument Mode], [Analog Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Analog Demodulation], [Return]).
Press [demodulation setup], [ch1 result], [FM].
Press [Pause|Single] to simulate a transient and take data.
3. Now rescale to examine the results:
Press [A], [Auto Scale].
Press [B], [Ref Lvl/Scale], [Y per div], 200, [exponent], 3, [Hz].
The display should appear as shown below.

FM demodulation analysis. The bottom trace shows the frequency

variation in the transient signal. Comparing it with the time signal
in the previous figure shows that demod results are meaningless during
periods of no signal.

3-5
 Characterizing a Transient Signal

4. Now we’ll look at the amplitude response of the signal with AM demodulation:
Press [Instrument Mode], [demodulation setup], [ch1 result], [AM].
Press [Pause|Single].
5. Now scale both traces:
Press the blue [Shift], [A]. Traces A and B should be active (both LEDs
on).
Press [Auto Scale] to automatically scale the active traces.
The display should now appear as shown below.

AM demodulation analysis (bottom trace). Note that there is

more ringing (more cycles before settling) in the amplitude
of the transient signal than was seen in the frequency analysis
(compare bottom trace of this figure with that of FM demodulation
in the previous figure)

3-6
4

Making On/Off Ratio

Measurements
This chapter shows you how to measure the on/off ratio of a burst signal.
This type of signal is typical in communication applications which use a
burst carrier. You will use a signal from the Signals Disk to simulate a
phase-modulated burst carrier.

4-1
 Making On/Off Ratio Measurements

To set up time gating

1. Select the baseband mode and initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Load a burst signal from the Signals Disk into a register and play it through the
source:
Insert the Signals Disk in the internal disk drive.
Connect the SOURCE to the channel 1 INPUT.
Press [Save/Recall], [default disk], [internal disk].
Press [Return], [catalog on].
Rotate the knob to highlight PMBURST.DAT
Press [recall trace], [from file into D1], [enter].
Press [Source], [source on], [source type], [arbitrary].
3. Select a window:
Press [ResBW\Window], [main window], [gaussian top].

The display should now appear as below.

A 5 MHz burst carrier, phase-modulated with a 25 kHz signal.

4-2
 Making On/Off Ratio Measurements

4. Turn on a second trace and configure it to display stable time data:

Press [Display], [2 grids].
Press [Trigger], [trigger type], [internal source].
Press [B].
Press [Measurement Data], [main time ]
([main time ch1] for a 2-channel analyzer).
Press [Auto Scale].
5. Set up a gate to encompass the first burst:
Press [Time], [gate on], [ch1 gate dly].
Pre ss [Marker|Entry] to turn on the Entry LED.
Rotate the knob to align the left gate marker with the beginning of the
first burst.
Press [gate length]
Rotate the knob to align the right gate marker with the end of the first
burst.
6. Turn off the grids to highlight the gate markers:
Press [Display], [more display setup], [grids off]
The display should now appear as below.

The lower trace displays the time domain signal with a gate
encompassing the first burst. The upper trace displays the
frequency spectrum of the gated burst.

4-3
 Making On/Off Ratio Measurements

To measure the on/off ratio

This assumes you have already set up time gating as in “To set up time
gating.”
1. Turn on averaging:
Press [Average], [average on].
2. Turn on and zero the offset marker on the spectrum display:
Press [A], [Shift], [Marker], [Shift], [Marker⇒].
3. Move the gate to the “off” portion of the time display:
Press [B], [Time], [ch1 gate dly].
Rotate the knob until the gate markers encompass the “off” portion of the
signal.
The display should appear as below.

This measurement allows you to determine how much of the carrier leaks
through to the off portion of a burst transmission and therefore establishes
the dynamic range of the transmission system. In this particular example
the dynamic range is low because of the noise inherent in playing a signal
through the arbitrary source.

On the upper trace the offset marker is set at the “on” signal level.
When the gate is moved to the “off” portion of the signal,
the marker reading reflects the difference between the “on” portion
and the “off” portion of the signal.

4-4
5

Making Statistical
Power Measurements
This chapter shows you how to make statistical power measurements, such
as CCDF (Complementary Cumulative Density Function), and peak,
average, and peak-to-average statistical measurements.

5-1
 Making Statistical Power Measurements

To display CCDF

This procedure shows you how to display the Complementary Cumulative

Density Function (CCDF). The procedure uses the analyzer’s source to
generate random noise and to display the CCDF of the random noise.

1. Preset the analyzer.

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Select the Vector instrument mode.
Press [Instrument Mode] [Vector].
3. Connect the analyzer’s source to the channel 1 input.
4. Set the source level to 0 dBm, select random noise, and activate the source.
Press [Source], [level] 0 dBm.
Press [source type], [random noise], [return]
Select [source on].
5. Set the range to 10 dBm.
Press [Range] 10 dBm.
6. Enable time-domain corrections.
Press [System Utility], [time domain cal on].
7. Set the frequency span to 5 MHz to band-limit the random noise signal (this
turns on the analyzer’s LO and zooms the time data).
Press [Frequency], [span], 5 MHz.
8. Display the CCDF trace.
Press [Measurement Data], [more choices], [CCDF ch1].

CCDF (complementary cumulative density function) is a statistical

power-measurement that is the complement of CDF, as follows:

CDF: Probability (Pinst ≤ Paverage)

CCDF: Probability (Pinst ≥ Paverage)

where: Pinst = Instantaneous power

Paverage = Average power

5-2
 Making Statistical Power Measurements

CCDF provides better resolution than CDF for low probability signals,
especially when the y-axis is in log format. Presetting the analyzer
automatically selects log format ([Data Format], [magnitude log(dB)]).

The analyzer plots CCDF using units of % for the y-axis and power (dB)
for the x-axis. Power on the x-axis is relative to the signal average power,
so 0 dB is the average power of the signal. In other words, a marker
reading that shows 12% at 2 dB means there is a 12% probability that the
signal power will be 2 dB or more above the average power.

The analyzer computes CCDF using all samples in the current time record.
Each successive measurement adds additional samples to the CCDF
measurement. Pressing [Measurement Restart] or changing most parameters
under the MEASUREMENT keygroup restarts the CCDF measurement.

Tips CCDF measurements operate on time data. By default, time data is

uncalibrated. Therefore, make sure you enable time-domain corrections as
done in this procedure before making CCDF measurements.
For accurate CCDF measurements on burst signals, use triggering and
time-gating to include only the burst in the CCDF measurement. Including the
signal off-time degrades measurement accuracy. To learn how to use triggering
and time gating, see “To set up time gating” in chapter 4.

The CCDF measurement displays the number of

samples used to compute CCDF (Cnt:) and the
average power of your signal (Avg:).

When the y-axis is log format, the

CCDF display shows two curves: the
CCDF curve of your signal and the
CCDF curve for an ideal band-limited
Gaussian-noise signal.
In log format, the analyzer
automatically plots the ideal curve
(using the same color as the
graticule) so you can compare it with
that of your signal.
For comparison, this procedure set
the span to 5 MHz to band-limit our
random noise signal.
CCDF of Random Noise

5-3
 Making Statistical Power Measurements

To display peak, average, and peak/average statistics

This procedure shows you how to use features under [Marker Function] to
display peak, average, and peak-to-average statistical power measurements.
You can use these features to obtain the same results you get with CCDF
measurements. Unlike CCDF measurements, you can display these
statistical power measurements in any instrument mode as long as the
active trace contains time-domain data. This is useful because these
statistical power measurements give you a way to view power statistics
using the analog, digital, video, and wideband CDMA instrument modes.

Use CCDF measurements for the power distribution of main time results.
Use these statistical power measurements for demodulation results or
results from math functions.
1. Perform the previous task.
The previous task generates the random noise signal used by this task. It
also sets the input range and enables time-domain corrections.
2. Display time-domain data.
Press [Measurement Data], [main time ch1].
3. Set the statistical percentage to 99.8%.
Press [Marker Function], [peak/average statistics], [peak percent], 99.8%.
4. Select a statistical power measurement.
Press [peak power].
5. Turn on the statistical power measurement.
Press [statistics on],.

In many applications, the instantaneous power of a signal can be treated as

a random variable. Depending on the associated statistics, it may or may
not make sense to define power in terms of an absolute value. Instead,
power is defined in probablistic terms.

For example, you may determine that the instantaneous power of a given
signal is less-than-or-equal to 3.5 dBm 99.8% of the time. In this case, you
would say that the peak power is 3.5 dBm at a peak percent of 99.8%, or
the peak power will be below 3.5 dBm 99.8% of the time. Alternatively,
you could say that the instantaneous power will exceed 3.5 dBm 0.2% of
the time (100% – 99.8% = 0.2%). You can represent this probability
mathematically as:

Probability (Pinst ≤ 3.5 dBm) = 99.8% or, more generically:

Probability (Pinst ≤ Ppeak) = Peak percent

where: Pinst = Instantaneous power

Ppeak = Peak power
Peak percent = probability associated with Ppeak

5-4
 Making Statistical Power Measurements

Using the [Marker Function] hardkey, the analyzer lets you set the peak
percent and then display peak, average, or peak-to-average statistical power
for these configurations (otherwise the statistical softkeys are inactive):
l The instrument mode is not Scalar.
l The measurement contains time-domain data (x-axis is time).

The analyzer computes statistical power measurements using all samples in

the current time record. Each successive measurement adds additional
samples to the measurement. Pressing [Measurement Restart] or changing most
parameters under the MEASUREMENT keygroup restarts the measurement.
The [num samples] softkey displays the number of samples used in the
measurement.

Changing [peak percent] or selecting a different statistical power computation

does not restart the measurement. For example, if you measure peak
power and then select average power, the analyzer recomputes average
power using all samples from the peak power measurement. This lets you
view different statistics on the same data.

Important Statistical power measurements operate on time data. By default, time data is
uncalibrated. Therefore, make sure you enable time-domain corrections as
done by step 1 in this procedure before making these measurements.

In this example, the peak

power of this signal will
be below this value
99.8% of the time.

Peak power statistical-power measurement

5-5
6

Creating Arbitrary
Waveforms
This chapter shows you how to generate arbitrary waveforms using the
analyzer’s arbitrary source. You can generate arbitrary waveforms that
contain up to 16,384 samples of real or complex data. Under certain
conditions, you can extend the arbitrary-source length to include up to
32,768 samples of real or complex data.

6-1
 Creating Arbitrary Waveforms

To create a waveform using a single, measured trace

You use trace data to generate arbitrary waveforms. You can generate
short or long waveforms. Short waveforms have up to 4096 samples of
complex data or 8192 samples of real data. Long waveforms have more
than 4096 samples of complex or 8192 samples of real data.

There are two ways of generating trace data. You can use measured data
or you can use a computer program (such as MATLAB1) on your
computer to generate trace data. This and the next task show you how to
generate arbitrary waveforms using measured data. Subsequent tasks show
you how to generate arbitrary waveforms using computer-generated data.

The steps below show you how to use a single, measured trace to create a
short arbitrary waveform. See ‘’To create a waveform using multiple,
measured traces’’ to learn how to create long arbitrary waveforms.

1. Initialize the analyzer and select the Vector instrument mode:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [Vector].

2. Connect your signal to the analyzer (this example uses the analyzer’s source to
provide a 1 MHz fixed-sine signal).
Connect the SOURCE output to the CHANNEL 1 input.
Press [Source], [source on].

3. Select time-domain data.

Press [Measurement Data], [main time].

4. Start the measurement:

Press [Meas Restart].

5. Save the trace data into a data register:

Press [Save/Recall], [save trace], [into D1].

6. Configure the arbitrary source to use your data:

1 MATLAB is a registered trademark of the The MathWorks, Inc.

6-2
 Creating Arbitrary Waveforms

Press [Source], [source type], [arbitrary) , [arb data reg], [D1].

The analyzer’s arbitrary source is now generating the same waveform as

that displayed in the original trace.

The maximum number of samples in the source waveform is dependent on

the sample rate used to create the arbitrary-source data. The maximum
number of samples that the arbitrary source can output varies between
16,384 and 32,768 samples. For details, see ‘’To output the maximum
number of samples’’ later in this chapter.

The analyzer’s arbitrary source requires time-domain data. Because of this,

there are basically three steps to follow when using a single trace to create
an arbitrary waveform:

1 Display the trace using time-domain measurement data.

2 Save the trace into a data register.
3 Turn on the arbitrary source and select the data register.

To create an arbitrary
waveform, display your
signal in the time domain
and save the resulting trace
to a data register. Then
configure the arbitrary
source to use the data
register.

6-3
 Creating Arbitrary Waveforms

To create a waveform using multiple, measured traces

Arbitrary waveforms that contain more than 4096 samples of complex or

8192 samples of real data are considered long waveforms. To create a long
waveform using measured data, you must use multiple traces (a waterfall
or spectrogram display).

1. Initialize the analyzer and select the Vector instrument mode:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [Vector].
2. Connect your signal to the analyzer (this example uses the analyzer’s source to
provide a 1 MHz fixed-sine signal).
Connect the SOURCE output to the CHANNEL 1 input.
Press [Source], [source on].
3. Set the analyzer’s frequency span to include all components of your signal. If
possible, use a cardinal span to ensure the arbitrary source can output the
maximum number of samples (for details, see ‘’To output the maximum
number of samples’’ later in this chapter).
Press [Frequency], [center], 1 MHz.
Press [span], 39.0625 kHz.
4. Create a contiguous waterfall or spectrogram display that contains
time-domain data.
See ‘’To create a contiguous waterfall or spectrogram display.’’
5. Save the contiguous waterfall or spectrogram data into a data register:
Press [Save/Recall], [save more], [save trace buffer], [into D1].
6. View the data register contents on trace B (this step is optional):
Press [B], [Measurement Data], [data reg], [D1] .
Press [Display], [waterfall setup], [waterfall on].
7. Configure the analyzer to measure from the input channel instead of the
time-capture buffer:
Press [Instrument Mode], [measure from input], [remove capture].
8. Turn on the arbitrary source:
Press [Source], [source type], [arb data reg], [D1], [Return].
Press [arbitrary], [Return].
Press [source on].

The analyzer displays ‘’Loading arb source from register D1’’ and
‘’Arbritary Source length: XXX samples.’’

6-4
 Creating Arbitrary Waveforms

9. Start the measurement and view the results:

Press [A], [Meas Restart].
The arbitrary source is now generating your signal.

Waterfall and spectrogram displays store trace data in the trace buffer.
Both displays use the same trace buffer, therefore it doesn’t matter which
display you use when you save the trace buffer. The [buffer depth] softkey
determines the size of the trace buffer. For example, a buffer depth of 20
means the trace buffer can contain up to twenty traces, regardless of how
many traces are displayed.

If the analyzer displays OUT OF MEMORY when you try to save data into
a data register, you need to reconfigure the analyzer’s memory. You may
want to press [System Utility], [options setup] to see if your analyzer has option
UFG. Option UFG adds an additional 4 MB of memory (and LAN
capability) to your analyzer.

HINT A good way to increase the amount of memory available for data registers is to
reduce [max freq points]. The value of this softkey determines the maximum
number of points in a trace and also reserves memory for other internal
operations. Press [System Utility], [memory usage], [configure meas memory], [max freq points] to
change this parameter.

The arbitrary source may not be able to use all data in the data register.
The arbitrary source can use up to 16,384 samples of real or complex data.
Under certain conditions, the arbitrary source can use up to 32,768 samples
of real or complex data (see ‘’To output the maximum number of samples’’
later in this chapter).

In this example, the data

register contains 20 traces.
The value of [buffer depth]
determines the number of
traces saved to the data
register.

6-5
 Creating Arbitrary Waveforms

To create a short waveform using ASCII data

There are several computer programs that let you create arbitrary
waveforms (such as MATLAB or MATRIXx2). This procedure shows you
how to load a short, computer-generated waveform into the analyzer’s
arbitrary source. Short waveforms contain 4096 complex or 8192 real
points, or less.

1. Use your program to create a waveform that contains no more than 4096
complex or 8192 real points (for larger waveforms, see ‘’To create a long
waveform using ASCII data’’).
2. Save your waveform as an ASCII file.
3. Convert your file from ASCII to SDF.
With the SDF utilities installed on your computer, type the following from
a DOS prompt:

ASCTOSDF [/z:cf] /x:0, ∆t source_file destination_file

where: /z:cf specifies the center frequency (use with complex data).
/x:0 specifies the start time as 0 seconds.
∆t is the interval between samples.
source_file is the name of your ASCII file
destination_file is the name of the SDF file.

4. Copy the SDF file to a 3.5" floppy disk.

5. Load the floppy disk into the analyzer and recall the SDF file into a data
register.
Press [Save/Recall], [default disk], [internal disk], [Return].
Press [catalog on] and select your file.
Press [recall trace], [from file into D1].

6. Configure the arbitrary source to use the data register.

Press [Source], [source type], [arb data reg], [D1], [Return].

7. Turn on the arbitrary source.

Press [arbitrary], [Return], [source on].

The arbitrary source is now generating your waveform.

2 MATRIXx is a product of Integrated Systems, Inc.

6-6
 Creating Arbitrary Waveforms

The analyzer stores trace data in Standard Data Format (SDF). Therefore,
you must use the Standard Data Format utilities to convert your data to
the SDF format recognized by the analyzer. For details about the SDF
utilities, see the Standard Data Format Utilities: User’s Guide shipped
with your analyzer.

The following paragraphs show you how the AMSIG.DAT file on the
Signals Disk was created. This signal was created using MATLAB. It is an
amplitude-modulated signal that uses a 25 kHz sinewave to modulate a
5 MHz carrier. The sample frequency is 25.6 MHz. Here are the equations
and commands used to create this signal (> is the MATLAB prompt):

>t=0:1023;
>x=t/1024*2*pi;
>y=(sin(200*x)).*(.7+.2*sin(x ));R >save amsig.asc y /ascii
>quit

where: t is a 1x1024 array of numbers from 0 to 1023.

x is a 1x1024 array of numbers from 0 to (almost) 2 pi.
y is a 1x1024 array of numbers composing 200 cycles of a sinusoid
signal that is amplitude modulated by one cycle of a sinusoid
signal
that has an index of modulation of 0.7.

The MATLAB file was converted to SDF format using the ASCTOSDF
utility, as follows:

ASCTOSDF /x:0,3.90625e-8 amsig.asc amsig.dat

where: /b: specifies the block size

/x: specifies the start time (or trigger delay) and ∆t.
1 1 −8
∆t = = = 3.90625 e
sample frequency 25.6 MHz

You can derive the frequency span using the following formula (use n =
2.56 since the ASCTOSDF command did not include the /z:cf argument):
1 1.28 for complex (zoom) data
span = where n =
∆t × n 2.56 for real (baseband) data

6-7
 Creating Arbitrary Waveforms

To create a long waveform using ASCII data

There are several computer programs that let you create arbitrary
waveforms (such as MATLAB or MATRIXx). This procedure shows you
how to use a long, computer-generated waveform with the analyzer’s
arbitrary source. Long waveforms have more than 4096 complex or 8192
real points.

1. Using your computer program, create your waveform and save it to an ASCII
file. Note the number of samples and the ∆t of your waveform.

2. Create a waterfall or spectrogram display that has the same number of samples
and ∆t as your waveform.

See “To create a fixed-length waterfall display.”

3. Using the results from the previous step, create a contiguous waterfall or
spectrogram display.

See “To create a contiguous waterfall or spectrogram display.”

4. Save the contiguous waterfall or spectrogram display (the trace buffer) to disk.
Press [Save/Recall], [default disk], [internal disk], [Return].
Press [save more], [save trace buffer], [into file].

5. Copy the trace-buffer file from disk onto your computer and put it in the same
directory as the ASCII file you created in step 1.

6. Use the sdfydata utility to replace the data in the trace_buffer file with data
from your ASCII file.

SDFYDATA sdf_file ASCII_file

where: sdf_file is the trace-buffer file.

ASCII_file is the ASCII file that you created in step 1.

6-8
 Creating Arbitrary Waveforms

7. Load the modified trace-buffer file into one of the analyzer’s data registers.
On your computer, copy the modified trace-buffer file to floppy disk.
Insert the disk in the analyzer’s disk drive.
Press [Save/Recall], [recall more], [catalog on] and select your file.
Press [recall trace buffer], [from file into D1].

8. Configure the arbitrary source to use the data register.

Press [Source], [source type], [arb data reg], [D1], [Return].

9. Turn on the arbitrary source.

Press [arbitrary], [Return], [source on].

The analyzer stores trace data in Standard Data Format (SDF). You
substitute trace data from a waterfall or spectrogram display using the
sdfydata Standard Data Format (SDF) utility. This utility is one of several
utilities included in the Standard Data Format Utilities: User’s Guide
shipped with your analyzer. The sdfydata utility automatically converts the
ASCII data to SDF format as it copies the ASCII data into the SDF file.

For additional details about installing and using the SDF utilities, see the
Standard Data Format Utilities: User’s Guide shipped with your analyzer.

6-9
 Creating Arbitrary Waveforms

To create a contiguous waterfall or spectrogram display

Contiguous traces are needed when you use a waterfall or spectrogram

display to generate an arbitrary-source waveform. You use waterfall or
spectrogram displays to generate arbitrary waveforms that contain more
than 4096 samples of complex data or 8192 samples of real data.

1. Prepare time-capture RAM for your signal:

Press [Instrument Mode], [capture setup].
Check [buffer length] to verify that your signal will fit in time-capture RAM.
Increase [buffer length] if necessary.

2. Fill time-capture RAM with your signal:

Press [fill buffer].

3. Set the overlap to 0%.

Press [Time], [ovlp: avg off 0 %].

4. Select time-domain data.

Press [Measurement Data], [main time].

5. Select a waterfall or spectrogram display and set the buffer depth to 20.
Press [Display], [waterfall setup], [waterfall on].
Press [buffer depth 20].

6. Start the measurement:

Press [Meas Restart].

Storing your signal in time-capture RAM and playing it back with 0%

overlap ensures that the traces in a waterfall or spectrogram display are
contiguous. To use waterfall or spectrogram data to drive the arbitrary
source, you need contiguous traces to eliminate phase discontinuities in the
arbitrary-source waveform. Make sure to display your data in the time
domain since the arbitrary source requires time-domain data.

This procedure uses a buffer depth of 20 to accommodate all procedures in

this chapter. Be sure to set the buffer depth according to your
measurement needs.

6-10
 Creating Arbitrary Waveforms

To learn about waterfall and spectrogram displays, see ‘’Using Waterfall

And Spectrogram Displays (Opt. AYB)’’ in the Operator’s Guide and see
online help for the [waterfall setup] and [spectrogram setup] softkeys.

This waterfall display was

created with an elevation of 25
pixels and trace height of
30 pixels. Notice that the traces
are contiguous (there is no phase
discontinuity between traces).
When looking at a waterfall
display, remember that the top
trace is the most recent trace,
the trace before it is the previous
trace, and so forth. Therefore, to
see if a waterfall display is
contiguous, you must piece the
waveforms together from the
bottom up.

Contiguous Waterfall Display With 20 Traces

6-11
 Creating Arbitrary Waveforms

To create a fixed-length waterfall display

There are several computer programs that let you create arbitrary
waveforms (such as MATLAB or MATRIXx). If the waveform is a long
waveform (it contains more than 4096 complex points or 8192 real points),
you must create a waterfall or spectrogram display on the analyzer, copy
the waterfall or spectrogram data to your computer, and use the Standard
Data Format (SDF) utilities to replace the waterfall or spectrogram data
with the data from your computer program. The waterfall or spectrogram
data must have the same sample frequency, length, and ∆t (time-interval
between points) as your computer-generated waveform.

This procedure shows you how to create a waterfall or spectrogram display

that contains the number of samples and ∆t that you need. See ‘’To
create a long waveform using ASCII data’’ to learn how to perform the
remaining steps.

This procedure configures the analyzer to create a waterfall display for a

complex signal that has a sample frequency of 3.84 MHz and a total source
length of 5 ms. Substitute your values where necessary.
1. Initialize the analyzer and select the Vector instrument mode:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [Vector].
2. Set the rbw mode to arbitrary and rbw coupling to auto:
Press [ResBW/Window], [rbw mode arb], [rbw coupling auto], [Return].
3. Compute the total number of samples in the waveform:
Msamples
Total Samples = 3.84 × 5 ms = 19,200 samples
sec
4. Divide the total number of samples into N equal-length segments. Note that:
l Each segment must contain an integer number of samples:
l For real signals, each segment must have between 128 and 8192 samples.
l For complex (zoom) signals, each segment must have between 64 and 4096
samples.
For this example, there are several possible solutions, such as:

10 segments of 1920 samples.

12 segments of 1600 samples.
20 segments of 960 samples.

6-12
 Creating Arbitrary Waveforms

5. Set the number of frequency points for your signal. The number of frequency
points must be greater than the number of samples in a segment divided by
2.56 (real) or 1.28 (complex):

1920 samples
For 10 segments of 1920 samples: = 1500 −−> Use 1601 freq. pts.
1.28
1600 samples
For 12 segments of 1600 samples: = 1250 −−> Use 1601 freq. pts.
1.28
960 samples
For 20 segments of 960 samples: = 750 −−> Use 801 freq. pts.
1.28

For this example, use 801 frequency points:

Press [ResBW/Window], [num freq points], 801.

6. Set the frequency span to obtain the desired sample frequency. For real
signals, divide the sample frequency by 2.56; for complex signals, divide by 1.28:

3.84 MHz
Frequency span (complex) = = 3 MHz
1.28

Press [Frequency], [span], 3, [MHz].

7. Set the time-record length equal to the length of one segment:
960 samples
Length of one segment = = 250 µ sec
3.84 MHz

Press [Time], [main length], 250 [µ s].

8. You are now ready to create a contiguous waterfall or spectrogram display for
your computer-generated data. Take note of the number of segments (time
records) needed for the waterfall or spectrogram display (from step 5) and see
‘’To create a long waveform using ASCII data’’ for further instructions.

Here is a contiguous waterfall display

created using the parameters from
this procedure. The overall length is
250 µs and the marker shows ∆t at
260.42 ns. You add ∆t to the stop time
to determine the length of each trace.
You then multiply the length of each trace
by the total number of traces to compute
the length of the waterfall display.

You may notice that the stop time is not

250 µs. The stop time does not include
∆t for the last point.

6-13
Creating Arbitrary Waveforms

To determine number of samples and ∆t

If you are using a waterfall or spectrogram display to import data from

your computer-generated waveform, the waterfall or spectrogram must have
the same number of samples and ∆t as your computer-generated waveform,
as explained in “To create a fixed-length waterfall display.”

The time-interval between points (∆t) is calculated according to the

following
1 1.28 for zoom measurements
formula:∆ t = where n =
( frequency span ) × n 2.56 for baseband measurements

The number of points (samples) in a time record is easy to determine with

the right configuration. With [rbw coupling auto] and [rbw mode arb], the [num freq pts]
softkey determines the number of points, as shown in the following table
(all softkeys are located under the [ResBW/Window] hardkey).

Time Record Length Using [rbw coupling auto] And [rbw mode arb]

Number of Frequency Points Time-record Length (in points)

(value of [num freq pts] softkey) Zoom Baseband
51 64 128
101 128 256
201 256 512
401 512 1024
801 1024 2048
1601 2048 4096
3201 4096 8192

The numbers in the above table are valid only when [rbw coupling auto] and
[rbw mode arb] are selected. For any other combination of rbw coupling
or rbw mode, you must measure the time-record length and use the
following formula to determine the number of samples in a time
record:

Time record length

Number of points =
∆t

6-14
 Creating Arbitrary Waveforms

To output the maximum number of samples

You use time-domain data (samples) in a data register to drive the

analyzer’s arbitrary source. The arbitrary source can use up to 16,384
samples. If the samples were created using a cardinal frequency span, the
arbitrary source can use 32,768 samples.

The number of samples you can save to a data register may be limited by
memory configuration. The analyzer displays OUT OF MEMORY if there is
insufficient memory when you try to save data into a data register.

HINT A good way to increase the amount of memory available for data registers is to
reduce [max freq points]. The value of this softkey determines the maximum
number of points in a trace and also reserves memory for other internal
operations. Press [System Utility], [memory usage], [configure meas memory], [max freq points] to
change this parameter.

It’s possible for a data register to contain more data (more samples) than
the arbitrary source can use. The maximum number of samples that the
arbitrary source can use is 16,384 samples. If you used a cardinal
frequency span to create the arbitrary waveform (to create the data in the
data register), the arbitrary source can use 32,768 samples. If you did not
use a cardinal frequency span, the maximum number of samples that the
arbitrary source can use will be between 16,384 and 32,768 samples.

If possible, use a cardinal frequency span to create your arbitrary

waveform. Spans slightly larger than a cardinal span slightly reduce the
number of samples that the arbitrary source can use. Increasing the
frequency span further reduces the number of samples that the arbitrary
source can use, the worst case being when the frequency span is just
below a cardinal span. Cardinal frequency spans are spans that fit the
following formula:

10 MHz
Cardinal Spans = , where n is a whole number.
2n

Based on this formula, cardinal spans are 10 MHz, 5 MHz, 2.5 MHz, and
1.25 MHz, and so on. For some instrument configurations, the maximum
frequency span is less than 10 MHz. For these configurations, use the
same formula to determine the cardinal spans and discard values that
exceed the maximum frequency span. For example, the maximum
frequency span of the external receiver is 7 MHz, in which case the
cardinal spans are 5 MHz, 2.5 MHz, 1.25 MHz, and so on.

6-15
7

Using Waterfall and

Spectrogram Displays
(Opt. AYB)
This chapter shows you how to view signals in an almost three
dimensional way by displaying multiple traces as a function of time.

7-1
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To create a test signal

This procedure creates the test signal used throughout this chapter to
demonstrate waterfall and spectrogram features. You use the analyzer’s
source to generate a sine wave, connect the sine wave to the analyzer’s
channel 1 input, and then overrange the analyzer to simulate a spectral
display with multiple tones.

1. Preset the analyzer.

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Select the Vector instrument mode.
Press [Instrument Mode] [Vector].

3. Connect the analyzer’s source to the channel 1 input.

4. Activate the source and set the source level to 0 dBm.

Press [Source] [level] 0 dBm.
Select [source on].

5. Set the range to -30 dBm to overload the analyzer’s ADC.

Press [Range] – 30 dBm.

After completing this step, the analyzer displays OV1 to inform you that
the signal is overloading the analyzer’s ADC (in this case, the range is too
low). We do this intentionally to simulate a signal with multiple tones.

6. Turn averaging on to reduce the noise floor.

Press [Average] [average on].

7. Perform an auto scale to properly position the trace.

Press [Auto Scale].

7-2
 Using Waterfall and Spectrogram Displays (Opt. AYB)

Hint You may want to save the current measurement state to non-volatile RAM
(NVRAM). That way, if you don’t have time to finish the procedures in this
chapter, you can quickly reproduce the test signal. To do this, press
[Save/Recall] [default disk] [non-volatile RAM disk] [return] [save state]. Then enter the name
of the file you want to create and press [enter]. To recall the state, press
[Save/Recall] [catalog on] [recall state], select the file that you created, and press [enter].

Test Signal

7-3
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To set up and scale a waterfall display

This procedure uses the signal created at the beginning of this chapter to
show you how to set up and scale waterfall displays.

1. Perform the procedure at the beginning of this chapter to create a test signal.

2. Turn on the waterfall display for trace A.

Press [A] to activate trace A.
Press [Display] [waterfall setup] [waterfall on].

3. Press [Meas Restart] to start a new waterfall measurement.

4. Set the trace height.

Press [trace height] 100 [pixels].

5. Set the elevation.

Press [elevation] 20 [pixels].

6. Set the desired skew.

Press [azimuth] 10 [pixels].

7. Set the size of the trace buffer.

Press [buffer depth] 10 [enter].

In a waterfall display, new traces are added to the top of the display as
older traces flow to the bottom. The analyzer displays marker information
for the most recent (top) trace. Later in this chapter you will learn how to
select and display marker information for other traces in the waterfall
display.

Trace height determines the vertical space (in pixels) allotted to each trace.
The height of a trace within that vertical spacing depends on the y-axis
scaling, which is set with [Ref Lvl/Scale] [Y per div]. It may be easier to think
of trace height as defining the height of a box that each trace must fit in.

Whereas trace height sets the height of a box that each trace must fit in,
elevation determines the vertical space (in pixels) between those boxes. If
the elevation is less than the trace height, the “boxes” overlap—which
means the traces overlap. If the elevation is larger than the trace height,
the “boxes” don’t overlap.

7-4
 Using Waterfall and Spectrogram Displays (Opt. AYB)

Azimuth determines the shift, or skew, of the waterfall display. Aximuth

tells the analyzer how far, in pixels, to shift a trace from the previous
trace. Negative numbers shift the trace left; positive numbers shift the
trace right.

Buffer depth determines the number of traces stored in the waterfall

buffer. Larger numbers require more memory.

Hint Additional features such as threshold, baseline, and hidden line are also
available. For details about these features, press [Display] [waterfall setup]. Then
see online help for the keys that enable these features.

Hint You may find it easier to change parameters for softkeys which require a
numeric entry by using the knob. To activate the knob for entries, press
[Marker|Entry] and illuminate the Entry LED.

Z-axis X-axis Y-axis

value of value of value of
marker. marker. marker.

Y-axis

Z-axis

X-axis

Waterfall Display

7-5
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To select a trace in a waterfall display

This procedure shows you how to select traces in a waterfall display.

You can select any trace in the waterfall buffer. You can select a trace by
number, or by its z-axis value.

1. Follow the instructions in “To setup and scale a waterfall display” to create a
waterfall display.

2. Display two grids.

Press [Display] [2 grids].

3. Select the same measurement data for both grids.

Press [Shift] [B] to activate traces A and B.
Press [Measurement Data] [spectrum].

4. Turn on and couple markers for both traces A and B.

Press [Marker] [marker on].
Press [couple mkrs on].

5. Press [Pause|Single] to pause the measurement.

6. Turn on the trace selection feature.

Press [Marker Function] [trace select on].

The most recent trace (top) trace is selected if trace select is off. To
select any other trace, you must turn trace select on.

7. Select the desired trace.

Press [trace] and enter the number of the trace you want to select.

OR

Press [Marker|Entry] to highlight the Entry LED and rotate the knob to
select the desired trace.

7-6
 Using Waterfall and Spectrogram Displays (Opt. AYB)

You can select a trace by number or by its z-axis value (in seconds).
Trace number 1 is the first, or oldest trace in the waterfall buffer. To
select a trace by its z-axis value, press [trace] followed by the z-axis value.

Hint To display the current z-axis value in the [trace] softkey, press [trace] [s] . To
display the current trace number, press [trace] [enter].

If the marker is on, the analyzer displays marker information for the
selected trace. The z-axis marker value (see below) is the elapsed time
from when you pressed [Meas Restart] to when the trace was created.
Pausing a measurement does not reset the z-axis clock. If you pause and
resume a measurement, the z-axis value of the next trace is still referenced
to the last time you pressed [Meas Restart].

Marker values for the selected trace:

Z-axis value, X-axis value, and Y-axis value.

When you select a trace in a

waterfall display, other grids
that have identical
measurement data, such as
this one, show the selected
trace. If the measurement
data were different, this
grid would always show the
most recent trace acquired.
Using Trace Select In a Waterfall Display

7-7
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To use markers with waterfall displays

This procedure uses the results of the previous procedure to show you
how to use markers and offset markers with waterfall displays.

1. Perform the previous procedure to create a waterfall display and enable trace
selection.

2. Select the oldest trace in the waterfall display.

Press [Marker Function] [trace] 1 [enter].

3. Move the marker to the highest peak.

Press [Shift] [Marker].

[Shift] [Marker] is a shifted function (as indicated by the blue labeling

above the [Marker] hardkey) that moves the marker to the highest peak on
the selected trace. Later you will learn how to move the marker to the
highest peak in the waterfall buffer.

4. Turn on and zero the offset marker.

Press [Shift] [Marker →] to turn on and zero the offset marker.

Another way to set the position of the offset marker is with the
[offset posn setup] softkey (located under the [Marker] hardkey). This softkey
displays another softkey menu that lets you manually specify the x-, y-,
and z-axis location of the offset marker.

5. Move the marker to the next highest peak on the selected trace.
Press [Marker Search] [next peak].

With the offset marker on, marker values are relative to the offset marker.
Since the marker is on the same trace as the offset marker, the z-axis
marker value is still zero (0).

7-8
 Using Waterfall and Spectrogram Displays (Opt. AYB)

6. Move the marker to the next trace in the waterfall display.

Press [Marker Function] followed by the up-arrow key.

In this example, notice that the z-axis marker value is approximately 41.2
ms, which is the elapsed time between the two traces. In other words,
41.2 ms elapsed from when the analyzer acquired trace 2 to when it
acquired trace 1. The amount of time needed to acquire each trace is
dependent on several factors and may be different than that shown in this
example.

When the offset marker is on, marker

values are relative to the offset
marker. In this example, the selected
trace was acquired approximately
41.2 ms after the previous trace.

A diamond shows the

location of the marker.
A square shows the
location of the offset
marker.

Using Markers with a Waterfall Display

7-9
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To use buffer search in waterfall displays

This procedure shows you how to use buffer search to do marker-search

operations over all traces in the waterfall buffer. Start with step 2 if you
just finished the previous procedure.

1. If you just finished the previous procedure, skip this step.

Otherwise, configure the analyzer for this procedure by following the
instructions in “To select a trace in a waterfall display.”

2. Turn buffer search on.

Press [Marker Search] [buffer search on].

With buffer search on, marker-search operations include all traces in the
waterfall buffer. With buffer search off, marker-search operations are
confined to the selected trace.

Some marker-search operations are unavailable when buffer search is on.

The softkeys for these operations are ghosted to inform you that the
operations are unavailable.

3. Move the marker to the highest peak in the waterfall buffer.

Press [marker to peak]
OR
Press [Shift] [Marker].

[Shift] [Marker] is a shifted function, as indicated by the blue label above

the [Marker] hardkey, that duplicates the [marker to peak] operation.

4. Move the marker to the next highest peak in the waterfall buffer.
Press [Marker Search] [next peak].

Notice that the marker moved to the next highest peak in the waterfall
buffer which, in this example, is on a different trace.

5. Move the marker to the lowest point in the waterfall buffer.

Press [Marker Search] [marker to minimum].

7 - 10
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To set up a spectrogram display

This procedure shows you how to set up and view a spectrogram display.

1. Perform the procedure at the beginning of this chapter to create a test signal.

2. Turn on the spectrogram display for trace A.

Press [A] to activate trace A.
Press [Display] [spectrogram setup] [spectrogram on].

3. Set the size of the trace buffer.

Press [buffer depth] 50 [enter].

4. Press [Meas Restart] to start a new spectrogram measurement.

Notice that you set up a spectrogram display in much the same way as a
waterfall display. Both displays share the same trace buffer. Therefore,
you may switch between waterfall and spectrogram displays without losing
data.

A spectrogram display is simply another method of looking at trace data.

In a spectrogram, the y-axis is reduced to a single line, where color
represents different y-axis values. Therefore, each trace occupies a single,
horizontal line on the display. Because of this, a single-grid spectrogram
display requires about 300 traces to fill the entire display.

A spectrogram uses a colorbar to represent the y-axis. The colorbar shows

how colors are distributed along the y-axis and the y-axis value of each
color.

The flow of trace data in a spectrogram display is opposite a waterfall

display. In a waterfall display, traces flow from the top to the bottom of
the display, with the most recent trace at the top. In a spectrogram
display, traces flow from the bottom to the top of the display, with the
most recent trace at the bottom.

For additional information, see online help for the [spectrogram setup] softkey.

7 - 11
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To enhance spectrogram displays

This procedure shows you how to use advanced features to enhance the
spectrogram display.

1. Perform the previous procedure, “To set up a spectrogram display.”

2. Pause the measurement after the spectrogram fills the entire display.
Press [Pause|Single].

3. Adjust the mapping of amplitude to color to obtain the best display.

Press [Marker|Entry] to illuminate the Entry LED.
Press [enhance] and rotate the knob to obtain the best spectrogram display.

The [enhance] softkey lets you redistribute colors in the colorbar. A value
of 50% evenly distributes the colors. A value of 0% compresses the colors
into the top of the colorbar, whereas a value of 100% compresses them
into the bottom of the colorbar. Online help provides additional
information on this feature.

4. Select different color maps to see the effects on the spectrogram display.
Rotate the knob to until [enhance] is back to its default value of 50%.
Press [map color] [color reverse]
Press [grey normal]
Press [grey reverse]
Press [color normal] to return to the default color map.

Color maps change the colors used in the color bar. Different color maps
offer different perspectives. Essential information may be buried, or
obscure in one color map, but prominent in another.

7 - 12
 Using Waterfall and Spectrogram Displays (Opt. AYB)

5. Change the number of colors used in the color map.

Press [return].
Press [number colors] 2 [enter].
Press 5 [enter].
Press 10 [enter].
Press 64 [enter] to return to the default number of colors.

In this example, changing the number of colors erases the upper portion
of the spectrogram display. This happens because the traces in this area
are not in the spectrogram buffer. In other words, the spectrogram buffer
does not contain enough traces to fill the entire display. Changing the
number of colors forces the analyzer to recompute scaling factors for the
y-axis. The analyzer can do this only for traces in the spectrogram buffer.

6. Adjust the threshold to remove noise from the spectrogram display.

Press [threshold].
Rotate the knob to obtain the desired display.

As demonstrated, you can use the threshold feature to hide unwanted

information. When you specify a threshold, the analyzer displays colors at
or above the threshold; below the threshold, the analyzer uses the bottom
color in the colorbar.

You specify threshold as a percentage of the colorbar. A threshold of 0%

displays the entire signal. A threshold of 50% only displays colors in the
upper half of the colorbar—colors below this point (or threshold) are
displayed in the same color as the bottom color in the colorbar.

Hint You may overlay another type of trace on a spectrogram and observe, for
example, the spectrum or PSD of the input signal. To do this, press [Display]
[view/overlay traces] and turn on the trace that you want to use as an overlay.

Color map, enhance, and threshold affect all traces that contain
spectrogram displays. For example, if traces A and B contain spectrogram
displays, changing the color map for trace A also affects trace B.

7 - 13
Using Waterfall and Spectrogram Displays (Opt. AYB)

To use markers with spectrogram displays

You use markers in a spectrogram display the same way you use markers
in a waterfall display. To learn how to use markers (and select traces) in
a spectrogram display, display a spectrogram instead of a waterfall and
perform the following procedures:

l To select a trace in a waterfall display.

l To use markers with waterfall displays.
l To use buffer search in waterfall displays.

Note the following differences when performing the above procedures with
a spectrogram dispay.
l Spectrograms use a vertical line instead of a diamond to show the location of
the marker.

l Spectrograms use a vertical line and a horizontal line with a square to show the
location of the offset marker.

l Spectrograms use a horizontal line to show the selected trace when trace select
is on.

7 - 14
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To save waterfall and spectrogram displays

You save a waterfall or spectrogram display by saving the trace buffer.

The trace buffer contains the traces that make up the waterfall or
spectrogram display. Both waterfall and spectrogram displays share the
same trace buffer (which is why you can switch between waterfall or
spectrogram displays without losing data).

1. Select the desired depth for the trace buffer:

Press [Display] [waterfall setup] [buffer depth].
OR
Pr ess [Display] [spectrogram setup] [buffer depth].

2. Activate the waterfall or spectrogram display that you want to save. For
example, if the display is in trace A, press [A].

3. Save the trace buffer in a file or data register:

Press [Save/Recall], [save more], [save trace buffer].

Waterfall and spectrogram displays share the same trace buffer. Therefore,
it doesn’t matter which is displayed when you save the trace buffer. When
you recall the trace buffer, you can examine trace-buffer data in either
waterfall or spectrogram format.

The size of the trace buffer determines the number of traces that are
saved. For example, a buffer depth (trace buffer size) of 10 means ten
traces will be saved, regardless of how many traces are displayed. Save
the trace buffer to a file if you want to recall the trace buffer after a
power down.

Memory constraints may affect trace-buffer size. For details, see online
help for the [remove trace buffers] softkey (under [System Utilities] [memory usage]).

7 - 15
 Using Waterfall and Spectrogram Displays (Opt. AYB)

To recall waterfall and spectrogram displays

This procedure shows you how to recall waterfall and spectrogram

displays. See the previous procedure for instructions on saving waterfall
and spectrogram displays.

1. Activate the trace where you want the waterfall or spectrogram display. For
example, if you want to display a waterfall display in trace A, press [A].

2. Turn on the waterfall or spectrogram display.

Press [Display] [waterfall setup] [waterfall on].
OR
Press [Display] [spectrogram setup] [spectrogram on].

3. If the trace buffer is in a file, recall the trace buffer from the file into a data
register (if the trace buffer is in a data register, skip this step):
Press [Save/Recall], [recall more], [recall trace buffer].
Cho ose a file and data register.

4. Display the data register:

Press [Measurement Data], [data reg]
Choose the desired data register.

Any trace buffer may be recalled as either a spectrogram or a waterfall,

regardless of which display was selected at the time the buffer was stored.
If the data register contains only one trace, then the waterfall or
spectrogram only displays the single scan.

7 - 16
8

Using Digital
Demodulation
(Opt. AYA)
This chapter shows you how to use digital demodulation to demodulate
and view digitally modulated signals. You may perform the tasks in this
chapter using signals from the Signals Disk, or you may use these tasks as
a model for demodulating your own signals.

8-1
 Using Digital Demodulation (Opt. AYA)

To prepare a digital demodulation measurement

This task shows you one way to set up a digital demodulation

measurement. The task uses a NADC signal from the signals disk. Several
other tasks in this chapter use this setup to teach you how to use digital
demodulation.

1. Initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [vector].
2. Supply a NADC signal to the channel 1 INPUT or perform the following steps to
load a NADC signal from the Signals Disk into a data register and play it
through the analyzer’s aribtrary source:
Insert the Signals Disk into the internal disk drive.
Connect the SOURCE to the channel 1 INPUT.
Press [Save/Recall], [default disk], [internal disk].
Press [Return], [catalog on].
Rot ate the knob to highlight PI4DQPSK.DAT
Press [recall trace], [from file into D1], [enter].
Press [Source], [source on], [source type], [arbitrary].
3. Select the optimum range:
Press [Range].
Press the down-arrow key until the Channel-1 Over and Half LEDs are on.
Press the up arrow key one press at a time until the Over LED turns off.

For additional details about selecting the optimum range, see online help
for the [Range], [ch1 range] softkey.
4. Select a center frequency and span:
Press [Frequency], [center], 5, [MHz]
Press [span], 100, [kHz].

The center frequency tunes the analyzer to the carrier frequency. To

obtain reliable carrier lock, the center frequency must be close to the
carrier frequency. For details, see ‘’Carrier Locking’’ in the Digital
Demodulation Concepts (Opt. AYA) chapter.

Selecting the correct frequency span is also important when using digital
demodulation. The span must be wide enough to include all signal
components, and yet not too wide, or the measurement may be affected
by excessive noise and slower speed. For details, see ‘’Span
considerations for digitally demodulationed measurements’’ in the
Digital Demodulation Concepts (Opt. AYA) chapter.

8-2
 Using Digital Demodulation (Opt. AYA)

5. Set up the trigger:

Press [Trigger], [trigger type], [internal source], [return], [ch1 delay], 1, [ms].

This example uses a signal which has been supplied from the Signals Disk.
When you supply another signal to the channel 1 input you need to select
appropriate center frequency, span, range, and triggering parameters prior
to demodulating the signal.

The spectrum of a digitally modulated carrier before demodulation.

8-3
 Using Digital Demodulation (Opt. AYA)

To demodulate a standard-format signal

This task shows you how to demodulate the NADC signal on the Signals
Disk.
1. Configure the analyzer for a digital demodulation measurement.
If you haven’t already done so, perform the steps in the previous
task, ‘’To prepare a digital demodulation measurement.’’
2. Digitally demodulate the signal:
Press [Instrument Mode], [Digital Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Digital Demodulation], [Return].)
3. Choose standard demodulation setup parameters:
Press [demodulation setup], [demod format], [standard setups], [NADC].
4. Modify the standard parameters for this specific signal:
Press [Time], [result length], 100, [sym]
Press [pulse search off]
Press [Auto Scale]

If you are demodulating a signal which matches a standard signal type, you
can automatically configure the analyzer for that standard by pressing
[standard setups] and then choosing the appropriate type. The parameters set
when you choose a standard are: demod format, span, symbol rate, meas
filter, ref filter, alpha/BT, result length, pulse search, and points per
symbol. If your signal is not of a standard type you may select individual
parameters in the [demodulation setup] menu. To learn how to do this, see ‘’To
demodulate a non-standard-format signal’’ in the Analyzing Digitally
Demodulated Signals chapter.

A time display of a demodulated signal

8-4
 Using Digital Demodulation (Opt. AYA)

To select measurement and display features

The analyzer provides many different ways of viewing demodulated data.

This task shows you how to display demodulated data two different views
in two grids.

1. Demodulate your signal as shown in the previous task.

2. Select multiple display grids:

Press [Display], [4 grids quad ].

3. Change the measurement data for trace B:

Press [B], [Measurement Data], [error vector spectrum]

4. Change the data format for trace A:

Press [A], [Data Format], [polar IQ constell ation].

The [Measurement Data] menu allows you to select the type of data you want
to see, while the [Data Format] menu selects how you want to display that
data. You may select different measurement data and data formats for up
to four traces by activating each trace individually.

Each grid shows a different measurement type with an appropriate data format.

8-5
 Using Digital Demodulation (Opt. AYA)

To set up pulse search

In this example you learn how to perform pulse search on a burst signal.
This example uses a signal, provided on the Signals Disk, which is a record
of the output of a keyed NADC radio transmission.

1. Load “MEAS_PI4.DAT” from the Signals Disk into the arbitrary source:
Perform steps 1, 2, and 3 in ‘’To prepare a digital demodulation
measurement.’’ For step 2, load the “MEAS_PI4.DAT” signal from the
Signals Disk instead of ‘’PI4DQPSK.DAT.’’

2. Select appropriate setup parameters:

Press [Range] and select the optimum range.
Press [Frequency], [center], 5, [MHz]
Press [span], 100, [kHz].
Press [Trigger], [trigger type], [internal source], [return], [ch1 delay], 1, [ms].

If you don’t know how to select the optimum range, see ‘’To prepare a
digital demodulation measurement.’’

3. Select appropriate demodulation parameters:

Press [Instrument Mode], [Digital Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Digital Demodulation], [Return].)
Press [demodulation setup], [demod format], [standard setups], [NADC].

4. Select pulse and result lengths for this particular signal:

Press [Time], [search length], 500, [sym]
Press [result length], 156, [sym].

5. Select measurement, display, and analysis features. For example, to view the
default displays for all four traces:
Press [Display], [4 grids quad ]

6. The search length must be longer than result length and must include at least
one entire pulse. If the search length is long enough to include more than one
pulse, only the first pulse is demodulated. For additional details, see the
Digital Demodulation Concepts (Opt. AYA) chapter and see online help for
the [pulse search] and [search length] softkeys.

8-6
 Using Digital Demodulation (Opt. AYA)

Demodulating a pulsed signal

8-7
 Using Digital Demodulation (Opt. AYA)

To set up sync search

In this task you learn how to synchronize your measurement by using a

specific bit pattern within the chain of bits. You learn how to define sync
words and set an offset. Since sync search is often used with pulsed
signals, this example assumes you have already acquired and demodulated
a pulsed signal as shown in the previous task.
1. Select two displays and format them:
Press [Display], [2 grids]
Press [A], [Measurement Data], [IQ measured time].
Press [Data Format], [magnitude log(dB)].
Press [D].
2. Enter a sync bit pattern:
Press [Time], [sync setup], [pattern], [clear entry], 000101, [enter].
3. Select an offset:
Press [offset], 6, [sym].
4. Turn on sync search:
Press [Return], [sync search on]

The search length must be longer than the combination of result length,
sync pattern, and offset. The sync pattern may include up to 32 symbols.
The offset may be positive or negative. See the online help topics for
more information on these keys.

The sync word is highlighted when sync search is completed successfully

8-8
 Using Digital Demodulation (Opt. AYA)

To select and create stored sync patterns

When using sync search you can enter a sync bit pattern as in the previous
task, or you can load up to six of your own sync patterns into softkeys F3
through F8 and then use the softkeys to select a sync bit pattern. This
task assumes you have completed the previous task.
1. Insert the Signals Disk into the analyzer’s disk drive.
2. Load an example of user-defined sync patterns:
Press [Save/Recall], [catalog on].
Scroll to highlight SYNC_KEY.TXT
Press [recall more], [recall sync/state defs], [enter].
3. Choose one of the user-defined sync patterns:
Press [Time], [sync setup], [offset], 15, [sym], [user sync patterns].
Pres s one of the six user-defined softkeys to change sync patterns.
If you select Sync 1, Sync 5, or Sync 6 you see what happens if the
analyzer cannot find the sync pattern. The analyzer demodulates the signal
but displays the message “SYNC NOT FOUND.” When this happens the
result is positioned at the start of data collection. In this case the sync is
not found because the combination of offset and sync word place the
result length beyond the pulse. The other four sync words show the result
length on the leading edge, trailing edge, or center of the pulse.

You can create your own sync bit pattern definitions for the softkeys. See
the file “STAT_DEF” on the Signals Disk. The file may be viewed and
edited with any ASCII editor and the results may be saved to disk. If you
have IBASIC installed, you may use it as an editor. See “SYNC_KEY.TXT”
and “STATES.TXT” files to see a sync pattern and state definition that were
created using IBASIC to modify portions of the “STAT_DEF” file.

Up to 6 sync patterns may be loaded into softkeys to facilitate changing sync

patterns.

8-9
 Using Digital Demodulation (Opt. AYA)

To demodulate and analyze an EDGE signal

This task shows you how to demodulate an EDGE (Enhanced Data rates
for GSM Evolution) signal. The EDGE signal used in this task was
generated with the HP/Agilent E4433 ESG Series Signal Generator, with a
frequency of 5 MHz, amplitude of 0 dBm, and a framed (pulsed) data
format. Option B7A must be installed to demodulate EDGE signals.
1. Initialize the analyzer:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
2. Connect your EDGE signal to the channel 1 INPUT.
3. Set the center frequency (this example uses a 5 MHz EDGE signal):
Press [Frequency], [center], 5, [MHz]
4. Select the optimum range:
Press [Range].
Press the down-arrow key until the Channel-1 Over and Half LEDs are on.
Press the up arrow key one press at a time until the Over LED turns off.

For additional details about selecting the optimum range, see online help
for the [Range], [ch1 range] softkey.
5. Turn on digital demodulation:
Press [Instrument Mode], [Digital Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Digital Demodulation], [Return].)
6. Configure the analyzer to demodulate the EDGE signal:
Press [demodulation setup], [demod format], [standard setups], [EDGE].
7. Display four grids to view the vector diagram (trace A), error-vector trace
(trace B), eye diagram (trace C), and symbol table (trace D).
Press [Display], [4 grids quad ].
8. Autoscale traces A, B, and C:
Press [Shift] [B], [Shift] [C]
Press [Auto Scale].

8 - 10
 Using Digital Demodulation (Opt. AYA)

9. Display the ideal vector diagram in trace C and compare it to trace A:

Press [C], [Measurement Data], [IQ reference time]
Press [Data Format], [polar IQ vector]
Press [more format setup], [symbol dots]
Press [Shift] [A], [Auto Scale].

Selecting the EDGE standard setup automatically sets the demodulation

format, frequency span, symbol rate, filtering, and several other
demodulation parameters (for a complete list, see help for [standard setups]).

Because EDGE signals incorporate rotation and the EDGE filter introduces
ISI (inter-symbol interference), symbol locations in the vector and
constellation diagrams should appear random. However, at one
point/symbol (the value set by the EDGE standard setup) the analyzer
removes the effects of ISI so you can view a “clean” vector diagram.
Above one point/symbol, the analyzer does not remove the effects of ISI.

10. Set points/symbol to 2 to view the effects of ISI on the vector diagram:
Press [A], [Display], [single grid]
Press [Time], [points/symbol], 2, [enter].

Quad display showing individual IQ measured and IQ reference vector diagrams, the
error vector trace, and the symbol table (at 1 point/symbol). Single display showing
same vector diagram, but at 2 points per symbol. At 1 point/symbol, the analyzer
removes the effects of ISI so you can view a “clean” vector diagram. Above 1
point/symbol, the effects of ISI are not removed.

8 - 11
 Using Digital Demodulation (Opt. AYA)

To troubleshoot an EDGE signal

This task shows you how to use the error-vector trace and the symbol
table to diagnose problems with an EDGE signal. To simulate an error
condition, the task uses the analyzer’s source to mix a 5.1 MHz, –35 dbM
sine wave with the EDGE signal loaded in the previous task.
1. Demodulate your EDGE signal as shown in the previous task.
2. Make sure points/symbol are set to one:
Press [Time], [points/symbol], 1, [enter].
3. Mix the analyzer’s source with your EDGE signal.
Connect a T-connector to the CHANNEL 1 input.
Connect the analyzer’s SOURCE output to the T-connector.
Connect your EDGE signal to the T-connector.
4. Configure the source to output a 5.01 MHz, –35 dbM, sine wave:
Press [Source]
Press [level], [–35], [dbM]
Press [sine freq], [5.01], [MHz]
Press [source on]
5. Display the error-vector in trace A and autoscale the trace:
Press [A], [Measurement Data], [error vector time]
Press [Data Format], [polar IQ vector]
Press [Auto Scale],
6. Display two grids and put the symbol table in the lower trace.
Press [Display], [two grids].
Press [D] (By default, the symbol table is in trace D).

The error-vector vector diagram and symbol table provide good insight
into the quality of your EDGE signal. The next step shows you another
useful display: the error-vector spectrum.
7. Display the spectrum of the error-vector in trace A and auto-scale the trace.
Press [A], [Measurement Data], [Ierror vector spectrum]
Press [Auto Scale].

You use the same methods to troubleshoot an EDGE signal as you do

other types of digital signals. For example, you still use error parameters
in the symbol table to help you determine the quality of an EDGE signal.
The symbol table reports error parameters such as Mag Err (Magnitude
Error), Phase Err (Phase Error), Freq Err (Frequency Error), EVM (Error
Vector Magnitude), and SNR (Signal-to-Noise Ratio).

8 - 12
 Using Digital Demodulation (Opt. AYA)

Mag Err and Phase Err are especially useful to determine if your signal
contains AM, PM, spurious, or excessive noise errors. AM errors increase
Mag Err; PM errors increase Phase Err; spurious and noise errors increase
both Mag Err and Phase Err.

The EDGE demodulation format adds two new error parameters to the
symbol table: pk EVM and 95% EVM. pk EVM is the mean (average) of
the peak EVMs—one per measurement (it is a mean of the peaks, not a
peak of the peaks). 95% EVM is the error-vector-magnitude (EVM) below
which 95% of the individual symbol EVM’s occur.

The error-vector trace shows the error vector between the measured signal
(IQ measured) and the ideal signal (IQ reference) at the symbol locations.
The error-vector vector diagram shows the magnitude and phase errors at
the symbol locations. The error-vector spectrum shows undesired spectral
components at the symbol locations.

The error-vector vector diagram is a good

troubleshooting tool for EDGE signals. Here, the
error-vector vector diagram shows the effect of the
spur added to our EDGE signal—equal amplitude and The 5.01 MHz spur is clearly seen
phase errors. in the error-vector spectrum trace.

Two 2-grid displays. The first display shows the vector diagram of the error-vector
trace and the symbol table in the lower grid. The second display shows the
spectrum of the error-vector trace and the symbol table in the lower grid.

8 - 13
 Using Digital Demodulation (Opt. AYA)

To demodulate and analyze an MSK signal

This example uses an MSK signal from the Signals Disk to show you how
to demodulate an MSK signal and view MSK phase transitions.

1. Load “MSK.DAT” from the Signals Disk into the arbitrary source:
Perform steps 1, 2, and 3 in ‘’To prepare a digital demodulation
measurement.’’ For step 2, load the “MSK.DAT” signal from the Signals
Disk instead of ‘’PI4DQPSK.DAT.’’

2. Select appropriate setup parameters:

Press [Range] and select the optimum range.
Press [Frequency], [center], 5, [MHz]
Press [span], 1, [MHz].
Press [Trigger], [trigger type], [internal source], [return], [ch1 delay], 74, [us].

If you don’t know how to select the optimum range, see ‘’To prepare a
digital demodulation measurement.’’

3. Demodulate the signal:

Press [Instrument Mode], [Digital Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Digital Demodulation], [Return]).

When you turn on digital demodulation, you may see this message:
Maximum span limited by symbol rate and maximum span/symbol rate ratio.
This message informs you that the frequency span is limited by the
currently selected symbol rate. In other words, the symbol rate used in
the last digital demodulation measurement is too small. The next step
automatically selects the correct symbol rate and restores the frequency
span to 1 MHz, so you can ignore this message. For additional details
about symbol rate and frequency span interactions, see ‘’Parameter
Interactions’’ in the Digital Demodulation Concepts chapter.

4. Configure the digital demodulator for a GSM measurement:

Press [demodulation setup], [demod format], [standard setups], [GSM]
Press [Time], [pulse off], [sync off]

5. Format traces A and C to view unwrapped phase versus group delay:

Press [ A]
Press [Data Format], [phase unwrap]
Press [Display], [view/overlay traces], [C on]
Press [C], [Measurement Data], [IQ measured time], [Data Format], [group delay]
Press [Shift], [A], [more format setup], [symbols dots ]
Press [Auto Scale]

8 - 14
 Using Digital Demodulation (Opt. AYA)

6. Format traces B and D to view the reference versus the measured trellis
diagram:
Press [Display], [2 grids], [view/overlay traces], [D on]
Press [B], [Measurement Data], [IQ reference time]
Press [D], [Measurement Data], [IQ measured time]
Press [Shift], [B]
Press [Data Format], [eye diagram trellis]
Press [more format setup], [eye length], 4, [enter], [Auto Scale]

The upper grid shows the relationship between instantaneous frequency and
instantaneous phase. The trellis diagram in the lower grid presents another view of
phase response.

8 - 15
 Using Digital Demodulation (Opt. AYA)

To demodulate a two-channel I/Q signal

Note This measurement can only be performed with a 2-channel analyzer—you must
have option AY7 (option AY7 adds a second input channel).

If you have separate baseband I and Q signals available for your

measurement, you may demodulate them directly if you have a two-channel
analyzer. This type of demodulation preserves the original transmitted
relationship between the I and Q signals.

1. Apply real I and Q signals to Channel 1 and Channel 2 respectively. Be sure to

use the inputs on the upper (IF) section of the analyzer.

2. Select the special baseband receiver mode:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (ch1 + j*ch2)]
89441A: [IF section (ch1 + j*ch2)]
Press [Preset].

3. Adjust the frequency span to encompass the signal with a span of at least 78
kHz.

4. Make sure that time domain calibration is on under [System Utility]

5. Select identical parameters for both channels under the [Range] key.

6. Select identical parameters for both channels under the [Input] key.

7. Select identical parameters for both channels under the [Trigger] key.

8. Proceed with digital demodulation as shown previously.

For more information on this type of measurement see online help for the
[input section (ch1 + j*ch2)] (89410A) or [IF section (ch1 + j*ch2)]
(89441A) key.

8 - 16
9

Using Video
Demodulation
(Opt. AYH)
This chapter shows you how to use digital video demodulation to
demodulate and view digitally-modulated video signals. You may perform
the tasks in this chapter using signals from the Signals Disk, or you may
use these tasks as a model for demodulating your own signals.

9-1
 Using Video Demodulation
(Opt. AYH)

To prepare a VSB measurement

This task shows you how to load and view the 8 VSB signal located on the
Signals Disk. If you have your own 8 VSB signal, use the steps below and
enter the demodulation parameters for your signal.

1. Initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [vector].
Press [System Utility], [memory usage], [configure meas memory)], [max time pts], 4096.

2. Supply an 8 VSB signal to the channel 1 INPUT or perform the following steps
to load an 8 VSB signal from the Signals Disk into the analyzer’s time-capture
RAM:
Insert the Signals Disk into the internal disk drive.
Press [Save/Recall], [default disk], [internal disk].
Press [Return], [catalog on].
Rotate the knob to highlight 8VSB.CAP
Press [recall more], [recall capture buffer], [enter].

3. Turn on averaging:
Press [Average] [average on].

4. Measure and scale the displayed trace:

Press [Measurement Restart], [Auto Scale].

When you measure time-capture data, the analyzer automatically sets its
frequency span and center frequency to that used to capture the data.
Therefore, you did not need to set these parameters in the above steps.

If you are not measuring time-capture data, you must set the center
frequency, frequency span, and range. If these parameters are incorrect,
the analyzer may not lock to your carrier, measurement speed may be
reduced, or you may see excessive errors in the demodulated results. For
details about setting these parameters, see ‘’Carrier locking’’, ‘’Input Range’’,
and ‘’Span considerations’’ in the Video Demodulation Concepts chapter.
The next task, ‘’To determine the center frequency for a VSB signal.’’,
shows you how to determine the correct center frequency.

9-2
 Using Video Demodulation
(Opt. AYH)
VSB measurements typically require a large portion of measurement
memory. Therefore, it is a good idea to choose the maximum value (4096)
for [max time points] (see step 1). For details about [max time points], see online
help (press [Help], then press [max time points]).

Note Before demodulating a VSB signal, view the signal in Vector mode to verify that
the pilot is on the left (low side) of the spectrum. If it isn’t, you must configure
the analyzer to demodulate a high-side pilot. For further details, see
‘’To demodulate a VSB signal.’’

To demodulate VSB signals,

the pilot must be on the left
(low side) of the spectrum. If
it isn’t, you must configure the
analyzer to demodulate a
high-side pilot, as shown in
‘’To demodulate a VSB
signal.’’.

Spectrum of an 8 VSB signal

9-3
 Using Video Demodulation
(Opt. AYH)

To determine the center frequency for a VSB signal

Choosing the correct center frequency is important for all digital video
demodulation measurements. This task shows you how to determine the
correct center frequency for VSB measurements. To learn how to
determine the correct center frequency for QAM measurements, see the
‘’Video Demodulation Concepts’’ chapter. Note that you cannot use the
time-capture signal from the signals disk to perform this task. This task
uses the analyzer’s frequency counter, which cannot be used on time
capture data.

1. Initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [vector].

2. Increase the display resolution:

Press [ResBw/Window].
Press [num freq pts] and set to 1601.
Press [rbw mode arb].
Press [rbw] and press the down-arrow key to select the smallest resolution
bandwidth.

3. Position the marker on the pilot:

Press [Shift], [Marker].

4. Turn on the frequency counter:

Press [Marker Function], [freq counter on].

5. Compute the ideal center frequency:

Symbol Rate
Center Frequency (LOW SIDE PILOT) = + (Pilot Frequency)
4

Symbol Rate
Center Frequency (HIGH SIDE PILOT) = (Pilot Frequency) −
4

In this example, the symbol rate is 10.762 MHz, the frequency counter
shows the pilot frequency at 3.308894876 MHz, and the pilot is on the low
side of the spectrum. Using the formula for low-side pilot, the ideal
center frequency is 5.999394876 MHz. A center frequency of 6 MHz is
close enough to ensure carrier lock.

9-4
 Using Video Demodulation
(Opt. AYH)
6. Set the center frequency to the computed value:
Press [Frequency], [center], 6 [MHz)].

Note If your pilot is on the high (right) side of the spectrum, you must configure the
analyzer to demodulate a high-side pilot. For further details, see the next task:
‘’To demodulate a VSB signal.’’

Frequency counter readout. In this

example, this is the frequency of the pilot
signal.

In this example, the pilot is

on the low (left) side of the
spectrum.

8 VSB Signal With Low-Side Pilot

9-5
 Using Video Demodulation
(Opt. AYH)

To demodulate a VSB signal

This task shows you how to demodulate a VSB signal. The task uses the
8 VSB time-capture signal that you loaded into the analyzer in ‘’To prepare
a VSB measurement’’.

1. Prepare the analyzer for a VSB measurement as shown in the ‘’To prepare a
VSB measurement’’.

2. Turn averaging off and restart the measurement:

Press [Average] [average off].
Press [Measurement Restart].

3. Demodulate the signal:

Press [Instrument Mode], [demod type], [Video Demodulation], [Return].

4. Select the correct demodulation parameters for the 8 VSB signal (if you are not
using the signal provided on the Signals Disk, enter the parameters for your
signal) :
Press [demodulation setup].
Press [demod format], [VSB 8], [Return].
Press [symbol rate], 10.762, [MHz]
Press [result length], 800, [sym]
Press [ref filter], [raised cosine]
Press [Return], [meas filter], [root raised cosine].
Press [alpha/BT], .1152
Press [Time], [points/symbol], 5.
Press [pulse search off], [sync search off].

5. If you are using the 8 VSB signal from the Signals Disk, skip the next step.

6. If the pilot is on the right (high side) of the spectrum, configure the analyzer to
demodulate a high-side pilot:
Press [Instrument Mode], [demodulation setup], [more], [freq spectrum mirror].

To learn what [freq spectrum mirror] does, see online help (press [Help], then
press [freq spectrum]).

9-6
 Using Video Demodulation
(Opt. AYH)
7. View the constellation and eye diagram:
Press [Display], [2 grids],
Press [A], [Measurement Data], [IQ measured time]
Press [Data Format], [polar (IQ) constellation].
Press [B], [Measurement Data], [IQ measured time]
Press [Data Format], [eye diagram I].

With VSB signals, symbol locations (detection decision points) are derived
from the real portion (I) of the demodulated data. This is evident in the
constellation diagram where you see symbols aligned vertically in 8
locations (16 locations for 16 VSB) along the I-axis. The vertical lines in
the constellation diagram indicate ideal symbol locations.

Hint Displayed data must contain real data to see symbol locations for VSB signals.
For example, if you press [Data Format] and select the imaginary part of the data
or the Q eye-diagram, you won’t see symbol information.

Constellation and eye diagram for 8 VSB signal

9-7
 Using Video Demodulation
(Opt. AYH)

To prepare a QAM or DVB QAM measurement

This task shows you one way to set up a QAM or DVB QAM measurement.
The task uses the RF section (0-10 MHz) receiver and a 32 DVB QAM
signal from the signals disk. Several other tasks in this chapter use this
setup to teach you how to use video demodulation.
1. Initialize the analyzer:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [vector].
Press [System Utility], [memory usage], [configure meas memory)], [max time pts], 4096.

2. Supply a 32 DVB QAM signal to the channel 1 INPUT or perform the following
steps to load a 32 DVB QAM signal from the Signals Disk into a data register
and play it through the analyzer’s aribtrary source:
Insert the Signals Disk into the internal disk drive.
Connect the SOURCE to the channel 1 INPUT.
Press [Save/Recall], [default disk], [internal disk].
Press [Return], [catalog on].
Rotate the knob to highlight 32DVBQAM.DAT
Press [recall trace], [from file into D1], [enter].
Press [Source], [source on], [source type], [arbitrary].
3. Select the optimum range:
Press [Range].
Press the down-arrow key until the Channel-1 Over and Half LEDs are on.
Press the up arrow key one press at a time until the Over LED turns off.

For additional details about selecting the optimum range, see online help
for the [Range], [ch1 range] softkey.
4. Select a center frequency and span:
Press [Frequency], [center], 5, [MHz]
Pr ess [span], 2.5, [MHz].

The center frequency tunes the analyzer to the carrier frequency. To

obtain reliable carrier lock, the center frequency must be close to the
carrier frequency. For details, see ‘’Carrier Locking’’ in the
Video Demodulation Concepts (Opt. AYA) chapter.

Selecting the correct frequency span is also important when using video
demodulation. The span must be wide enough to include all signal
components, and yet not too wide, or the measurement may be affected
by excessive noise and slower speed. For details, see ‘’Parameter
Interactions’’ in the Video Demodulation Concepts (Opt. AYA) chapter.

9-8
 Using Video Demodulation
(Opt. AYH)
5. Set up the trigger:
Press [Trigger], [trigger type], [internal source], [return], [ch1 delay], 3, [ms].

This example uses the 32 DVB QAM signal from the Signals Disk. This
signal was generated following a procedure similar to that shown in
chapter 9, ‘’To create an ideal digitally modulated signal.’’ The method
used to create this signal results in some invalid data at the beginning of
the time record. The 3 milli-second trigger delay removes the invalid data
from the measurement.

When you supply any signal to the channel 1 input you need to select
appropriate center frequency, span, range, and triggering parameters prior
to demodulating the signal. Display formats and measurement types may
be applied and changed after demodulation.

This task sets [max time pts] to its maximum value (4096), which allocates the
maximum amount of measurement memory for digital video demodulation.
This lets you choose larger result lengths, search lengths, and
points-per-symbol. For additional details about allocating memory for
digital video demodulation, see ‘’Parameter Interactions’’ in the Video
Demodulation Concepts chapter and see online help for [max time points],
(press [Help], then press [max time points]).

QAM and DVB QAM measurements treat I/Q origin offset differently. QAM
measurements remove I/Q origin offset, DVB QAM measurements do not
remove I/Q origin offset. Both demodulation formats report I/Q origin
offset (in the symbol table).

The spectrum of a 32 DVB QAM signal before demodulation.

9-9
 Using Video Demodulation
(Opt. AYH)

To demodulate a QAM or DVB QAM signal

This task shows you how to demodulate the 32 DVB QAM signal generated
in ‘’To prepare a QAM or DVB QAM measurement’’. Prior to demodulating
a video signal you must select the correct center frequency, frequency
span, and range as shown in that task. You use the same procedure to
demodulate both QAM and DVB QAM signals.

1. Prepare the analyzer for a DVB QAM measurement as shown in the previous
task.

2. Demodulate the signal:

Press [Instrument Mode], [demod type], [Video Demodulation], [Return].

3. Select the correct demodulation parameters for the 32 DVB QAM signal (if you
are not using the signal provided on the Signals Disk, enter the parameters for
your signal) :
Press [Instrument Mode], [demod type], [Video Demodulation], [Return] .
Press [demodulation setup].
Press [demod format], [DVB QAM 32], [Return].
Press [symbol rate], 1, [MHz]
Press [result length], 400, [sym]
Press [ref filter], [raised cosine]
Press [Return], [meas filter], [root raised cosine].
Press [alpha/BT], .15
Press [Time], [points/symbol], 5.
P ress [pulse search off], [sync search off].

4. View the created signal versus the reference:

Press [Display], [2 grids],
Press [A], [Measurement Data], [IQ measured time]
Press [B], [Measurement Data], [IQ reference time]
Press [Shift], [A] to activate both traces.
Press [Data Format], [polar (IQ) vector].
Press [more format setup], [symbol dots] (if [symbol dots] is already selected, you still
must press this key to force both traces to display symbols as dots).

5. Scale both traces:

Press [Auto Scale]

If you are familiar using digital demodulation, you may have noticed that
setting up a video demodulation measurement and a digital demodulation
measurement is identical for QAM and DVB QAM measurements (you can
demodulate 16 QAM and 32 QAM signals with digital demodulation or
video demodulation).

9 - 10
 Using Video Demodulation
(Opt. AYH)

Hint You use [demodulation setup] to set demodulation parameters, you use
[Measurement Data] to select the measurement calculation used on demodulated
data, and you use [Data Format] to select a display format (trace coordinates). To
learn more about these keys and the choices under them, see online help.
Online help contains detailed descriptions for all keys (press [Help], then press
the desired key). The next task uses [Measurement Data] and [Data Format] to
display a constellation diagram and the error-vector trace.

Two time displays of a demodulated signal: IQ measured versus IQ Reference

9 - 11
 Using Video Demodulation
(Opt. AYH)

To select measurement and display features

You can display demodulated data in many different formats. This task
uses the demodulated 32 DVB QAM signal from the previous task to show
you just a few ways of viewing demodulated data.
1. Select multiple display grids:
Press [Display], [4 grids quad ].
2. Change the data format for trace A:
Press [A], [Data Format], [polar IQ constellation].
3. Change the measurement data for trace B:
Press [B], [Measurement Data], [error vector time].
4. Scale traces A, B, and C:
Press [Shift], [A], [Shift], [C] to activate traces A, B, and C.
Press [Auto Scale].

By default, selecting 4 grids displays the current trace in trace A, the

error-vector trace in trace B, the eye diagram in trace C, and the symbol
table in trace D. Thus you can view demodulated data in four different
ways at the same time.

You can change the [Measurement Data] and [Data Format] for any trace.
Simply activate the trace (for example, press [A] to activate trace A), then
select the desired measurement data and data format.

Each grid shows a different measurement type with an appropriate data format.

9 - 12
 Using Video Demodulation
(Opt. AYH)

To set up sync search (QAM only)

In this task you learn how to synchronize your measurement by using a

specific bit pattern within the chain of bits. You learn how to define sync
words and set an offset. Sync search operates the same for both digital
and video demodulation. This example uses the 32 DVB QAM signal
created in ‘’To prepare a QAM or DVB QAM measurement’’. Note that you
cannot use sync search with VSB measurements.
1. Select two displays and format them:
Press [A], [Data Format], [magnitude log(dB)].
Press [Display], [2 grids]
Press [D].
2. Select the search length for this particular signal:
Press [Time], [search length], 1000, [sym].
3. Enter a sync bit pattern:
Press [sync setup], [pattern], [clear entry], 00101, [enter].
4. Select an offset:
Press [offset], 6, [sym].
5. Turn on sync search:
Press [Return], [sync search on]

The search length must be longer than the combination of result length,
sync pattern, and offset. The sync pattern may include up to 32 symbols.
The offset may be positive or negative. See the online help topics for
more information on these keys.

The sync word is highlighted when sync search is completed successfully

9 - 13
 Using Video Demodulation
(Opt. AYH)

To select and create stored sync patterns (QAM only)

When using sync search you can enter a sync bit pattern as in the previous
task, or you can load up to six of your own sync patterns into softkeys F3
through F8, and then use the softkeys to select a sync bit pattern. This
task uses the results of the previous task.
1. Insert the Signals Disk into the analyzer’s disk drive.
2. Load an example of user-defined sync patterns:
Press [Save/Recall], [catalog on].
Scroll to highlight SYNC_KEY.TXT
Press [recall more], [recall sync/state defs], [enter].
3. Choose one of the user-defined sync patterns:
Press [Time], [sync setup], [offset], 15, [sym], [user sync patterns].
Pres s one of the six user-defined softkeys to change sync patterns.
If you select Sync 1, Sync 5, or Sync 6 you see what happens if the
analyzer cannot find the sync pattern. The analyzer demodulates the signal
but displays the message “SYNC NOT FOUND.” When this happens the
result is positioned at the start of data collection. In this case the sync is
not found because the combination of offset and sync word place the
result length beyond the pulse. The other four sync words show the result
length on the leading edge, trailing edge, or center of the pulse.

You may create your own sync bit pattern definitions for the softkeys. See
the file “STAT_DEF” on the Signals Disk provided with this documentation.
The file may be viewed and edited with any ASCII editor and the results
may be saved on a disk. If you have IBASIC installed, you may use it as
an editor. You may view the files “SYNC_KEY.TXT” and “STATES.TXT” to
see a sync pattern and a state definition which were created by using
IBASIC to modify portions of the “STAT_DEF” file.

Up to 6 sync patterns may be loaded into softkeys

9 - 14
 Using Video Demodulation
(Opt. AYH)

To demodulate a two-channel I/Q signal

Note This measurement can only be performed with a 2-channel analyzer—you must
have option AY7 (option AY7 adds a second input channel).

If you have separate baseband I and Q signals available for your

measurement, you may demodulate them directly if you have a two-channel
analyzer. This type of demodulation preserves the original transmitted
relationship between the I and Q signals.
1. Apply real I and Q signals to Channel 1 and Channel 2 respectively. If you have
an 89441A, be sure to use the inputs on the upper (IF) section of the analyzer.
2. Select the special baseband receiver mode:
Press [Instrument Mode], [receiver], then press:
89410A: [ input section (ch1 + j*ch2)].
89441A: [ IF section (ch1 + j*ch2)].

3. Adjust the frequency span to encompass the signal with a span of at least 78
kHz.
4. Make sure that time domain calibration is on under [System Utility]
5. Select identical parameters for both channels under the [Range] key.
6. Select identical parameters for both channels under the [Input] key.
7. Select identical parameters for both channels under the [Trigger] key.
8. Proceed with digital demodulation as shown previously.

For more information on this type of measurement see online help for the
[input section (ch1 + j*ch2)] (89410A) or [IF section (ch1 + j*ch2)] (89441A) key.

9 - 15
10

Analyzing Digitally
Demodulated Signals
(Options AYA and AYH)
This chapter shows you how to analyze signals demodulated with Digital
Demodulation or with Video Demodulation. The tasks in this chapter use
digital demodulation, but the same steps apply to video demodulation. You
will learn how to select measurement data and compatible data formats,
use polar markers, examine symbol tables, display state definitions, and
examine errors.

10 - 1
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To demodulate a non-standard-format signal

This task uses a 16 QAM signal from the Signals Disk to teach you how to
demodulate a non-standard signal. The task uses digital demodulation. If
you want to perform this task using video demodulation, choose video
demodulation instead of digital demodulation in step 4. The Video
Demodulation Concepts (Opt. AYH) chapter explains how video
demodulation differs from digital demodulation.

1. Initialize the analyzer:

Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [Vector]

2. Load “16QAM.DAT” from the Signals Disk and play it through the arbitrary
source:
Perform steps 1, 2, and 3 in ‘’To prepare a digital demodulation
measurement’’ in chapter 6. For step 2, load the “16QAM.DAT” signal
from the Signals Disk instead of ‘’PI4DQPSK.DAT.’’

3. Select appropriate setup parameters:

Press [Range] and select the optimum range.
Press [Frequency], [center], 500, [kHz]
Press [span], 78, [kHz].
Press [Trigger], [trigger type], [internal source], [return], [ch1 delay], 150 [us].

If you don’t know how to select the optimum range, see ‘’To prepare a
digital demodulation measurement’’ in chapter 6.

4. Digitally demodulate the signal:

Press [Instrument Mode], [Digital Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Digital Demodulation], [Return] ).

10 - 2
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

5. Choose demodulation setup parameters:

Press [Time], [pulse search off], [sync search off].
Press [points/symbol], 5, [enter]
Press [Instrument Mode], [demodulation setup], [demod format], [QAM 16]
Press [Return], [symbol rate], 24.3, [kHz].
Press [result length], 100, [sym]
Press [meas filter], [root raised cosine]
Press [Return], [ref filter], [raised cosine]
Press [Return], [alpha / BT], .35, [enter]

6. Display the vector diagram:

Press [Data Format], [polar (IQ) vector].

As shown in this task, with non-standard signals you must select individual
parameters for demodulation setup. For furhter information on
demodulation-setup parameters, see the Digital Demodulation Concepts
chapter if you are using digital demodulation; see the Video Demodulation
Concepts chapter if you are using video demodulation.

Vector diagram for 16 QAM signal

10 - 3
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To use polar markers

This task shows you how to select the polar-marker format (magnitude and
phase or real and imaginary) and polar-marker units (dBm, Watts, or volts).
This task is a continuation of the previous task.
1. Select result coordinate power calculation:
Press [Instrument Mode], [demodulation setup], [more], [normalize off].
2. Select the desired polar marker format:
Press [Marker], [polar mkr setup], [format mag & phase], [W].
3. Press [Auto Scale]
4. Rotate the knob to examine time points along the trajectory.
5. Select a different number of points per symbol:
Press [Time], [points/symbol], 10, [enter].

The vector diagram shows all time points on the trajectories between
decision points. More points per symbol creates a smoother vector
diagram. If you want to examine the marker value only at the decision
points you may select a constellation diagram. The cross-hairs show the
ideal detection-decision points (for details, see online help for [Data Format]
[more format setup] [ideal state]).

The measurement values reflect your

choice of polar marker format and
whether the measurement reflects
normalized or power units

In a vector diagram, the marker allows you to track all time points.

10 - 4
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To view a single constellation state

In this task you use the marker as a reference to reposition a constellation

state to the center of the screen and zoom in. This task is a continuation
of the previous task.
1. Turn normalization on, select constellation data format, then pause the
measurement:
Press [Instrument Mode], [demodulation setup], [more], [normalize on].
Press [Data Format], [polar IQ constellation].
Press [Pause|Single].
2. Rotate the knob to move the main marker to a constellation point in the upper
right corner.
3. Reposition the constellation:
Press [Shift], [Marker Function].
4. Rescale and zoom the display:
Press [Ref Lvl/Scale], [Y per div], .03, [enter].
Press [Marker|Entry] and use the knob to zoom in and out.

The marker lets you view and zoom a single constellation state

10 - 5
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To locate a specific constellation point

You now use the offset marker as a pointer to snap the main marker to a
constellation point. This is more convenient than searching linearly
through time in order to position the main marker on a desired
constellation point. This task is a continuation of the previous task.
1. Turn on the offset marker, and place it on the main marker:
Press [Shift], [Marker→].
2. Move the offset marker close to the desired point:
Press [Marker], [offset posn setup], [offset x posn]
Rotate the knob counterclockwise until the x position reads approximately
690m.
3. Snap the main marker to the constellation point closest to the offset marker:
Press [Marker Search], [marker to offset mkr].
4. Enhance the brightness of the crosshairs identifying the optimal constellation
point:
Press [Display], [more display setup], [color setup], [color index], 7, [enter]
Press [luminosity], 60, [%].
This change remains in effect for all trace grids until you return the
luminosity back to 48% using the same procedure.
5. Before proceeding to later tasks, return to previous display and measurement
conditions:
Press [Auto Scale]
Press [Pause|Single]
Press [Marker|Entry]

The offset marker lets you place the marker on a specific constellation point

10 - 6
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To use X-axis scaling and markers

This task shows you how to zoom-in on a selected portion of the x-axis.
This task is a continuation of the previous task.
1. Select both A and B as active traces:
Press [Display], [2 grids].
Press [A], [Shift], [B].
2. Select measurement data and data format :
Press [Measurement Data], [error vector time].
Press [Data Format], [magnitude linea r], [more format setup], [symbol bars]
Press [Auto Scale].
3. Examine a portion of the X-axis with the X scale markers:
Press [A], [Ref Lvl/Scale], [X scale markers].
Press [center ref], 40, [sym].
Press [width], 20, [sym].
Press [scale at markers].
4. Change the location of the markers using the knob:
Press [Marker|Entry] to highlight the ENTRY LED.
Press [right ref] and rotate the knob to relocate the right-reference marker.
Press [left ref] and rotate the knob to relocate the left-reference marker.
5. Return the knob to marker mode and the display to full scale before performing
later tasks:
Press [Marker|Entry],
Press [X full scale].

Bars appear at symbol

decision points

X-axis markers let you examine portions of the trace

10 - 7
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To examine symbol states and error summaries

This task shows you how to display the symbol table, which contains
demodulated bits and numeric error information. This task also shows you
how to couple the marker in the symbol table to the constellation diagram
so you can see the bits that correspond to a state. This task is a
continuation of the previous task.

1. Select two display grids:

Press [Display], [2 grids].

2. Display the symbol table in the upper grid:

Press [A], [Measurement Data], [symbol table/error summary].

3. Display a constellation diagram in the lower grid:

Press [B], [Measurement Data], [IQ measured time]
Press [Data Format], [polar IQ constell ation].

4. Turn on symbol dots on the constellation display:

Press [more format setup], [symbol dots ]

5. Press [Auto Scale].

6. Couple the markers on the two grids:

Press [Marker], [couple mkrs on].

7. Rotate the knob to move the marker from symbol to symbol in both the symbol
table (trace A) and the constellation diagram (trace B).

8. Turn on averaging to observe averaged numeric error data:

Press [Average], [average on].

9. Turn averaging off before continuing to other tasks.

Press [average off].

Display online help for more information on these keys and topics.

10 - 8
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

A Marker 3.00000 sym 15.000

EVM = 859.14 m%rms 1.4978 % pk at sym 55

MagErr = 609.17 m%rms 1.4317 % pk at sym 14
Phase Err = 623.96 mdeg 2.1174 deg pk at sym 44
Freq Err = 1.7722 Hz
IQ Offset = -66.106 dB SNR = 39.101 dB

0 001010100110 1111 01010001000010011010111110010010

48 000011110000000010000111100100000110000011010001
96 000011111011110001110111011101101010111110100110
144 000001100010111010110011100111111001010000110000
Averaging may be applied to
192 110000011110010101110100101111010000111111000010
240 011011110101010110100001000000000111100010111110
the numeric error summaries
288 011110001110111100101111100001101111001110011101
336 000011000110110011011000001011110000010001001110

TRACE B: Ch1 16QAMMeasTime

B Mkr 3.000000 sym 998.80 m -134.74 deg
1.5

Const

300
m
/div

-1.5

The symbol state table, which displays the binary bits for each symbol at
the decision points, may be viewed and compared to other displays of the data.

10 - 9
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To view and change display state definitions

This task shows you how to view and change the state definitions
corresponding to the detection decision points (symbol locations). You can
view and change state definitions for most modulation formats. This task
is a continuation of the previous task.

1. Examine the current state definitions:

Press [Instrument Mode], [demodulation setup]
Pre ss [demod format], [display state definitions].

Pressing [display state definitions] displays the state definitions for the currently
selected demodulation format. To see the state definitions for a different
demodulation format, select the demodulation format, then press
[display state definitions]. If the selected demodulation format doesn’t require
state definitions (such as MSK or pi/4 DQPSK), pressing [display state definitions]
does nothing.

2. Recall new state definitions:

Insert the Signals Disk in the analyzer’s disk drive.
Press [Save/Recall], [catalog on]
Rotate the knob to highlight STATES.TXT
Press [recall more], [recall sync/state defs], [enter].

3. Examine the new state definitions:

Press [Instrument Mode], [demodulation setup]
Press [demod format], [display state definitions].

To learn how to create your own state definitions, see the “STAT_DEF” file
on the Signals Disk. This file contains instructions and sample state
definitions for all modulation formats that have modifiable state definitions.
You can edit the file with any ASCII editor or with the IBASIC editor (if
you have the IBASIC option installed). You may want to view the
“STATES.TXT” file on the Signals Disk, which was created by modifying
and saving a portion of the “STAT_DEF” file.

10 - 10
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

Note Note that for video demodulation (option AYH), you cannot display or change
the state definitions for DVB QAM. State definitions for DVB QAM are fixed as
defined in the European Telecommunication Standard (online help for the
[DVB QAM] softkey shows the state definitions as defined in this standard).

State definitions are shown in a format that corresponds to the constellation diagram.

10 - 11
 Analyzing Digitally Demodulated Signals (Options AYA and AYH)

To view error displays

This task shows you how to view several different error displays, such as
the error-vector magnitude (EVM), magnitude error, and phase error at
each symbol point. This task is a continuation of the previous task.
1. Select four grids:
Press [Display], [4 grids stack ].
2. Select the symbol table/error summary for the top grid:
Press [A], [Measurement Data], [symbol table/error summary ].
3. Select an error vector display on trace B:
Press [B], [Measurement Data], [error vector time ]
Press [Data Format], [magnitude linear ].
4. Select an error magnitude display on trace C:
Press [C], [Measurement Data], [IQ error mag]
Press [Data Format], [part real].
5. Select an error phase display on trace D:
Press [D], [Measurement Data], [IQ error phase]
Press [Data Format], [phase wrap ].
6. Activate the three lower traces and scale them:
Press [D], [Shift], [C], [Shift], [B].
Press [Auto Scale].
7. Format the display to show bars only at the decision points:
Press [Data Format], [more format setup], [symbols bars ]
Press [Time], [points/symbol], 1, [enter], [Auto Scale].
8. Rotate the knob to view errors at the decision points.

The combined error vector

magnitude of all decision points.

The error vector magnitude

of each decision point.
The magnitude error of each
decision point.
The phase error of each
decision point.

Use multiple grids to view error with various measurement data and data formats.

10 - 12
11

Creating User-defined
Signals (Options AYA
and AYH)
This chapter shows you how to create your own digitally modulated signals.

11 - 1
 Creating User-defined Signals (Options AYA and AYH)

To create an ideal digitally modulated signal

You may create a digitally modulated signal by using noise as the input and
saving the reference signal of your selected demodulation format. You may
check the created signal by playing it through the source as shown in the
following task. This example creates a 32QAM signal but most digitally
modulated signal types may be created in a similar way (see text on
following page). The waveform created by the following procedure is
included on the Signals Disk as “32QAM.DAT”.

This task uses digital demodulation. If you want to perform this task using
video demodulation, choose video demodulation instead of digital
demodulation in step 3.
1. Initialize the analyzer:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [vector].
2. Select a center frequency and span:
Press [Frequency], [center], 5, [MHz]
Press [span], 61, [kHz].
3. Select a digital demodulation format:
Press [Instrument Mode], [Digital Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Digital Demodulation], [Return] ).
Press [demodulation setup].
Pre ss [demod format], [QAM 32].
4. Select demodulation format parameters:
Press [Return], [symbol rate], 30.5, [kHz],
Press [result length], 240, [sym]
Press [ref filter], [root raised cosine]
Pres s [Return], [alpha / BT], .5, [enter]
Press [more], [normalization on].
Press [Time], [points/symbol], 5, [enter],
Press [pulse search off], [sync search off].
5. Select a reference time display (the Data Format is irrelevant):
Press [Measurement Data], [IQ reference time]
Press [Data Format], [polar (IQ) constellation]
Press [Auto Scale].
6. Save the signal to a register:
Press [Save/Recall], [save trace], [into D1].
7. You may want to the save the created signal to a disc for more permanent
storage:
Press [Save Recall], [save trace], [into file], and enter a file name.

11 - 2
 Creating User-defined Signals (Options AYA and AYH)

You may use the analyzer to create custom arbitrary waveforms

corresponding to digital communication signals. Since the I/Q reference
signal is the ideal representation of a format type, a properly saved version
of the reference signal provides an ideal waveform. The internally
generated waveforms may be used as test signals (to test an amplifier, for
example).

The following guidelines may help you create a model arbitrary waveform:
l You cannot use this procedure to create 8 VSB or 16 VSB signals.
l This procedure is unreliable with 64 QAM, 64 DVB QAM, or 256 QAM signals.
You may or may not be able to create a valid signal using these formats.
l You must be sure that no external signal is applied to channel 1—the
analyzer’s internal noise is used to create the signal.
l Although span is irrelevant in creating signals, a span of twice the symbol rate
results in faster demod updates when you play back the waveform.
l Points per symbol should be either 5 or 10 (if you are creating an MSK signal
you must use at least 10 points per symbol).
l You should use a result length which is at least 10 symbols longer than what
you want to use as a test signal. This permits a 5 symbol truncation at each end
of the record to eliminate possible invalid data caused by discontinuities
between the beginning and end of the waveform. For theoretically complete
settling, approximately 5 symbols are neeed at each end of the waveform.
However, burst system specifications disregard settling issues because it is
impossible to instantly settle a Nyquist filter. The majority of the settling is
complete after 1 symbol. With five symbols of settling, the effect on error vector
magnitude is below algorithm residual error for a root raised cosine measure
filter.
l You must select IQ reference time.
l If you select a modulation type which employs distributed filtering, you must
select an appropriate filter type. For example, some format types define the
reference filter as a raised cosine type, since the reference must normally
account for square root filtering in the transmitter and square root filtering in
the receiver. The cascade of the two is full filtering. However, to simulate a
transmitter, only half filtering, that is, square root filtering should be used.
Therefore, when you save this type of reference you must define the filter type
as root raised cosine so that when you play it back as a stimulus it is partially
filtered, allowing the demodulator to apply the additional filtering.

11 - 3
 Creating User-defined Signals (Options AYA and AYH)

To check a created signal

This section assumes you have created a signal as shown on the previous
page, have not changed any setup parameters, and have not preset the
instrument.

This task uses digital demodulation. If you want to perform this task using
video demodulation, choose video demodulation instead of digital
demodulation in step 4.

1. Connect the source output to the channel 1 input and play the signal through
the arbitrary source:
Press [Source], [source on], [source type], [arbitrary]. The default data register is
D1.

2. Select the optimum range:

Press [Range].
Press the down-arrow key until the Channel-1 Over and Half LEDs are on.
Press the up arrow key one press at a time until the Over LED turns off.

For additional details about selecting the optimum range, see online help
for the [Range], [ch1 range] softkey.

3. Select the internal trigger and choose a 5 symbol delay to truncate invalid data
at the beginning of the record:
Press [Trigger], [trigger type], [internal source], [Return], [ch1 delay], 164, [us].

4. Select the correct demodulation parameters:

Press [Instrument Mode], [Digital Demodulation] (with option AYH, press [Instrument Mode],
[demod type], [Digital Demodulation], [Return] ).
Press [demodulation setup].
Pre ss [result length], 230, [sym]
Press [ref filter], [raised cosine]
Press [Return], [meas filter], [root raised cosine].

5. View the created signal versus the reference:

Press [Display], [2 grids],
Press [A], [Measurement Data], [IQ measured time]
Press [B], [Measurement Data], [IQ reference time]

11 - 4
 Creating User-defined Signals (Options AYA and AYH)

6. Format both traces simultaneously:

Press [Shift], [A], [Data Format], [polar IQ constellation]
Press [more format setup], [symbol dots ].
Press [Auto Scale]

Be careful when selecting a reference filter if the demodulation format uses

distributed filtering. The demodulation format used in this example (32
QAM) uses distributed filtering. Therefore, a root-raised-cosine reference
filter was needed to create the ideal signal (to represent filtering at the
transmitter), whereas a raised-cosine reference filter was needed to
demodulate the ideal signal (to represent the total filtering in the system).

Truncate the analysis time record at the beginning (with trigger delay of
about 5 symbols) and at the end (with a result length about 10 symbols
shorter) to eliminate invalid data caused by discontinuities between the
beginning and end of the waveform. Use this truncated signal as a
stimulus for component or system tests.

This signal-versus-reference check is also a good opportunity to see the

effects of changing the symbol rate or filtering in order to demonstrate the
effects of setup parameters which are incompatible with the incoming
signal.

11 - 5
 Creating User-defined Signals (Options AYA and AYH)

To create a user-defined filter

This task shows you how to create filters that you can use as the
measured or reference filter. This task uses digital demodulation. If you
want to perform this task using video demodulation, choose video
demodulation instead of digital demodulation in step 4.

1. Create a file defining your desired filter:

You may use common software packages such at MATLAB or Mathcad to
define the filter shape. For digital demodulation, you must use 20 points
per symbol and you may use a maximum of 20 symbols. For video
demodulation, you must use 40 points per symbol and you may use a
maximum of 20 symbols. A good choice for a total number of defined
points is 401 for digital demodulation and 801 for video demodulation (this
allows you to define a center point symbol in order to achieve symmetry)

2. Convert the file to SDF using the Standard Format Data Utilities supplied with
the analyzer:
As an example, use a PC to convert an ASCII file to SDF:
asctosdf <source file> <destination file>
The destination disk must be compatible with your analyzer’s disk drive
(DS,HD).

3. View the filter’s impulse response:

Press [Save/Recall], [recall trace], [from file into Dx].
Press [Measurement Data], [data reg]; select the data register that has your
trace.

4. Apply the filter to your signal:

Press [Instrument Mode], [Digital Demodulation] (with option AYH, press
[Instrument Mode], [demod type], [Digital Demodulation], [Return] ).
Press [demodulation setup].
Pre ss [meas filter] or [ref filter], [user defined].
Select a user data register Dx.
The Signals Disk contains an example of a user defined filter. The trace
version of the filter is stored as “GAUSS1.DAT” and the ASCII version is
stored as “GAUSS1.ASC” It is a Gaussian filter 6 symbols (121 pts) wide,
with 20 points per symbol, and a BT of 1.0. You might use this type of
filter on an MSK signal.

The documentation for the Standard Data Format Utilities which

accompanies this analyzer also includes a section (Mathcad Examples)
describing how to create waveforms and filters for use with this analyzer.

11 - 6
 Creating User-defined Signals (Options AYA and AYH)

The trace display of “GAUSS1.DAT”.

11 - 7
12

Using Adaptive
Equalization (Options
AYA and AYH)
This section shows you how to use Adaptive Equalization. Adaptive
equalization removes linear errors from modulated signals by dynamically
creating and applying a compensating filter. Adaptive equalization is only
available in Digital and Video Demodulation instrument modes.

12-1
 Using Adaptive Equalization (Options AYA and AYH)

To determine if your analyzer has Adaptive Equalization

To use Adaptive Equalization, your analyzer must have the options and
hardware shown below. The following steps show you how to determine if
your analyzer has these options and hardware.
l Options AYA (Vector Modulation Analysis) OR option AYH (Digital Video
Modulation Analysis).
l A42 Memory assembly greater than Revision A.

1. Display the OPTIONS CONFIGURATION table.

Press [System Utility], [options setup].
2. Check that option AYA or option AYH is installed in your analyzer. The
analyzer displays YES under INSTALLED if an option is installed.
3. Check that your A42 Memory assembly’s revision is greater than Rev A.
Set the power switch to off (O).
Set the power switch to on ( l ) to run the power-up tests.
Press [System Utility], [more], [diagnostics], [test log on].

The test log contains the results of the power-up tests. The power-up
tests report the revision of the A42 Memory assembly. Make sure your
analyzer has a revision greater than Rev A.
4. If your analyzer does not have all of the above options and hardware, you must
purchase the options or hardware that you are missing. To do this, contact
your Agilent Technologies sales representative or your local Agilent
Technologies Sales and Service office (listed on the inside, rear cover of the
Operator’s Guide).

The power-up tests report the

revision of your A42 Memory
assembly. This analyzer does
NOT need a new A42 Memory
Assembly because the revision
(Rev) is greater than Rev A.
Test log

12-2
 Using Adaptive Equalization (Options AYA and AYH)

To load the multi-path signal from the Signals Disk

This task shows you how to load a 16-QAM, multi-path, time-capture signal
from the Signals Disk. Other tasks in this section use this signal to teach
you how to use adaptive equalization.

1. Initialize the analyzer and select the Digital Demodulation instrument mode:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [demod type], [Digital].

2. If your analyzer has the optional, second input-channel installed, turn it off:
Press [Input], [channel 2], [ch2 state off].

3. Load the time-capture data into the capture buffer:

Insert the Signals Disk in the analyzer’s disk drive.
Press [Save/Recall], [default disk], [internal disk] to select the internal disk drive.
Press [Return] (bottom softkey), [catalog on]to display the files on the disk.
Rotate the knob until the file EQSIGNAL.DAT is highlighted.
Press [recall more], [recall capture buffer], [enter] (takes about 2 minutes).

Using time-captured data causes the analyzer to automatically measure

from the capture-buffer instead of the input channel and sets the center
frequency, span, and resolution bandwidth to those used when the data
was captured. In a typical equalization measurement you would set these
parameters instead of loading captured data.

12-3
 Using Adaptive Equalization (Options AYA and AYH)

To demodulate the multi-path signal

This task shows you how to demodulate the multi-path signal that you
loaded in the previous task.

1. Load the multi-path signal as instructed in the previous task.

2. Set demodulation parameters for this signal:

Press [Instrument Mode], [demodulation setup].
Press [symbol rate], 5 MHz.
Press [alpha/BT], 0.15 [enter].
Press [meas filter], [root raised cosine], [return].
Press [ref filter], [raised cosine], [return].
Press [result length], 500 [sym].
Press [demod format], [QAM 16], [return].
Press [Time], [points/symbol], 1, [enter].
3. Configure different displays for the demodulated data:
Press [Display], [4 grids quad].
Press [A] , [Measurement Data], [more choices], [channel frequency resp].
Press [B], [P10equalizer impulse resp], [Data Format], [magnitude log(dB)].
P ress [C], [Mea sure ment Data], [symbol table/error summary]
Press [D] , [Measurement Data], [IQ measured time], [Data Format],
[polar (IQ) constellation], [more format setup], [symbol dots].

4. Start the measurement:

Press [Meas Restart].

Traces A and B display the frequency response and the impulse response
of the equalization filter. You can view these displays even when you are
not using the equalization filter. In this example equalization is turned off,
therefore the equalization filter does not change and these displays remain
constant.

Traces C and D display the symbol table and constellation diagram for the
multi-path signal. The signal contains a significant amount of distortion
which makes it difficult to demodulate. The next task uses equalization to
compensate for distortion in the signal which significantly improves these
displays.

By default, the equalization filter is defined to have a unit impulse response

which yields a flat frequency response.

12-4
 Using Adaptive Equalization (Options AYA and AYH)

In this example, traces A and B

show the default frequency
response and impulse response of
the equalization filter. By
default, the equalization filter has
a unit impulse response.

This signal is difficult to

demodulate due to linear
distortion. The next task uses
equalization to compensate for
the linear distortion.

Demodulated signal and equalization filter displays

12-5
 Using Adaptive Equalization (Options AYA and AYH)

To apply adaptive equalization

This task shows you how to apply adaptive equalization to the multi-path
signal that you demodulated in the previous task.

1. Perform the previous task.

2. Display the equalization-filter menu:
Press [Instrument Mode], [demodulation setup], [more].
3. Set equalization parameters for this measurement:
Press [eq filt len], 41 [sym].
Press the up or down arrow key until the convergence is 2 e – 6.
4. Configure the equalization filter to update with each measurement:
Press [eq adapt run].
5. Reset the equalization filter:
Press [eq reset].
6. Enable the equalization filter:
Press [eq filter on].
7. Restart the measurement:
Press [Meas Restart].
8. Autoscale traces A and B (the frequency response and impulse response of the
equalization filter) as the analyzer shapes the equalization-filter.
Press [A]
Press [Shift], [B]
Press [Auto Scale].

This example lets you watch as the analyzer shapes the equalization filter.
The analyzer estimates new filter coefficients with each measurement, and
then uses the new coefficients to adapt the filter for the next measurement.

By default, the equalization filter has a unit impulse response when the
analyzer is first turned on, if you press [Preset] or [eq reset], or if you change
instrument modes or [points/symbol]. Aside from these conditions, the analyzer
uses the last computed coefficients when you enable equalization. For
example, if you used equalization in a previous measurement, the analyzer
uses the coefficients from the previous measurement unless you press
[Preset] or [eq reset], or change instrument modes or [points/symbol]. Therefore, it
is good practice to press [eq reset] to reset the filter coefficients before you
start a measurement.

12-6
 Using Adaptive Equalization (Options AYA and AYH)

The [convergence] determines how quickly the old and new filter-coefficients
converge. Larger values converge faster. Values that are too large can
cause the adaptation algorithm to become unstable or fluctuate from stable
to unstable. Filter length, points-per-symbol, modulation format, and result
length interact to determine the best value for convergence. Good results
are normally achieved using values between 10-7 and 10-6 .

HINT At the start of your measurement, set the convergence high to quickly shape
the filter. Then decrease the convergence to fine-tune the filter to the optimum
shape.

The equalization filter length ([eq filt len]) affects the number of taps in the
equalization filter. For multi-path environments, longer filter lengths are
needed to estimate good filter coefficients.

The following parameters affect measurement speed when using adaptive

equalization.

l [result length]
l [eq filt len]
l [points/symbol]
For additional details, see online help for the [eq filter on/off] softkey.

In this example, trace A shows the

frequency response and trace B
shows the impulse response of the
equalization filter. With [eq adapt
run] selected, the traces change
with each measurement as the
analyzer updates the filter
coefficients.

Applying equalization to the measurement

12-7
 Using Adaptive Equalization (Options AYA and AYH)

To measure signal paths

This task shows you how to use the equalization filter’s impulse response
to identify and measure paths in a multi-path signal. This task uses the
multi-path signal on the Signals Disk and is a continuation of the previous
task.

1. Perform the previous task.

2. Configure the display to show the impulse response of the equalization filter in
a single grid:
Press [B], [Display], [single grid].

3. Change the x-axis units to seconds:

Press [Ref Lvl/Scale], [X & Y units setup], [X units], [s].

4. Display bars at the symbol locations:

Press [Data Format], [more format setup], [symbol bars].

5. Move the marker to the peak impulse (this is the main signal path):
Press [Shift], [Marker].

The marker readout shows the main impulse at 0 seconds with

approximately 0 dB of loss.

6. Move the marker to the next peak (this is the second signal path):
Press [Marker Search], [next peak].

The marker readout shows the second signal path at − 800 ns with
approximately 15 dB of loss relative to the main impulse (the strongest
path).

7. Move the marker to the next peak (this is the third signal path):
Press [next peak] again.

The marker readout shows the third signal path at 3 µ s with

approximately 20 dB of loss.

The impulse response in this example shows several peaks. The three
highest peaks correspond to the main signal path plus two multi-path
signals. The remaining peaks correspond to the two multi-path signals, as
described in the next paragraph.

12-8
 Using Adaptive Equalization (Options AYA and AYH)

Each point in the impulse-response display corresponds to a tap in the

feed-forward equalizer (FFE). In a FFE, large coefficients that are
separate from the main tap correspond directly to alternate signal paths.
The smaller peaks are a result of the same alternate signal paths that
created the large peaks. In other words, a signal with one strong alternate
path will have more than two impulses on the display (the main impulse
and the impulses due to the alternate path). The additional impulses will
be lower than the two main impulses.

Path 1 Path 2 Path 3

In this example, the signal has

3 paths. The strongest path
is not the shortest path. The
shortest path passed through
a building causing it to be
attenuated. Normally, the
strongest path will be the
shortest, or direct path.
Using the impulse-response display to measure multi-path signals.

12-9
Using Adaptive Equalization (Options AYA and AYH)

To learn more about equalization

Adaptive equalization is a powerful feature that you can use in many

applications. The following paragraphs include additional information that
may help you use adaptive equalization for your application.
l Equalization is available only for Digital and Video Demodulation instrument
modes. Equalization is not available for EDGE measurements when the
EDGE(winRC) measured filter is selected. To use equalization with EDGE
measurements, set the measured filter to OFF.

l The primary application of the equalizer’s impulse-response display is for

evaluating multi-path environments. Multi-path environments usually require
longer filter lengths.

l The primary application of the equalizer’s frequency-response display is for

evaluating the transmitter or receiver signal-path for errors such as passband
ripple and group-delay distortion. Short filter lengths usually work well for
these types of measurements.

l By default, the equalization filter has a unit impulse response (only one tap in
the filter has a non-zero value and data simply passes through the filter). The
position of the unit impulse is a function of the filter length and is positioned to
provide the most optimum efficiency for most situations. The position cannot
be adjusted.

l The filter length and points/symbol determine the number of taps in the
equalization filter, as follows:

# Taps = (( filter_length − 1 ) times points_per_symbol )) + 1

Press [Instrument Mode], [demodulation setup], [more], [eq filt len] to set the filter length.
Press [Time], [points/symbol] to set the points/symbol.

l Generally, there is no advantage to using more than 2 points/symbol when using

equalization. You may want to use more than 2 points/symbol for better
resolution of such displays as eye diagrams, but the tradeoff is slower
measurement speed.

l To see the channel frequency-response over the entire bandwidth of your signal,
use 2 points/symbol or greater. You cannot see the channel frequency-response
over the entire bandwidth of your signal if you use 1 point/symbol.

l Online help for the equalization-filter softkeys includes additional information.

Select Digital or Video demodulation, then press [Help], [Instrument Mode],
[demodulation setup], [more] and any equalization softkey.

12-10
13

Using Wideband CDMA

(Options B73, B79, and
080)
This chapter shows you how to make Wideband CDMA measurements. For
a conceptual overview of Wideband CDMA, see the chapter titled
Wideband CDMA Concepts.

13 - 1
 Using Wideband CDMA (Options B73, B79, and 080)

To view a W-CDMA signal

All tasks in this chapter were created using a 3GPP 1999 forward-link
signal (option 080). The signal was generated with an HP/Agilent E4433
ESG Signal Generator at 5 MHz, –10 dBm, and using the channel
definitions shown below. If you are using a different W-CDMA signal
(i.e., trial 1998 forward-link signal (option B73) or ARIB 1.0-1.2 forward-link
signal (option B79)), the tasks in this chapter still apply . However, in the
next task (To demodulate a W-CDMA signal), make sure you set the
demodulation parameters to match your signal.

Note Older E4433 ESG Signal Generator’s may flip the frequency spectrum at 5 MHz.
If necessary, correct for this when you demodulate the signal in the next task
(To demodulate a W-CDMA signal) by pressing [Instrument Mode], [demodulation setup],
[freq spectrum mirror].

Channel Rate (kSym/s) Spread Code Power

PSCH N/A N/A – 15.01
SSCH N/A N/A – 15.01
CPICH 15 0 – 20.01
DPCH 30 10 – 16.01
DPCH 30 11 – 16.01
DPCH 30 12 – 16.01
DPCH 120 8 – 9.01
DPCH 120 9 – 10.01
DPCH 120 10 – 11.01
DPCH 60 32 – 12.01
DPCH 60 33 – 12.01
DPCH 60 34 – 12.01
DPCH 60 35 – 12.01
DPCH 480 6 – 4.45

The following steps show you how to view your W-CDMA signal in the
Vector instrument mode to verify that the signal is present and that the
analyzer’s center frequency, span, and input range are set correctly.

13 - 2
 Using Wideband CDMA (Options B73, B79, and 080)

1. Connect your signal to the Channel 1 input.

2. Initialize the analyzer and select the Vector instrument mode:
Press [Instrument Mode], [receiver], then press:
89410A: [input section (0-10 MHz)].
89441A: [RF section (0-10 MHz)].
Press [Preset].
Press [Instrument Mode], [Vector]
3. Select a center frequency and span:
Press [Frequency], [center], 5, [MHz]
Press [span], 5, [MHz].
To lock onto your signal, the analyzer’s center frequency must be set to
the center frequency of your signal. To make accurate measurements, the
frequency span must be set to include your entire signal. This signal
requires a center frequency and span of 5 MHz.
4. Select the optimum range:
Press [Range].
Press the down-arrow key until the Channel-1 Over and Half LEDs are on.
Press the up arrow key one press at a time until the Over LED turns off.
For additional details about selecting the optimum range, see online help
for the [Range], [ch1 range] softkey.
5. Start the measurement:
Press [Measurement Restart]

Spectrum of the
W-CDMA signal.
Averaging was turned on
to smooth the trace.

Wideband CDMA Signal

13 - 3
 Using Wideband CDMA (Options B73, B79, and 080)

To demodulate a W-CDMA signal

This task shows you how to demodulate the W-CDMA signal that you
loaded in the previous task. If you are using your own W-CDMA signal
instead of the one from the previous task, you may need to change some
of the parameters set in step 5 to match those of your signal.
1. Load the W-CDMA signal as instructed in the previous task.
2. Turn on Wideband CDMA demodulation.
Press [Instrument Mode], [demod type], [Wideband CDMA].
3. Set the maximum W-CDMA span to 5 MHz.
Press [System Utility], [memory usage], [configure meas memory]
Press [mx WCDMA span], 5 MHz.
4. Verify that the maximum number of time points is 4096:
Press [max time pts], 4096 [enter].
If necessary, reallocate memory until you can set [max time pts] to 4096 points.
Smaller values limit the number of symbols that the analyzer can
demodulate and display.
5. Set demodulation parameters for this signal:
Press [Instrument Mode], [demodulation setup].
Press [demod format], [3GPP forward link], [W-CDMA 5 MHz], [return], [return].

Selecting [3GPP forward link] configures the analyzer to make base-station,

W-CDMA measurements, as proposed in the 3GPP 1999 forward link
standard. Pressing the [W-CDMA 5 MHz] standard-setup softkey automatically
sets the chip rate (3.84 MHz), main length (15 slots), scramble code (0),
and filter alpha (0.22) to match that of a 5 MHz, 3GPP 1999 forward-link
signal. For details about standard setups softkeys, see online help for the
[3GPP forward link] softkey.
6. If necessary, mirror the frequency spectrum.
Press [Instrument Mode], [demodulation setup], [freq spectrum mirror]

As mentioned in the previous task, some W-CDMA signals may have a

flipped (mirrored) frequency spectrum. If this is true for your signal, you
must perform this step.
7. Select the composite code-domain power display:
Press [Measurement Data], [code domain], [composite].
8. Start the measurement:
Press [Pause/Single].
9. After the measurement finishes, autoscale the results:
Press [Auto Scale].

13 - 4
 Using Wideband CDMA (Options B73, B79, and 080)

One important parameter set by this procedure is the maximum W-CDMA

span. This parameter allocates memory for W-CDMA measurements. Since
W-CDMA measurements require large amounts of memory, set this
parameter to the smallest frequency span that you will measure.

If the analyzer is unable to lock to your signal, verify that you are using
the correct chip rate, scramble code, and center frequency. Also, verify
that [freq spectrum mirror] is selected if the spectrum of your signal is flipped
(mirrored). The analyzer’s chip rate and scramble code must match that of
your signal. The analyzer’s center frequency must be within 500 Hz. of
your signal’s center frequency.

By default, the analyzer displays the composite code-domain power display,

which shows all layers simultaneously. So you can differentiate between
active layers, the analyzer uses a different color for each code layer. In
some measurements, you may have to use x-scale markers to see the color.
This is because individual channels at the slower layers are represented by
a single line in the code-domain power display.

Tip Code-domain power is relative to the total signal power in the code domain. To
display absolute power, press [Instrument Mode], [demodulation setup], [normalize off].

Each code layer uses a different

color. The marker in this
illustration is on channel 8 in
code layer 5 (120 ksym/s).

X-axis annotation is based on the

slowest code layer. For 3GPP
1999 forward-link signals, the
slowest code layer is code layer 9,
also known as code layer 7.5
ksym/s.

Composite Code-Domain Power Display

13 - 5
 Using Wideband CDMA (Options B73, B79, and 080)

To view data for a single code layer

This task builds on the previous task to show you how to view
code-domain power for a single code layer. Single code-layer displays are
useful if the composite display does not accurately identify which layer a
channel resides in. This can happen under these conditions:
l A time slot contains power control.
l A time slot contains excessive noise.
l Spreading codes are not allocated correctly (layers are transmitted on top of one
another).

1. Perform the previous task.

2. Display two grids:

Press [Display], [2 grids].

3. Display code-domain power for the 60 ksym/s layer in trace B:

Press [B].
Press [Measurement Data], [code domain], [60 ksym/s].

You can display code-domain power for all layers simultaneously (using
composite code-domain power) or you can display code-domain power for
a single code layer.

Like composite code-domain power, the single code-layer display shows all
active channels. However, the only channels that are colored are those in
the selected layer (using the same color as that used in the composite
display). Active channels in other code layers are not colored.

This example shows four active channels in the 60 ksym/s code layer:
channels 32 through 35. These channels are colored. Other channels, such
as channels 16 through 21 also contain power; however, these channels are
not colored since the power is not in the 60 ksym/s code layer.

13 - 6
 Using Wideband CDMA (Options B73, B79, and 080)

Trace B shows code-domain power for a single

code layer: code layer C6 (60 ksym/s). This
code layer has 64 code channels (codes 0 to 63
on the x-axis). Channels that are active in this
code layer are colored. Channels that are active
but not in this code layer are not colored.

Composite and single code-domain power displays

13 - 7
 Using Wideband CDMA (Options B73, B79, and 080)

To view data for a single code channel

This task builds on the previous task to show you how to view the vector
diagram for individual channels. The power varies in each channel. The
task turns normalization off so you can see the vector diagram change size
when you change channels.

1. Perform the previous task.

2. Display the vector diagram in trace B:
Press [B].
Press [Measurement Data], [time domain]., [IQ measured].
Press [Data Format], [polar (IQ) vector].
3. Turn normalization off to compare the absolute power between different
channels:
Press [Instrument Mode], [demodulation setup], [normalize off].
4. Select the first 120 ksym/s channel in the composite code-domain display:
Press [A].
Move the marker until the marker readout shows Chan: 120 ksym/s C5(8).
5. Display the vector diagram for the selected channel:
Press [Marker →], [mkr→ layer/channel].
6. Autoscale the results:
Press [Auto Scale].
7. Display the vector diagram for the second 120 ksym/s channel:
Move the marker until the marker readout shows Chan: 120 ksym/s C5(9).
Press [Marker →], [mkr → layer/channel].
8. Using a different method, display the constellation diagram for the third
120 ksym/s channel:
Press [Time], [code channel], 10, [code].

All time-domain measurement data (except composite time domain) and

the symbol table show results for a single channel in a single code layer
(you use the [Measurement Data] hardkey to select time-domain measurement
data and the symbol table). By default, the analyzer uses channel 0 in
slowest code layer. This procedure shows you two ways to select a
different channel and layer—using the [mkr → layer/channel] softkey or the [code
channel] and [code layer] softkeys.

13 - 8
 Using Wideband CDMA (Options B73, B79, and 080)

The beginning steps show you how to use the [mkr → layer/channel] softkey
(under [Marker →]). The final step shows you how to use the [code channel] and
[code layer] softkeys (under [Time]). In fact, the [mkr → layer/channel] softkey is
simply a shortcut that sets [code channel] and [code layer] to the current marker
value. For this shortcut to work, the active trace must be a code-domain
power display.

Notice in this procedure that the vector diagram for the second and third
channels decreases in size. This occurs because power is absolute
(normalization is off) and the power decreases in each channel.

Also notice that there are 16 marker positions in the composite

code-domain power display for each channel in the 120 ksym/s code layer.
This occurs because marker readouts are based on the slowest code layer,
in this case the 7.5 ksym/s layer, and there are sixteen 7.5 ksym/s channels
per 120 ksym/s channel.

Position the marker on the channel of

interest, then press [Marker →],
[mkr → layer/channel] to display the vector
diagram for that channel in the lower
trace. You can use this procedure with
any time-domain display in the lower
trace, including the symbol table/error
summary display.

Diagram for Code Channel C5(8)

13 - 9
 Using Wideband CDMA (Options B73, B79, and 080)

To view data for one or more slots

This task builds on the previous task to show you how to use time gating.
Time gating lets you view measurement data for selected slots.

1. Perform the previous task.

2. Display the IQ magnitude error:

Press [B].
Press [Measurement Data], [time domain], [IQ error mag].

3. Autoscale the results:

Press [Auto Scale].

4. Display the IQ magnitude error for the first slot:

Press [Time], [gate length], 1, [slot].
Press [gate delay], 0, [slot].

Changing [gate length] or [gate delay] automatically selects [gate on] to enable time
gating.

5. Display the IQ magnitude error for the second slot:

Press [gate delay]
Press the up arrow key (to increment gate delay to 1 slot).

6. Display the IQ magnitude error for the second slot in the previous code channel:
Press [code channel].
Press the down arrow key (to decrement code channel to code 9).

This task introduces time gating, which is an advanced feature that lets you
select one or more slots for analysis. With time gating off, the
measurement includes all slots (as set by [Time], [main length]).

With time gating on, the measurement includes the slots selected by [gate
length] and [gate delay]. Gate length selects the number of slots; gate delay
determines which slot in the measurement to use as the starting slot. For
example, a gate delay of zero selects the first slot in the measurement as
the first slot in the gated results; a gate delay of one selects the second
slot; a gate delay of two selects the third slot, and so forth.

13 - 10
 Using Wideband CDMA (Options B73, B79, and 080)

Tip 1 You do not have to start a new measurement when you use time gating. You
can change the gate length and gate delay to see different gated results on the
same measurement data.

Time gating provides a convenient way

to view results for one or more slots.
As in the previous task, the results are
displayed for a single code channel in a
single code layer, as determined by the
[Time], [code channel] and [code layer]
softkeys.

IQ magnitude error for C5 (9), slot two (3.84 MHz chip rate)

13 - 11
 Using Wideband CDMA (Options B73, B79, and 080)

To view the symbol table and error parameters

This task builds on the previous task to show you how to use the symbol
table.

1. Perform the previous task.

2. Display the symbol table for the gated results in trace A:

Press [A].
Press [Measurement Data], [symbol table/error summary].

3. Position the marker on any symbol—in this case symbol 17:

Press [Marker/Entry] until the Marker LED is highlighted.
Rotate the knob until the marker is on symbol 17.

The symbol table shows the demodulated bits, error parameters, and slot
and timing information for the selected layer and channel (as you learned
in To view data for a single code channel, the [code layer] softkey determines
the layer; the [code channel] softkey determines the channel).

The previous task turned on time gating to include only one slot in the
measurement. In the symbol table, Slot shows the beginning slot in the
measurement.

Tip For details about the symbol table and error summary results, see online help
for the [symbol table/error summary] softkey.

13 - 12
 Using Wideband CDMA (Options B73, B79, and 080)

The symbol table and error summary

information for slot 2, code layer 5,
channel 9.

Symbol Table/Error Summary results for C5 (9), slot two

13 - 13
 Using Wideband CDMA (Options B73, B79, and 080)

To use x-scale markers on code-domain power displays

This task builds on the previous task to show you how to use x-scale
markers to ‘’zoom’’ in on channels in a code-domain power display.

1. Demodulate the W-CDMA signal as shown earlier in this chapter.

2. Display a single grid:
Press [Display], [single grid].
3. Display code-domain power for the 15 ksym/s layer in trace A:
Press [A].
Press [Measurement Data], [code domain], [15 ksym/s].
4. Autoscale the results:
Press [Auto Scale].
5. Using x-scale markers, select channels 0 to 25:
Press [Ref Lvl/Scale], [X scale markers], [left ref], 0 [code].
Press [right ref], 25 [code].
Press [scale at markers].
6. Turn off x-scale markers:
Press [Ref Lvl/Scale], [X scale markers], [X full scale].

Using x-scale markers, you

can select which portion of
the x-axis you want to view.
This feature lets you zoom
in on selected channels.

Using x-scale markers to ‘’zoom’’ in on composite code-domain power

13 - 14
14

Using the LAN

(Options UTH & UG7)
The tasks in this section show you how to configure and use the analyzer’s
optional LAN interface. The LAN interface is present only in analyzers that
have option UTH. X-Windows operation and FTP (File Transfer Protocol)
are available only in analyzers that have option UG7 (Advance LAN).

14 - 1
 Using the LAN (Options UTH & UG7)

To determine if you have options UTH and UG7

1. Turn on the analyzer.

2. Press [Local/Setup]. If softkey F5 is [LAN setup], you have option UTH. If the
[LAN setup] softkey does not exist, stop here-you do not have option UTH, nor do
you have option UG7.
3. Press [LAN setup]. If softkey F4 is [X11 display on/off], you have option UG7. If the
[X11 display on/off] softkey does not exist, you do not have option UG7.

The LAN interface is present only in analyzers that have option UTH. This
option lets you use telnet or C-programs to send GPIB commands to the
analyzer via the LAN.

Note Option UTH replaced option UFG. The only differenced between option UFG
and UTH is the amount of RAM. Option UFG had 4 MB of RAM, option UTH
has 20 MB of RAM. If your analyzer has option UFG, your analyzer does have
the LAN option.

Option UTH consists of a single printed circuit board (card) that contains
20 MegaBytes of memory, a LAN interface, and an additional GPIB port.
The LAN interface provides Ethernet (IEEE 802.3) LAN compatibility and
has two LAN ports: a ThinLAN BNC and a 15-pin AUI (MAU) connector.
The additional GPIB port is a controller-only port that communicates with
external GPIB devices, and provides a simple way to program external
receivers (such as downconverters) without tying up the primary GPIB port
or system controller. For details on using external receivers, see “Using
the Agilent 89411A Downconverter.”

Option UG7 enhances option UTH. Option UG7 provides remote

X-Windows capabilities, which lets you view the analyzer’s display and
control the analyzer from across the building or across the world. Option
UG7 also includes FTP (File Transfer Protocol) software. You can use
FTP to transfer data to and from the analyzer.

To order options, contact your local Agilent Technologies Sales and Service
Office.

14 - 2
 Using the LAN (Options UTH & UG7)

To connect the analyzer to a network

1. Turn off the analyzer.

2. If your network uses ThinLAN BNC cables, connect one of them to the
ThinLAN connector on the analyzer’s rear panel.
or
If your network uses MAUs, connect one of them to the AUI Port connector on
the analyzer’s rear panel.
3. Turn on the analyzer.
4. Press [Local/Setup] [LAN setup], [LAN port setup].
5. Press [port select] to display the option corresponding to the connector you used
in step 2: either “ThinLAN (BNC)” or “AUI (MAU).”

The ThinLAN connector only allows you to connect the analyzer to a

ThinLAN network. However, the AUI Port lets you to connect the analyzer
to ThinLan, ThickLAN, or StarLAN 10 networks via the appropriate
off-board MAU. (These networks are Agilent Technologies’s
implementation of IEEE 802.3 types 10BASE2, 10BASE5, and 10BASE-T.)

The analyzer should be connected to a network by only one of its LAN

connectors. Check with your network administrator if you have any other
questions about the LAN connections.

14 - 3
 Using the LAN (Options UTH & UG7)

To set the analyzer’s network address

1. Ask your network administrator to assign an Internet Protocol (IP) address to

your analyzer. Write down the address for use in step 4.
2. If your analyzer must communicate with computers outside of the local subnet,
ask your network administrator for the IP address and subnet mask required to
route data through the local gateway. Write down these values for use in steps
5 and 6.
3. On the analyzer, press [Local/Setup], [LAN setup], [LAN port setup].
4. Press [IP address], type the address obtained in step 1, then press [enter].
5. If you obtained a gateway address in step 2, press [gateway IP], type the address,
then press [enter].
6. If you obtained a mask value in step 2, press [subnet mask], type the value, then
press [enter].
7. Turn off the analyzer, then turn it back on to make the new settings permanent.

You must enter the addresses and the subnet mask using dotted decimal
notation (for example, 13.1.212.66). You can disable gateway routing by
setting [gateway IP] or [subnet mask] to 0.0.0.0.

14 - 4
 Using the LAN (Options UTH & UG7)

To activate the analyzer’s network interface

1. Press [Local/Setup], [LAN setup], then press [LAN power-on] to display “active.”
2. Turn off the analyzer, then turn it back on to make the new setting permanent.
When you are not using the network interface, you should press
[LAN power-on] to display “inactive.” This will free additional memory for
other uses.

14 - 5
 Using the LAN (Options UTH & UG7)

To send GPIB commands to the analyzer

1. Confirm that the first four tasks in this chapter have been completed.
2. If you do not know the network address of your analyzer, press [Local/Setup],
[LAN setup], [LAN port setup], then write down the value displayed under [IP address].
3. On the computer, type: telnet <IP_address>
(where <IP_address> is the network address of your analyzer).
4. On the computer, type the GPIB command that you want to send to the
analyzer. For example, to query the analyzer for its center frequency, type:
FREQ:CENTER?
5. To end your telnet session, type <Ctrl><D>.

The computer you use to send GPIB commands to the analyzer must be
attached to the network and configured with software that supports the
TELNET protocol. For additional information about using telnet, refer to
the documentation that came with your TELNET software.

Telnet is available only in analyzers that have option UTH. To determine if

your analyzer has this option, see this task: “To determine if you have
options UTH and UG7.”

14 - 6
 Using the LAN (Options UTH & UG7)

To select the remote X-Windows server

1. Determine the IP address of the computer you will use for remote X-Windows
operation. (Ask your network administrator for help if you don’t know how to
do this.) Write down the address for use in step 3.
2. Press [Local/Setup], [LAN setup].
3. Press [X11 IP address], type the address obtained in step 1, then press [enter].

After you have attached the analyzer to the network and configured it as
described in the previous two tasks, you can operate it remotely from any
computer that is attached to the network and running X-Windows. This
task shows you how to select the computer you want to use for remote
operation. The next task shows you how to initiate remote operation.

Remote X-Windows is available only in analyzers that have options UTH

and UG7. To determine if your analyzer has these options, see this task:
“To determine if you have options UTH and UG7.”

14 - 7
 Using the LAN (Options UTH & UG7)

To initiate remote X-Windows operation

1. Confirm that the previous six tasks have been completed.

2. On the remote computer (selected in the previous task), position the mouse
pointer in one of the windows, then enter the following command: xhost +
3. On the analyzer, press [Local/Setup], [LAN setup].
4. Press [X11 display] to display “on.”
5. On the remote computer, use the mouse to position the outline of the remote
X11 display, then click the left mouse button to continue.

When you complete this task, the outline of the remote X11 display is filled
in with a replica of the analyzer’s front panel. The computer maintains this
replica by using its LAN connection to get the latest trace, and state
information, from the analyzer.

You may find that the computer responds more slowly to other processes
(for example, key presses and mouse movements) while it is maintaining
the X11 display. If it responds too slowly, you can decrease the value of
[rate limit], which is located under [Local/Setup], [LAN setup]. This allows the
analyzer to respond more quickly to other processes by reducing the
amount of time it spends maintaining the X11 display.

Remote X-Windows is available only in analyzers that have options UTH

and UG7. To determine if your analyzer has these options, see this task:
“To determine if you have options UTH and UG7.”

14 - 8
 Using the LAN (Options UTH & UG7)

To use the remote X-Windows display

1. Confirm that the previous task has been completed.

2. Use the instructions in the following paragraphs to control the analyzer from
the remote X-Windows display.

l To press a key. Place the cursor on the key, then click the left mouse button.
l To activate shifted key functions. Place the cursor on the [Shift] key, then click
the left mouse button. (Text is now displayed in blue on keys with shifted
functions.)
l To modify parameters. Click on the key that activates the parameter you want
to modify, use the computer’s keyboard to type the new text or number, then
click on [enter] (or the appropriate units key.)
l To turn the knob. Place the cursor on the knob, then click the right or left
mouse button to turn it; right turns it clockwise, left turns it counter-clockwise.
l To position the marker. Place the cursor on or near the trace at the desired
x-axis location, then click the left mouse button.

You may want to pause the analyzer before changing its configuration via
the X-Windows display. The analyzer can respond more quickly to these
changes when it is paused. Click on [Pause|Single] to pause the analyzer.

Remote X-Windows is available only in analyzers that have options UTH

and UG7. To determine if your analyzer has these options, see this task:
“To determine if you have options UTH and UG7.”

14 - 9
 Using the LAN (Options UTH & UG7)

To transfer files via the network

1. Confirm that the first four tasks in this chapter have been completed.
2. If you do not know the network address of your analyzer, press [Local/Setup],
[LAN setup], [LAN port setup], then write down the value displayed under [IP address].
3. On the computer, type: ftp <IP_address>
(where <IP_address> is the network address of your analyzer).
4. On the computer, just press <Enter> (or <Return>) when you are prompted
for a name and password.
5. If you want to list the files in the root directory, type: ls
6. Change to the directory where you want the file transfer to occur by typing:
cd <directory>.
7. If the file is an ASCII file, set FTP to ASCII by typing: ascii. If the file is a binary
file, set FTP to binary by typing: binary.
8. If you want to transfer a file from the analyzer to the computer, type:
get <filename>
9. If you want to transfer a file from the computer to the analyzer, type :
put <filename>
10. To exit FTP, type <quit>

The computer you use to transfer files must be attached to the network
and configured with software that implements the following networking
application: TCP/IP’s File Transfer Protocol (FTP). Refer to the
documentation supplied with that software for additional information about
using FTP to transfer files.

You cannot transfer LIF files from the computer to the analyzer. LIF files
can only be transferred from the analyzer to the computer. If you transfer
a LIF file from the computer to the analyzer, the LIF file will be corrupted.

For information about the analyzer’s directory structure, see the “FTP (File
Transfer Protocol)” topic in online help. (Press [Help] [1] to select the
online help index, use the knob or the arrow keys to highlight “FTP (File
Transfer Protocol),” and press [4]).

FTP is available only in analyzers that have options UTH and UG7. To
determine if your analyzer has these options, see this task: “To determine if
you have options UTH and UG7.”

14 - 10
15

Using the Agilent

89411A Downconverter
The 89411A allows the use of the advanced analysis features of the 89410A
(the IF Section) to be applied to signals above the frequency limit of the
89441A.

15-1
The Agilent 89411A at a Glance

Agilent 89411A front panel

Agilent 89411 rear panel

Agilent 89411A block diagram

15-2
Descriptions

The 89411A is a fixed downconverter used to translate the 21.4 MHz IF

output on several Agilent Technologies RF/microwave spectrum analyzers
to a baseband frequency within the range of the 89410A. It translates the
entire IF bandwidth to a baseband frequency centered at 5.6 MHz, The
conversion gain of the 89411A can be varied to be compatible with several
different spectrum analyzers.

The following describes elements appearing in the front and rear panel
illustrations, the block diagram, and the setup diagram (on the next page).

(rear) The IF Input connector. This is a 21.4 MHz signal from the rear
panel of an RF or microwave spectrum analyzer. The input signal level
should be approximately −20 dBm to achieve optimum performance from
the 89411A.

(front) The Output connector provides a 5.6 MHz signal that goes to
the Channel 1 input connector on the Agilent 89410A Vector Signal
Analyzer. This signal should be approximately −15 dBm. The
downconversion gain step attenuator may be adjusted to change the
gain.

(rear) The Reference Input connector accepts a 10 MHz reference

signal from the RF or microwave spectrum analyzer.

(rear) The Reference Output connector provides a 10 MHz reference

signal that goes to the Agilent 89410A Vector Signal Analyzer.

(rear) The downconversion gain switch controls a step attenuator

which allows you to adjust the total gain of the downconverter from +5
dBm to −15 dBm in 5 dB steps, the goal being an output of approximately
−15 dBm. When the IF Input level is −20 dBm, an attenuator setting of 0
dB yields an output level of −15 dBm.

(front) The Reference Unlocked indicator. This indicator lights when

the 10 MHz reference input signal at is <0 dBm, or when there is a
malfunction in the 89411A local oscillator.

15-3
 Using the Agilent 89411A Downconverter

Connection and setup details for the Agilent 89411A

89410A

89411A

RF or microwave
spectrum analyzer

Agilent 89411 setup diagram

15-4
 Using the Agilent 89411A Downconverter

q If the RF/microwave analyzer is the HP/Agilent 8566A/B:

1. The frequency reference output of the 8566 is connected to the reference input
in the 89411. The reference output of the 89411 connects to the external
reference input of the 89410A (both signals ≥ 0 dBm).
2. 21.4 MHz IF output of the 8566 connects to the IF input of the 89411. This
signal level is nominally −20 dBm when the signal level on the 8566 is at its
reference level (top of screen).
3. The (front panel) baseband output of the 89411A is connected to the channel 1
input on the 89410. Its level is nominally −15 dBm when the step attenuator in
the 89411A is set to the rightmost position (labeled +5 dB). For optimum
results the input range of the 89410A should be set to −14 dBm.

Note If you have installed option UFG or UTH (4 MByte or 20 MByte extended RAM
and additional I/O), the SYSTEM INTERCONNECT port is provided only for
connection to the spectrum analyzer used with the Agilent 89411A 21.4 MHz
Down Converter.
The GPIB address for the port is one higher than the analyzer address. For
example if [Local/Setup], [analyzr addrs] is 19, the address of the port on option UFG
or UTH is 20. The port is also available via IBASIC at select code 10.

The HP/Agilent 8566A/B should be set up as follows:

Set the center frequency to the frequency of your signal. Set the frequency
span to 0 Hz, and the resolution bandwidth to 3 MHz. The reference level
should be adjusted so that the signal lies within 1 division of the top of the
screen. The vertical scaling should be set to linear rather than log. You
should also set the sweep time to a large value (e.g. 100 sec.) to prevent
the sweep retrace from causing unwanted transients in your measurement.

15-5
 Using the Agilent 89411A Downconverter

q If the RF/microwave analyzer is an MMS system:

Several possibilities exist here, depending on the combination of RF and IF
modules present. In addition, the frequency reference connections are
more involved. The conversion gain of the system depends on which IF
module supplies the 21.4 MHz IF signal, and the attenuator setting of the
RF section. The nominal conversion gains in the section that follows do
not include effects of frequency response of the RF section. The
conversion gain is generally smaller at higher frequencies, especially for the
RF sections with preselectors (e.g. 70905, 70906, and 70908). For
simplicity, the stated conversion gains assume the attenuation of the RF
section is set to 0 dB. However, to prevent damage to the mixer in the RF
section, it is recommended that the attenuation be set to at least 10 dB.
You should make sure the reference level and attenuator settings are
appropriate for your measurement. The reference level should be set to a
level equal to or larger than the largest signal you expect to measure.

At this time you should also set the following parameters:

Set the center frequency to the frequency of your signal and the frequency
span to 0 Hz. The reference level should be adjusted so that the signal lies
within 1 division of the top of the screen. You should also set the sweep
time to a large value (e.g. 100 sec.) to prevent the sweep retrace from
causing unwanted transients in your measurement.

The frequency reference module (70310A) should be ordered with the

standard configuration (i.e. do not order opt, 002 which deletes the
ovenized oscillator), unless there will always be a good externally supplied
reference signal present.
1. If the system contains a 70902A and a 70903A IF section, the signal flow is
normally from the RF section to the 70903A, then to the 70902A. The 21.4 MHz
OUT port on the 70903A drives the 70902A, and the AUX 21.4 MHz OUT port on
the 70902A drives the 89411A IF input. The conversion gain of the system is
nominally −5 dB in this case.
2. If the system contains only the 70902A IF section, the RF section is connected
to the 70902A, and the AUX 21.4 MHz OUT port of the 70902A drives the IF
input of the 84911A. The conversion gain is nominally −5 dB.
3. If the system contains only the 70903A, the RF section drives 21.4 MHz input,
and its 21.4 MHz OUT port drives the 84911A input. The conversion gain is
nominally +5 dB.
4. If the system contains only an RF section (such as the 70904A) then the output
can be taken from it directly to the 89411A input. The conversion gain of the
RF section is nominally 5 dB.

15-6
 Using the Agilent 89411A Downconverter

In some of the situations listed above, the combination of input signal

level, attenuation, and IF section employed may result in a IF level at the
89411 input which is too large. The conversion gain switch on the rear of
the 89411A should be set to provide a lower amount of conversion gain in
this case. To obtain optimum performance from the 89411A, its conversion
gain should be set, if possible, to give −15 dBm at the 89411A front panel
output for the largest measured signal. In some cases the amplitude may
be too small and you should lower the range setting of the 89411A to
obtain the best dynamic range.

Example If you have an RF signal level of +10 dBm into the MMS spectrum analyzer
configured as in (1) above, you would probably want to use the following setup.
The MMS spectrum analyzer’s reference level is set to +10 dBm resulting in 20
dB of RF attenuation and the signal at the 70902A AUX 21.4 MHz OUT port is
−15 dBm. The conversion gain of the 89411A should be set to 0 dB and the
signal provided to the 89411A is −15 dBm, as desired.

For more complete information on MMS system components, refer to the

operation manual supplied with your system, and to the Modular
Measurement System catalog.

15-7
 Using the Agilent 89411A Downconverter

Calibration

This task may be performed to eliminate the frequency response

contribution of the 89411A and attached RF/microwave spectrum analyzer
when making down-converted measurements.

The 89410A has the capability to control certain HP/Agilent spectrum

analyzers over the GPIB by issuing the appropriate commands for changing
the analyzer’s frequency and vertical scaling. These commands set the
center frequency, span, and linear display mode. The commands are
compatabile with the HP/Agilent 8566A/B, HP/Agilent 8568A/B and MMS
systems using L.O./Control module 70900A/B.

So, in addition to the signal cabling you would normally use with the
89411A, the GPIB should be connected and the address of the RF analyzer
should be checked so that it does not conflict with the address of any
other attached peripherals or other instruments. You should also set the
89410A to be system controller (under the ‘Local/Setup’ key). If you are
using an MMS (70000) system or have other devices which are capable of
being system controller, you should make sure they are not configured to
be system controller at this time.

Note If you have installed option UFG or UTH (4 MByte or 20 MByte extended RAM
and additional I/O), the GPIB port on the option board is used for control of an
external receiver. The main GPIB port can be used for other functions such as
allowing control of the 89410A from an external computer. The 89410A is
always the system controller on the option board GPIB interface while the
main GPIB interface can be configured as either system controller or
talker/listener.

Take the cable which would normally be connected from the 89411A
output to the 89410A channel 1 input and connect it from the 89410A
source to the channel 1 input instead. Set up the 89410A as follows:
1. Initialize the analyzer. Press [Preset].
2. Set up the IF parameters.
Press [Instrument Mode], [receiver], [IF section (0-10 MHz)].
Press [mirror freq off]*.
Press [external setup], [IF center], 5.6 [MHz] (required by the 89411A).
Press [IF bandwidth], 3 [MHz]†.
Press [minimum freq], 0 [Hz].

* For frequencies in the lowest conversion band of the RF analyzer one ‘mirroring’ will occur, and another
one is contributed by the 89411A.
† This should be set to the maximum RBW that the RF/ microwave analyzer supports.

15-8
 Using the Agilent 89411A Downconverter

Press [maximum freq] 22 [GHz] (appropriate for the HP/Agilent 8566A/B).

3. To enable control of the external spectrum analyzer via GPIB:

Press [rcvr control on]
Press [Local/Setup], [system controller], [peripheral addresses], [ext rcvr adrs], 18, [Enter]
(or use whatever address the attached RF/microwave spectrum analyzer is
set to).
4. Set up the other measurement parameters.
Press [Instrument Mode], [vector].
Press [System Utility], [time domain cal on].
Press [Frequency], [center], 5.6 [MHz].
Press [span], 3 [MHz].
Press [Range], −14 [dBm].
Press [ResBW/Window], [rbw mode arb].
Press [main window], [uniform].
Press [Return], [num freq pts], 801, [Enter].
5. Set up the source and turn it on:
Press [Source], [source type], [periodic chirp].
Press [Source], [level], −22, [dBm].
Press [source on].
6. Set up the trigger:
Press [Trigger], [trigger type], [internal source].
7. Set up averaging.
Press [Average], [average type], [time].
Press [Return], [num averages] , 100, [enter].
Press [average on].
8. Set up a time display:
Press [Measurement Data], [main time ch1].
Take a measurement: press [Meas Restart].
Wait for the average count to reach 100.
Store this data in D3: press [Save/Recall], [save trace], [into D3].
Turn averaging off: press [Average], [average off].

(continued on next page...)

15-9
 Using the Agilent 89411A Downconverter

9. Turn on the arbitrary source:

Press [Source], [source type], [arbitrary], [arb data reg], [D3].
Press [Source], [level], −32, [dBV pk].
Set up a frequency display: press [Measurement Data], [spectrum ch1].
Turn averaging on: press [Average], [average on].
Take a measurement: press [Meas Restart].
Wait for the average count to reach 100.
Save the trace in D1: press [Save/Recall], [save trace], [into D1].

Connect the source to the input of the RF/microwave analyzer and the
89411A output to the 89410A channel 1 input.

Set the center frequency of the RF/microwave analyzer to 5.6 MHz and the
span to 0 Hz Make sure the resolution bandwidth is set to 3 MHz and the
sweep time is set to a large value. Set the display scale to linear and
adjust the reference level to −22 dBm.
10. Turn averaging off:
Press [Average], [average off].
The source signal now appears on the 89410A screen.
Press [Math], [define F5], [meas data], [spectrum ch1], [/], [data register], [D1], [enter].
This defines function SPEC1/D1 to view the normalized trace you are
about to produce. Function F5 may already be set up.
11. Set the receiver to external:
Press [Instrument Mode], [receiver], [external)].
Set the frequency: press [Frequency], [center], 5.6, [MHz].
Set up the source using the source’s automatic span quantization:
Press [Source], [source type], [periodic chip], [arbitrary].
Set the range: [Range], −14, [dBm].
Set the Meas Data to the function: [Measurement Data], [math func], [F5].
Turn averaging back on: [Average], [average on].
Restart the measurement: press [Meas Restart].
When the data is collected, save trace in D1: [Save/Recall], [save trace], [into D1].

15-10
 Using the Agilent 89411A Downconverter

12. Display D1:

Press [Measurement Data], [data reg], [D1]. This is the conversion gain and IF
response of the RF analyzer and the 89411A (see figure).

Agilent 89411A correction data

13. Set the Measurement Data back to the function (F5) containing SPEC1/D1.
Now the screen displays the 89410A source signal with any frequency response
contributions of the RF/microware analyzer and the 89411A removed. This
procedure does not correct for RF unflatness contributed by the RF/microwave
analyzer, only IF effects are corrected. However, these effects are the most
important since they vary more significantly over the IF bandwidth (3 MHz in
this example).
14. You should now set the instrument to a configuration appropriate for your
measurement. For example, the source could be turned off, the trigger set
back to free run, the window type set to flat top, and averaging disabled.

15-11
16

Extending Analysis to
26.5 GHz with 20 MHz
Information Bandwidth
This chapter shows you how to use the HP/Agilent 71910A wideband receiver
to extend the frequency coverage and information bandwidth of an
Agilent 89400-series vector signal analyzer. Information in this chapter is from
Product Note 89400-13.

16 - 1
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Overview
Some applications require an information bandwidth and frequency coverage
beyond that offered by Agilent 89400-series vector signal analyzers.
Instruments such as the 89441A are limited to frequencies below 2.65 GHz and
information bandwidths of 7 MHz. This precludes the analysis of many spread
spectrum, radar and satellite signals which typically occupy more than 7 MHz
bandwidth and that may exist only at microwave frequencies.

By combining two products—the Agilent 89410A vector signal analyzer and the
HP/Agilent 71910A wideband receiver into a single measurement system—the
unique capabilities of the vector signal analyzer can be used on signals with 20
MHz bandwidth at frequencies up to 26.5 GHz. This chapter describes how to
configure, calibrate and operate an 89410A analyzer with the 71910A wideband
receiver.

The HP 89410A vector

signal analyzer and
HP 71910A wideband
receiver together form a
wideband vector signal
analyzer system.

Agilent 89410A and HP/Agilent 71910A Wideband Vector Signal Analyzer System

16 - 2
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

System Description
The wideband vector signal analyzer system consists of two major
components: an 89410A two-channel vector signal analyzer and a 71910A
wideband receiver with a wideband IF and quadrature outputs.
Note The 89441A consists of an 89410A (the IF section) and an 89431A (the RF section).
To use an 89441A with the 71910A, disconnect the RF section and connect the IF
section to the 71910A as described in this chapter.

The 89410A provides the user interface and display, and performs all of the
signal processing.

The 71910A is basically a microwave spectrum analyzer with additional

features to optimize it for and signal monitoring applications. In this
application, it converts the RF or microwave signal into a baseband signal
which can be further processed by the vector signal analyzer. As the front end
of the measurement system, it also provides the necessary gain or attenuation.

16 - 3
Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Simplified System Block Diagram

16 - 4
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Agilent 89410A Operation

The 89410A has two input channels (with option AY7), each with a bandwidth
of 10 MHz. Normally, this would represent the maximum bandwidth of the
signal to be analyzed. However, the 89410A is capable of treating the signals on
each channel as two parts of the same signal. That is, the signal going into
channel one represents the real part of a complex signal, and the signal going
into channel two represents the imaginary part. These two signals are usually
referred to as the in-phase and quadrature-phase components, or simply
I and Q.

The 89410A digitizes the I and Q signals which are, by themselves, real signals
and then combines them internally into a single complex signal of the form
I+jQ or CH1+jCH2, where j represents the square root of negative one. This
new complex signal, which exists only in digital form, has a maximum
bandwidth of 20 MHz, or twice the input bandwidth of the vector signal
analyzer.

HP/Agilent 71910A Operation

The 71910A wideband receiver (also called the MMS system) consists (at a
minimum) of an LO/controller module, an RF front-end module, a wideband IF
module, and a precision frequency-reference module. For optimum
performance of the entire system, the signal level should exist within a certain
range as it propagates through the system.

Although it is a simplification, the MMS system components can be viewed as

consisting of just a few blocks. They are an RF section with an attenuator and
conversion stage, an IF section with adjustable gain, and an IQ demodulator, as
shown in the previous block diagram.

The RF attenuator must be set to ensure the signal level reaching the
conversion stage doesn’t cause damage or distortion. This attenuator has a
step size of 5 dB.

The IF section has a bandwidth of 100 MHz which ensures a relatively flat
frequency response over the center 20 MHz used in this system. A flat IF is
important in vector signal analysis. An IF with a significant amount of
amplitude unflatness or group delay distortion would produce significant
errors. This is especially true for modulation analysis where the IF
characteristics would introduce distortion in the time domain characteristics of
the signal. For example, group delay distortion in the IF would result in
increased inter-symbol interference in a digitally-modulated signal.

16 - 5
Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

The IF section provides filtering and variable gain. The filtering can be used to
prevent unwanted signals from reaching subsequent blocks in the system and
causing distortion. The example program sets the filters to their widest
bandwidth to obtain the best accuracy. This ensures the minimum amount of
frequency response error (magnitude and phase) and ensures the highest
accuracy for the IQ demodulation. The gain of the IF section is set so that the
IQ demodulator operates over a signal level range where it is most linear. The
IF gain resolution is 1 dB. The 89410A input range is set to be compatible with
the the full-scale output of the IQ demodulator.

Mirrored Spectrums
The vector signal analyzer is responsible for reducing the measurement
bandwidth to 20 MHz and below. When the microwave receiver is used at
frequencies below 12.8 GHz, the spectrum obtained using I+jQ is mirrored
about the center frequency. There are two ways to compensate for this
mirroring. The first is to simply swap the I and Q outputs. While this works, it’s
inconvenient and makes calibration more difficult.

A simpler way to compensate for the mirroring is to conjugate the complex

signal. In other words, I–jQ instead of I+jQ. The vector signal analyzer has a
spectral mirror key which conjugates the CH1+jCH2 data. This corrects both
time and frequency domain results without affecting the I and Q calibration.

IBASIC Example Program

As a measurement system, there is an obvious need for software to link the two
instruments together. This software should provide a user interface to the
system, as well as provide for system calibration.

An Instrument BASIC example program is available which provides these

functions. The primary purpose of the example program is calibration.
However, it also provides for simple control of center frequency and reference
level.

Once the system is calibrated and the center frequency and reference level are
properly adjusted, the program is paused or terminated to provide access to all
the measurement and analysis features of the vector signal analyzer.

The example program is written in Instrument BASIC and is included on the

Instrument BASIC Example Programs disk (this disk comes with Instrument
BASIC, option 1C2 ). It can be run on the 89410A, or on an external controller.
You can also obtain the example program from your local Agilent sales
representative.

16 - 6
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

System Configuration
This section describes the components, the physical connections between
components, and the software required to create a wideband vector signal
analysis system. Two system configurations are described: one includes the
Modular Measurement System (MMS) display and one does not.

Important The system may or may not benefit from the 70004A display. This choice should be
based on the types of measurements you intend to make. If you are interested
primarily in making measurements of complex signals which can be viewed or
analyzed completely by the 89410A, the display may not be necessary.
If however you are interested in making scalar spectrum measurements over a
range of frequencies wider than the capabilities of the 89410A (20 MHz), or you
intend to use other spectrum measurement capabilities provided by the 71910A
system, then your system must include the display. The calibration program
facilitates measurement of 20 MHz information bandwidth complex signals, but
only provides access to a limited subset of the 71910A functionality. The other
capabilities of the 71910A receiver are only available with a dedicated display and
front panel keypad present. Most systems require the presence of the display.

Agilent 89410A Configuration

The 89410A must have the options and firmware shown in the following table.
The instrument firmware must be revision A.04.00 or later and can be upgraded
by ordering 89410U option UE2. Option AY7 adds a second input channel to
support measurements of the complex (I+jQ) output signal of the 71910A.
Option 1C2 adds Instrument BASIC which you need to run the IBASIC example
program (an IBASIC example program). The IBASIC example program
controls and calibrates the system.

Note Additional information on configuring the 89410A can be found in the Agilent 89400
Series Vector Signal Analyzers Configuration Guide (p/n 5964-3630E).

16 - 7
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

HP/Agilent 71910A Configuration

The 71910A receiver contains all the Modular Measurement System (MMS)
components which make wideband vector signal analysis possible. Minimally,
Option 004 (Analog I/Q Outputs) must be ordered and, depending on the other
measurements you may want to make with the system, the MMS display
(70004A) may also be needed. A system without the MMS display would be
ordered as an 71910A with Option 004, Option 011 and Option 012. In this
configuration, the display is deleted, but the 70310A reference module is added
(see the following table).

An existing 71209A Option 001 spectrum analyzer can be upgraded to support

wideband vector signal measurements by ordering an 70911A IF module
(Option 098 or Option 099 firmware upgrade required). You may also need to
upgrade the power supply in the 70001A mainframe. The receiver firmware
personality must be installed to facilitate the wideband vector signal
measurements described in this paper.

Note For additional information on firmware installation and system configuration, refer
to the 71910A User’s Manual.
A number of other modules and options are available for the family of MMS
spectrum analyzers. For more information, refer to the 71910A Product Overview
(p/n 5963-0087E) and the HP/Agilent 70000 Modular Measurement System Catalog
(p/n 5965-2818E).

16 - 8
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Recommended Configuration

Agilent 89410A Vector Signal Analyzer

(must be firmware revision A.04.00 or greater)
Option AY7 Second 10 MHz input channel
Option AYA Vector Modulation Analysis
Option 1C2 Instrument BASIC (includes the example program described earlier in this chapter)
HP/Agilent 71910A Wideband Receiver
(must be instrument firmware revision 94120 (B.05.00) or greater;
70910A RF module must be installed)
Configurations With the 70004A Display
Option 004 Analog I/Q outputs
Configurations Without 70004A Display
Option 004 Analog I/Q outputs
Option 011 Single mainframe configuration (deletes 70004A, 70902A, 70903A, and 70310A)
Option 012 Adds 70310A reference module
Other Accessories
8120-1838 50 ohm BNC–BNC cable, 12" (two required)
8120-6269 50 ohm BMC–SMB cable, 24" (one required)
10833A GPIB cable, 1 meter (one required)

16 - 9
Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

System Connections
System connections are shown in the following figures. The first figure shows
the system without the 70004A display. The second shows the system with the
70004A display and also includes the 70902A and 70903A IF modules, which
implement IF bandwidths to support traditional scalar spectrum analysis.

The rear panel views in the figures show the connections between the 89410A
and the 71910A. In both configurations, the 10 MHz frequency reference for the
system is provided by the 70310A reference module. This is the recommended
configuration.

If your 89410A includes Option UFG or UTH, you have a second GPIB
connector. Referring to either figure, note that the GPIB connection is made
from the main GPIB port of the 89410A (labeled GPIB) and not the connector
labeled System Interconnect.

Front panel
connections

Rear panel
connections

HP 89410A

HP 71910A
Option 004
Option 011
Option 012

System Without the 70004A Display

16 - 10
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Front panel
connections

Rear panel
connections

HP 89410A

HP 70004A
Display Section

HP 71910A
Option 004

HP 70001A
Mainframe

System With the HP/Agilent 70004A Display

16 - 11
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

System Connections (applies to all systems)

Connect This Output: To This Input:

Component Output Component Input
70900B LO Module LO OUT 70910A RF Module 1st LO IN
300 MHz OUT 1 70910A RF Module 300 MHz IN
300 MHz OUT 2 70911A IF Module 300 MHz IN
TUNE SPAN 70910A RF Module TUNE SPAN

70910A RF Module 321.4 MHz OUT 70911A IF Module 321.4 MHz IN

70911A IF Module WB VIDEO OUT 70911A IF Module WB VIDEO IN

321.4 MHz OUT 70911A IF Module 321.4 MHz OPT IN
VID OUT TO LO 70900B LO Module VIDEO IN

70310A Precision Reference 100 MHz 70900B LO Module 100 MHz IN

Module
10 MHz 89410A EXT REF IN

Additional connections needed when the system is configured with the 70004A display and the 70902A and 70903A IF
Modules.

70902A IF Module VIDEO OUT 70903A IF Module VIDEO IN

70903A IF Module 21.4 MHz OUT 70902A IF Module 21.4 MHz IN

VIDEO OUT 70911A IF Module NB VID IN

70910A RF Module 21.4 MHz OUT 70903A IF Module 21.4 MHz IN

16 - 12
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Operation
The vector signal analyzer provides for all signal processing, display and
analysis. With the exception of scalar analysis, all 89410A features are available
in the wideband vector signal analyzer system.

The example program only needs to provide control over center frequency,
input attenuation, spectral mirroring and system calibration. As a convenience,
it also initializes both instruments for wideband vector signal analysis. This
includes, but is not limited to, changing instrument modes, setting inputs to DC
coupling and adjusting RBW settings. The example program also lets you
initiate the MMS self-calibration routines.

The upper trace shows

the spectrum of a 10
Mb/s QPSK signal
measured using the
wideband vector signal
analyzer system. The
trace is 20 MHz wide.
The lower trace is an eye
diagram of the same
signal obtained using the
HP 89410A vector
modulation analysis
(Option AYA).

Spectrum and Eye Diagram of a 20 MHz, 10 Mb/s QPSK Signal

16 - 13
Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Controlling the Receiver

As previously described, the system can be configured with or without the
70004A display. When the display is present, the center frequency and reference
level can be adjusted on the microwave receiver. While this may seem more
convenient, there are good reasons for using the features found in the example
program to control the receiver from the vector signal analyzer.

Changing Center Frequency

For example, when changing the center frequency, the example program labels
the traces on the vector signal analyzer to reflect the center frequency of the
measurement. This is useful since the x-axis annotation always shows a center
frequency of zero hertz whenever the receiver mode is set to CH1+jCH2.

To facilitate measurements over a range of frequencies, a frequency increment

function has been added to the example program. This function is accessed
from a sub-menu via the Freq Step key in the main menu. The sub-menu
contains keys for setting the desired frequency increment as well as stepping
the center frequency up and down with just one key stroke. These functions
take care of setting the 71910A center frequency, updating the annotation on
the 89410A, and mirroring the spectrum display if necessary.

Setting the Mirror Frequency Key

Another important function that the example program handles is the setting of
the mirror frequency key, used to flip or invert the spectrum of the signal about
zero hertz. This operation may be necessary depending on the center
frequency. In the microwave spectrum analyzer, center frequencies below 12.8
GHz will produce an output spectrum that is inverted. The frequency inversion
is due to the mixing scheme used in the instrument. When setting the center
frequency, the example program will check to see if the spectrum will be
inverted (based on center frequency) and will set the mirror frequency key to
compensate. The mirror frequency key actually causes the time data to be
conjugated; that is, the sign of the Q channel data is negated. Conjugating the
time data will not only mirror the spectrum, it will also correct the time data for
modulation analysis.

Changing the Reference Level

The example program is even more useful for changing the reference level
(input attenuation). When the input attenuation is adjusted in the microwave
receiver, the gain of the signal path may change. If it does, the absolute level
displayed on the vector signal analyzer will be wrong. When changing the
reference level, the example program will also change the gain calibration to
reflect any nominal changes in the gain of the signal path. This ensures that the
reference level can change without affecting the level of the result.

16 - 14
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

With two instruments, there are actually two sets of input attenuators that need
to be adjusted. Obviously, the input to the microwave receiver is adjusted
based on the level of the signal to be measured. The vector signal analyzer
inputs are adjusted to reflect the I and Q signal levels at the output of the IF
module. This setting is fixed and has been chosen as a compromise between
noise and distortion performance. A variable in the example program can be
adjusted to provide different levels of performance.

When you set the reference level via the program, that level should be the
expected peak level of the signal you are measuring. Keep in mind that the
peak-to-average ratio for your signal may need to be taken into account. The
program assumes a value of 4 dB. If the signal exceeds the level you specify,
distortion of the signal or possible damage to the system may result.

To provide better measurements of low-level signals, the program lets you set
the RF attenuator to zero dB. If you set the reference level below –10 dBm, a
prompt appears asking you to confirm the removal of the last 10 dB of
attenuation. You should consider your answer carefully since removing all
RF attenuation could degrade the measurement. In particular, it could result in
increased distortion, possible exposure of the 70910A RF module’s input mixer
to damage, and reduced return loss. If you are not sure your signal level will
always be below –10 dBm, you should never set the RF attenuation to 0 dB
(answer ‘NO’). Leaving in 10 dB of attenuation will help protect the receiver’s
input circuitry, with the consequence that there is a small signal-to-noise
degradation of your measurement.

The maximum reference level allowed is +30 dBm; the minimum is –60 dBm.
Signal levels above +30 dBm could damage the 70910A RF module, while levels
below –60 dBm will exhaust the available IF gain resulting in poor S/N
performance. The program warns you when you request a signal level outside
of this range.

To optimize the overall signal-to-noise ratio of the measurement the IF step

gain is used to ‘interpolate’ between RF attenuator values. This allows the
overall signal level to be set where the best S/N can be achieved for all
reference level settings.

16 - 15
Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

In certain situations the 89410A front panel over-range indicators (amber

LEDs) may be lit, even when the reference level has been set above the level of
your input signal. This indication is conservative, and in most cases there is no
significant degradation of the measurement. Even so, if the over-range
indicators are lit, you should verify your setup and signal levels. This situation
is most likely to occur if the input frequency is below 50 MHz and the selected
reference level is below about –20 dBm. With this combination there is a
chance that some LO feedthrough, and other mixing products from the MMS
downconverter will be large enough to contribute significantly to the energy
appearing at the 89410A input and activate the over range indicators. In most
cases these signals are out of band (above 10 MHz) and will not affect the
measurement. If the out-of-band signals are larger than the desired signal,
there is a possibility of intermodulation products being generated. In almost all
cases, the out-of-band signals will be too small to cause any measurement error.

Resolution Bandwidth
The example program selects a 100 MHz resolution bandwidth for the down
converter. At low measurement frequencies, this setting may allow signals at
the IF above 10 MHz to pass with only a small amount of attenuation. The RBW
setting is a conscious trade-off between selectivity and calibration accuracy. If
the RBW filter were set to a smaller value, you might be able to eliminate some
of the unwanted conversion products for certain measurements. However, the
frequency rolloff of a narrower filter setting would reduce the accuracy of the
measurement.

DC Offset and LO Feedthrough

DC offset and LO (local oscillator) feedthrough will cause a residual response
at the center of the screen. The contribution of the 89410A to this response is
from DC offset in it’s input circuits. This term is measured and removed on a
periodic basis automatically by the 89410A unless you explicitly disable this
‘Auto-zero’ function (not recommended). When the program initializes the
89410A, this function is enabled, and should be left enabled to achieve the best
and most stable system calibration.

16 - 16
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Calibrating the System

The system calibration described earlier is very easy to perform using the
example program. Once the instruments are properly connected, the example
program is started—automatically configuring the instruments for wideband
vector signal analysis. Pressing the calibration softkey will initiate the
calibration. Unless the example program has been modified, the program will
prompt you to connect a 300 MHz, –10 dBm signal to the input.

Once connected, the calibration only takes a few moments to perform. After
calibration, you can adjust the center frequency and reference level or exit the
program to operate the vector signal analyzer. Normally, you will pause (by
pressing the BASIC key) rather than quit the program. This allows the program
to be easily continued when you want to change center frequency or reference
level, or to perform another calibration.

To minimize the contribution of the 71910A system components to any

feedthrough or offset terms, you can run the 71910A’s internal calibration
routine. If the 71910A system display is present, you can invoke the calibration
by pressing CAL ALL under the amplitude menu. If you are not using the
display, the IBasic program has a softkey to invoke this function. It is labeled
CAL MSA and is located under the top-level softkey labeled MSA Diag.

When making measurements at microwave frequencies, you may want to

consider performing the calibration at a frequency nearer your measurement
frequency instead of 300 MHz (the default value). You will need a signal whose
level is accurately known to do this. The program contains a constant
Cal_frequency which can be changed to run the calibration routine at the
desired frequency. The program provides no explicit support for measurements
and calibration above 26.5 GHz at this time. If any external conversions occur
as part of your measurement system, mirroring of the frequency spectrum may
be an issue.

The stability and longevity of the calibration are governed for the most part by
temperature drift of the various system components. If the temperature has
changed by more than 5 or 10 degrees C, you may be able to improve the
accuracy of your measurement by re-running the system calibration.

16 - 17
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Calibration Methods
The 89410A vector signal analyzer and the 71910A receiver are both capable of
self calibration. To obtain the best system performance, the instruments must
be calibrated together.

The IBASIC example program calibrates both instruments as a system. The

program compensates for system-induced errors, such as unequal cable lengths
for the I and Q signals. It also fine tunes factory calibrations, such as
quadrature, for optimal performance.

The calibrations affect more than just amplitude accuracy. As shown in the
following illustration, calibrations also improve the dynamic range of the
system.

Calibrations provide
better accuracy and
improve dynamic range.
Upper trace shows
spectrum of 300 MHz
calibrator as measured
before system
calibrations. Lower trace
shows measurement of
same signal after system
calibration.

Calibration Improves Accuracy and Dynamic Range

To perform the calibrations, the IBASIC example program uses the 300 MHz
calibrator built into the microwave spectrum analyzer. The program measures
the magnitude and phase of the calibration signal while adjusting the center
frequency of the microwave spectrum analyzer. It also measures the amount of
residual DC in the I and Q signals. From these measurements, the program
determines the IQ gain imbalance, IQ magnitude error, IQ delay mismatch, IQ
quadrature error and the DC offset. It then adjusts both instruments to
minimize the effects of the errors.

16 - 18
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Note For additional details on how errors in the signal path can affect the spectrum,
refer to Product Note 89400-13, which contains most of the information in this
chapter plus one additional section, Appendix A. Appendix A contains a theoretical
discussion on three topics: how complex signals allow a doubling of bandwidth
relative to real-only signals, how the spectrum can be mirrored simply by
conjugating the complex time data, and how measurement errors can produce
spurious signals which reduce the dynamic range of a measurement.

DC Offset
Residual DC limits the dynamic range of the measurement by introducing a
spurious term in the center of the spectrum (zero hertz). The DC offset
calibration measures the amount of residual DC in both the I and Q channels
and determines the optimal settings for the offset DACs in the 70911A IF
module.

For DC offset, the two channels are independent of one another and are
calibrated separately. The following procedure describes the calibration for the
I channel. The same procedure is also performed on the Q channel.

At the start of the DC offset calibration, the example program instructs the
89410A to perform an internal offset calibration. This calibration removes DC
from the vector signal analyzer inputs without consideration of the input signal.
This is done so that subsequent internal calibrations, which are performed
automatically by the vector signal analyzer, will not affect the system
calibration.

After the vector signal analyzer’s offset calibration, the example program
records the amount of DC remaining in the I channel along with the current
setting of the I channel offset DAC in the 70911A. The DAC is then adjusted by
a small amount and the amount of DC in the channel is measured again. The
two DC measurements (y1 and y2) and the two DAC values (x1 and x2) are
combined to determine the slope and offset of a line y=Ax+B where A is the
slope and B is the offset. The optimal setting for the DAC is then –B/A. This is
the value that drives y, the residual DC, to zero.

16 - 19
Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Channel Match
Channel match is also important for preserving the dynamic range of the
measurement. System errors that cause mismatch include gain imbalance,
delay mismatch and frequency response differences between the I and Q
signals.

A frequency response mismatch can occur in both the IF module and the vector
signal analyzer input channels. Only the center 20 MHz of the IF module’s 100
MHz bandwidth is used (I and Q each have 50 MHz bandwidth), so the
mismatch there is relatively small. The mismatches in the vector signal analyzer
are compensated by the instrument’s built-in calibration routines.

IQ Gain, Delay Match

The example program only attempts to match gain and delay. The program
determines the gain and delay parameters by measuring the I and Q signals
generated when the 300 MHz calibrator is connected. To simplify the
measurement, the vector signal analyzer is taken out of the channel combine
mode (CH1 + jCH2) and put into two-channel mode.

The center frequency of the spectrum analyzer is adjusted to 291 MHz. With a
300 MHz signal and a center frequency of 291 MHz, the I and Q outputs are both
9 MHz sine waves. In a perfect system, these 9 MHz sine waves would have a
phase difference of 90 degrees and would have identical amplitudes. In a real
system, the amplitudes will probably be different, and the phase difference will
be something other than 90 degrees.

As shown in the following illustration, the phase error has two components—
error due to quadrature error and error due to IQ delay mismatch. Quadrature
error in the IF module produces a phase error that is independent of the
relationship between the calibrator frequency and the center frequency of the
measurement. Delay mismatch produces a frequency-dependent phase error
term that is zero when the center frequency and the calibrator frequency are
the same (that is, the difference frequency is zero and I and Q have no AC
component).

To determine the delay and quadrature error, the center frequency is adjusted
in 2 MHz steps from 291 MHz to 309 MHz, and the phase differences between I
and Q are recorded. Then, using a least-mean-square algorithm, a line is fit to
the phase error verses frequency data.

The slope of the line corresponds to the delay mismatch, and the offset to the
quadrature error. The absolute signal level and gain match are measured at the
same time as the phase. At each center frequency, the magnitude squared of the
signal (I 2 + Q 2 ) is computed, as is the ratio between I and Q.

16 - 20
 Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

The square root of the average magnitude squared is used to set the overall gain
adjustment. This minimizes the absolute error across the 20 MHz bandwidth
rather than the error at the center frequency.

The average ratio is used to determine the gain mismatch. The gain and delay
parameters are used to configure the external calibration parameters of the
vector signal analyzer. It’s worth noting that the errors are quite small. For
example, the delay mismatch is usually less than 100 psec.

Measured Phase

For a sinusoidal input, +90 deg

I and Q outputs should Ideal Phase
also be sinusoidal and
should have a phase Slope is caused by
difference of exactly Phase Error delay mismatch
+/-90 degrees. (measured Delay)
Quadrature error and
Offset is caused by quadrature error
delay mismatch -10 Mhz
between channels 0 Hz 10 Mhz
introduce a phase error
that is a function of the
relationship between
the center frequency of
the measurement and
the frequency of the
input signal.
-90 deg

Effects of Quadrature Error and Delay Mismatch

16 - 21
Extending Analysis to 26.5 GHz with 20 MHz Information Bandwidth

Quadrature
The final calibration is for quadrature error. Although a value for quadrature
error was obtained in the channel match calibration, the number cannot be
used by itself to set the quadrature adjustment DAC in the IF module as the
mapping from DAC setting to phase is not known.

The DAC setting for the quadrature is determined much like the DAC setting
for the residual DC. First, the phase difference between the channel one and
channel two signals is measured and the current DAC setting recorded. A
second measurement of the phase difference is made after the DAC has been
adjusted by a small amount.

The two DAC values and the two phase difference values can then be used to
compute how far to adjust the DAC to compensate for the quadrature error
observed in the delay calibration. The example program performs this
calibration with the center frequency adjusted to 309 MHz. Although the
frequency is not critical, the center frequency should be different than the
calibrator frequency to avoid potential problems with errors introduced by
residual DC in I and Q.

The example program was designed to work with a 300 MHz calibrator. Any
frequency can be used, provided the calibrator frequency variable in the
program reflects that frequency. This might be desirable if better absolute
amplitude accuracy is desired at some frequency other than 300 MHz. Other
parameters, such as delay, are not greatly affected by the center frequency of
the microwave spectrum analyzer.

16 - 22
17

Choosing an Instrument
Mode
The analyzer provides three instrument modes—Scalar, Vector, and Analog
Demodulation. This chapter provides a brief overview of these instrument
modes and guidelines for choosing the best one. You can also order additional,
optional instrument modes—for details, see these chapters: Digital
Demodulation Concepts, Video Demodulation Concepts, and Wideband CDMA
Concepts.

17 - 1
Choosing an Instrument Mode

Why Use Scalar Mode?

If you’ve used a spectrum analyzer before, you should have no trouble making
scalar measurements with this analyzer. And even if you haven’t used a spectrum
analyzer before, you’ll find this analyzer quite easy to learn.

Scalar measurements are implemented differently than measurements in the

traditional swept-tuned spectrum analyzer, but the results will appear familiar to
you. For more details on the implementation of scalar measurements see “What
Makes this Analyzer Different” and “Fundamental Measurement Interactions”.

Scalar Measurements provide displays of amplitude versus frequency for both

narrow and wide spans. Use Scalar Mode when you need:
l Very narrow resolution bandwidth with high speed
l Very low noise floor with wide spans
l High signal-to-noise dynamic range
l Maximum flexibility of span, resolution bandwidth, and speed
l Spans greater then 7 MHz

Specific examples of applications for which you would use Scalar Mode include:
l Looking at spurious signals or noise
l Looking for low level distortion products

17 - 2
 Choosing an Instrument Mode

Measurement data flow diagram for Scalar Mode

17 - 3
Choosing an Instrument Mode

Why Use Vector Mode?

The Vector Mode provides extremely fast measurements of magnitude and phase.
It can also easily change between the frequency and time domains. These
capabilities are especially useful in characterizing transient or non-stationary
signals.

If you’ve used a swept-tuned spectrum analyzer before, you already know that
narrow resolution bandwidth measurements of small frequency spans are very
time-consuming. Traditionally, swept-tuned analyzers have required very long
sweep times for narrow resolution bandwidths due to the sweep rate of the narrow
filters. The narrower the resolution bandwidth, the more time it takes for the
resolution bandwidth filters to settle. In fact, a swept-tuned analyzer’s sweep time
is inversely proportional to the square of the resolution bandwidth. As you choose
increasingly narrower resolution bandwidths (for example, when trying to resolve
close-in sidebands) the time it takes to make a measurement increases
exponentially. This characteristic is common to all conventional swept-tuned
analyzers.

However, this analyzer can make narrow band measurements very quickly in
Vector Mode. This lets you make measurements for spans from 1Hz to 10 MHz
(with resolution bandwidths from 9.6 mHz to 3 MHz) much faster than you can with
a swept spectrum measurement. For vector measurements, the analyzer uses the
FFT (Fast Fourier Transform) to convert the input signal from the time domain to
the frequency domain. The result is measurement capability up to 1000 times faster
than conventional swept spectrum analyzers.

If you have previously used FFT or dynamic signal analyzers, Vector measurements
will seem familiar to you. However, with most FFT analyzers you are restricted to
low frequency measurements, often to less than 100 kHz. But in this analyzer,
Vector Mode lets you perform FFT measurements up to 10 MHz for the Agilent
89410A and beyond 1 GHz for the Agilent 89441A.

Use Vector Mode when you need:

l Time selective (gated) frequency domain measurements
l Time data display
l Phase information
l Time capture/playback
l Narrow span with high speed
Specific examples of applications for which you would use Vector Mode include:
l Characterizing spurious response of an oscillator
l Characterizing a burst signal in both the time and frequency domain

17 - 4
 Choosing an Instrument Mode

Measurement data flow diagram for Vector Mode

17 - 5
Choosing an Instrument Mode

Why Use Analog Demodulation Mode?

The best way to make many measurements often includes the ability to
characterize the amplitude, frequency, and phase relationships between signal
components. Analog Demodulation mode provides this capability. For additional
details on this mode see “Analog Demodulation Concepts” later in this book.

Analog demodulation is derived from Vector Mode. The previous block diagram
indicates where analog demodulated measurements branch off from vector
measurements in the measurement process. Analog demodulation alters the time
record so that all operations past this point operate on demodulated data.

AM, FM, and PM signals commonly consist of a high frequency carrier signal
modulated by a lower frequency audio, video, or data signal. For these signals,
analog demodulation removes the carrier frequency leaving the modulating signal
which is contained in the sidebands of the carrier signal. An example of this
application is included in “Demodulating an Analog Signal” earlier in this book.

The most obvious use for analog demodulation is in analyzing intentionally

modulated signals as noted above. However, analog demodulation is also useful in
analyzing signals which are not intentionally modulated. For example the user may
choose to:
l Apply phase and frequency demodulation to oscillator or synthesizer designs
to examine phase and frequency trajectories during frequency hops or to
establish phase lock loop lock-up time. For an example of this type of
application see the “Characterizing a Transient Signal” chapter, earlier in this
book.
l Demodulate a burst signal to evaluate wait time in order to determine when a
signal stabilizes.
l Amplitude and phase demodulate sidebands to determine the type of
modulation present in phase noise. An example of this application is shown in
the “Measuring Phase Noise” chapter, earlier in this book.
l Demodulate bearing noise in mechanical measurements to examine
frequency instability.

17 - 6
 Choosing an Instrument Mode

The Advantage of Using Multiple Modes

You can often take advantage of the analyzer’s flexibility by using more than one
instrument mode in a measurement scenario.

Scalar—the big picture

Use Scalar Mode to identify signals present in a wide span and to evaluate small
signals very close to the noise floor. For example, in evaluating an oscillator you
can use Scalar Mode to identify spurs and harmonic distortion products and
evaluate them with very narrow resolution bandwidths.

Vector—the important details

After you have identified signals in a wide-span Scalar measurement, use Vector
Mode to quantify and analyze the important signals with phase and time data
capabilities. In the oscillator example above you could use Vector Mode to
measure oscillator sidebands.

Analog Demodulation—another view of the details

Analog demodulation provides another way to look at signals by characterizing
amplitude, frequency, and phase relationships. In the oscillator characterization
example above, you can demodulate the sidebands to determine whether the noise
is amplitude modulated or phase modulated.

17 - 7
Choosing an Instrument Mode

Instrument Mode? Measurement Data? Data

Format?
Instrument modes
When you specify an instrument mode, you are asking the analyzer to acquire input
data and process it in a certain way. For example, with the Vector Mode, the
analyzer uses an FFT algorithm to convert a single time record into the most basic
frequency domain measurement available—the spectrum. For Analog
Demodulation Mode, the analyzer transforms a complex waveform into a real, time
domain waveform.

Look at the front panel. Notice how the [Instrument Mode] key is in the
MEASUREMENT group. This is because anything you change from this group of
keys affects the way the analyzer collects input data. In general, if you change
something with these keys (such as the instrument mode, or a different start or
stop frequency), the analyzer must make a new measurement.

If you look at the analyzer’s front panel, you can see that both the [Measurement Data]
and [Data Format] hardkeys are in the DISPLAY group. If you press a key in this
group, all you are doing is selecting which type of data you want to display and how
you want to display it—you aren’t changing the way the analyzer makes
measurements or the data that was measured.

Softkeys in the MEASUREMENT group, however do affect the way the analyzer
takes input data. For example, [Instrument Mode] is in this group. So is [Frequency],
which you use to change the start and stop frequencies. So if you need to change a
parameter under any of these menus (or select a different instrument mode), you
will need to take new data.

Measurement data
Suppose you’ve selected the Vector Mode. Now you can select different types of
measurement data, for example, the linear spectrum, frequency response, or main
time data. No matter which measurement data selection you make, the analyzer
does not have to acquire new data. When you select a type of measurement data,
you are simply asking the analyzer to display a particular piece of the measurement
data that’s already been acquired.

Data format
Once you select both an instrument mode and appropriate measurement data,
choosing a data format simply tells the analyzer how you want to look at the
selected measurement data. For example, if you are in the Vector Mode and you
are viewing power spectrum data, you could change the Y-axis scaling by selecting
linear magnitude or logarithmic magnitude.

17 - 8
 Choosing an Instrument Mode

Unique Capabilities of the Instrument Modes

Many features and capabilities are available in all three measurement modes but
others are availabile in selected modes. The following table shows, at a very
general level, those features and capabilities which are not universal. The ensuing
tables show in more details the features available in each instrument mode.

General Capabilities Unique to Specific Measurement Modes

Analog
Scalar Vector Demodulation
Phase information no yes yes

Time data no yes yes

Gating no yes yes

Time capture no yes yes

Network Measurements no yes yes

Selectable detectors yes no no

Manual Sweep yes no no

Very narrow resolution bandwidths yes no no

Very low rbw/span ratios yes no no

Spans greater than 10 MHz yes no no

17 - 9
 Choosing an Instrument Mode

Display Features Available by Instrument Mode

Analog
Scalar Vector Demodulation
Measurement data spectrum yes yes yes

PSD yes yes yes

time no yes yes

frequency response* no yes yes

math functions yes yes yes

data registers yes yes yes

coherence* no yes yes

cross Spectrum* no yes yes

correlation no yes yes

capture buffer no yes yes

* with optional second channel

17 - 10
 Choosing an Instrument Mode

Measurement Features Available by Instrument Mode

Analog
Scalar Vector Demodulation
Trigger types Free run yes yes yes

channel1 + yes yes yes

channel2 yes yes yes

IF channel no yes yes

internal source* yes** yes yes

GPIB yes yes yes

external yes yes yes

Source types CW yes** yes yes

random noise yes yes yes

periodic chirp yes yes yes

arbitrary yes** yes yes

Average types rms (and rms exponential) yes yes yes

time (and time exponential) no yes yes

peak hold yes yes yes

* available only in conjunction with the arbitrary source and periodic chirp
** available only when optional RF source is disabled
+ available only in receiver modes of 0-10 MHz

17 - 11
18

What Makes this Analyzer

Different?
As you become familiar with your new analyzer, you will find that most
measurements look like those made with other types of analyzers with which you
are probably familiar. This chapter introduces you to the reasons for the
similarities, differences, and increased capabilites of this analyzer as compared to
other types of analyzers.

18 - 1
What Makes this Analyzer Different?

Time Domain and Frequency Domain

Measurements
Measurements made in the time domain are the basis of all measurements in this
analyzer. The time domain display shows a parameter (usually amplitude) versus
time. You are probably familiar with time domain measurements as they appear in
an oscilloscope. Similar measurements may be viewed with the time measurement
data capability.

Frequency-domain displays show a parameter (again, usually amplitude) versus

frequency. A spectrum analyzer takes an analog input signal—a time-domain
signal—and converts it to the frequency domain (this conversion can be done in
several ways; we’ll talk about that a little later). The resulting spectrum
measurement shows the energy of each frequency component at each point along
the frequency spectrum.

Many signals not visible in the time domain (such as noise and distortion products)
are clearly visible in the frequency domain. Because spectrum displays show
frequency components distributed along the frequency axis, it’s possible to view
many different signals at the same time. This is why the spectrum analyzer is such
a useful tool for looking at complex signals—it lets you easily measure (and
compare) the frequency and amplitude of individual components.

Notice the difference between the time-domain and frequency-domain

displays of the same input signal.

18 - 2
 What Makes this Analyzer Different?

The Y-axis (amplitude)

Time-domain measurements are usually viewed with a linear X-axis and a linear
Y-axis (think of an oscilloscope). Frequency-domain measurements are sometimes
viewed with a linear Y-axis and a linear X-axis, but usually must be viewed with a
logarithmic Y-axis, since this is the only way to view very small signals and much
larger signals simultaneously.

Let’s look at the spectrum of a sine wave. Because the amplitude of any harmonic
is small relative to the fundamental frequency, it’s nearly impossible to view a
harmonic on the same display as the fundamental unless the Y-axis scale is
logarithmic. That’s why most measurements made with spectrum analyzers use a
logarithmic amplitude scale—a scale based on decibels. And since the dB scale is
by definition logarithmic, there’s no need to use logarithmically-spaced graticule
lines.

The X-axis (frequency)

Sometimes it’s convenient to use a logarithmic X-axis. Perhaps most familiar to
you is the frequency response measurement. This is traditionally displayed with a
log X-axis (frequency) versus a log Y-axis (relative magnitude).

But most measurements do not require a logarithmic frequency scale. In fact, when
making spectrum measurements it’s easier to characterize harmonics with a linear
X-axis scale since harmonics that are multiples of the same fundamental will
appear at evenly-spaced intervals.

18 - 3
What Makes this Analyzer Different?

What are the Different Types of Spectrum

Analyzers?
There are two broad categories of spectrum analyzers: swept-tuned analyzers and
real-time analyzers. Both swept-tuned analyzers and real-time analyzers have
been around for many years. However, within the past decade or so, spectrum
analyzers have become much more sophisticated. These newer spectrum analyzers
use digital signal processing to provide additional measurement capablity—and let
you interpret measurement results much more easily.

Both swept-tuned and real-time spectrum analyzers display amplitude versus

frequency. How they process and display this information, however, varies with
the specific type of analyzer. A real-time spectrum analyzer displays the energy at
all frequency components simultaneously. A swept-tuned spectrum analyzer
displays measurement results sequentially—in other words, not in ‘’real time.’’ This
is because a swept-tuned analyzer, in effect, uses a single narrow filter that is tuned
across a range of frequencies to produce a spectrum display.

Swept-tuned analyzers have been the traditional choice for higher frequency
applications—for example, 100 kHz and above. Real-time analyzers are generally
used for lower frequencies—for example, audio-frequency and vibration
measurements.

For an additional overview of spectrum/network measurements, see Spectrum and

Network Measurements by Robert A. Witte (Prentice Hall, Englewood Cliffs, New
Jersey, 1993).

Swept-tuned spectrum analyzers

Swept-tuned spectrum analyzers are descended from radio receivers. So it should
come as no surprise that swept-tuned analyzers are either tuned-filter analyzers
(analogous to a TRF radio) or superheterodyne analyzers. In fact, in their simplest
form, you could think of a swept-tuned spectrum analyzer as nothing more than a
frequency-selective voltmeter with a frequency range that’s tuned (swept)
automatically.

Modern swept-tuned analyzers (superheterodyne analyzers, in particular) are

precision devices that can make a wide variety of measurements. However, they
are primarily used to measure steady-state signals since they can’t evaluate all
frequencies in a given span simultaneously. The ability to evaluate all frequencies
simultaneously belongs exclusively to the real-time analyzer.

18 - 4
 What Makes this Analyzer Different?

Real-time spectrum analyzers

Despite the high performance of modern superheterodyne analyzers, they still can’t
evaluate frequencies simultaneously and display an entire frequency spectrum
simultaneously. Thus, they are not real-time analyzers. And the sweep speed of a
swept-tuned analyzer is always limited by the time required for its internal filters to
settle.

Parallel-filter analyzers
Another way to build a spectrum analyzer is to combine several bandpass filters,
each with a different passband frequency. Each filter remains connected to the
input at all times. This type of analyzer is called a parallel-filter analyzer. After an
initial settling time, the parallel-filter analyzer can instantaneously detect and
display all signals within the analyzer’s measurement range.

The particular strength of such an analyzer is its measurement speed—this allows it

to measure transient and time-variant signals. However, the frequency resolution
of a parallel-filter analyzer is much coarser than a typical swept-tuned analyzer.
This is because the resolution is determined by the width of the decimating filters.
To get fine resolution over a large frequency range, you would need many, many
individual filters—thus increasing the cost and complexity of such an analyzer.
This is why all but the simplest parallel-filter analyzers are expensive.

Typically, parallel-filter analyzers have been used in audio-frequency applications.

18 - 5
What Makes this Analyzer Different?

FFT analyzers
FFT spectrum analyzers (also referred to as dynamic signal analyzers) use digital
signal processing to sample the input signal and convert it to the frequency domain.
This conversion is done using the Fast Fourier Transform (FFT). The FFT is an
implementation of the Discrete Fourier Transform, the math algorithm used for
transforming data from the time domain to the frequency domain. This analyzer is
an example of an FFT analyzer that can make real-time measurements.

FFT spectrum analyzers are powerful instruments, since their processing power
can extract more information from an input signal than just the amplitude of
individual frequency components. For example, FFT analyzers can measure both
magnitude and phase, and can also switch easily between the time and frequency
domains. This makes them ideal instruments for the analysis of communication,
ultrasonic, and modulated signals.

If an FFT analyzer samples fast enough, all input data is evaluated and the analyzer
makes a real-time measurement. When operating in real time, FFT analyzers can
make the same measurements traditionally done with parallel-filter analyzers—and
make these measurements, if desired, with far greater frequency resolution.

In the past FFT analyzers have had the disadvantage of their restricted frequency
range—most FFT analyzers could not make measurements above 100 kHz. The
limiting factor has been the speed of the analog-to-digital converter used to sample
the analyzer’s input signal. This is why swept-tuned superheterodyne analyzers are
still used for RF and microwave measurements, though some newer-generation
swept-tuned analyzers, such as the HP/Agilent 8560 family of analyzers can also
make FFT measurements. As you will see, this trend toward hybrid technology has
gone one step further with the Agilent 89400 Series Vector Signal Analyzers.

See the related sidebar for FFT background information.

18 - 6
 What Makes this Analyzer Different?

FFT Background

FFT Basics
The Fourier transform integral converts data from The FFT algorithm works on sampled data in a
the time domain into the frequency domain. special way. Rather than acting on each data
However, this integral assumes the possibility of sample as it is converted by the ADC, the FFT waits
deriving a mathematical description of the until a number of samples (N) have been taken and
waveform to be transformed—but real-world transforms the complete block of data. The
signals are complex and defy description by a sampled data representing the time-domain
simple equation. The Fast Fourier Transform (FFT) waveform is typically called a time record of size-N
algorithm operates on sampled data, and provides samples.
time-to-frequency domain transformations without But the FFT analyzer cannot compute a valid
the need to derive the waveform equation. frequency-domain result until at least one time
The Fast Fourier Transform (FFT) is an record is acquired—this is analogous to the initial
implementation of the Discrete Fourier Transform, settling time in a parallel-filter analyzer. After this
the math algorithm used for transforming data from initial time record is filled, the FFT analyzer is able
the time domain to the frequency domain. Before to determine very rapid changes in the frequency
an analyzer uses the FFT algorithm, it samples the domain. A typical size for N might be 1024
input signal with an analog-to-digital converter (the samples in one time record.
Nyquist sampling theorem states that if samples During the FFT process, the FFT algorithm
are taken twice as fast as the highest frequency transforms the N time domain samples into N/2
component in the signal, the signal can be equally-spaced lines in the frequency domain. Each
reconstructed exactly). This transforms the line contains both amplitude and phase
continuous (analog) signal into a discrete (digital) information—this is why half as many lines are
signal. available in the frequency domain (actually, slightly
Because the input signal is sampled, an exact less than half the number of lines are used, since
representation of this signal is not available in either some data is corrupted by anti-aliasing filters).
the time domain or the frequency domain.
However, by spacing the samples closely, the
analyzer provides an excellent approximation of the
input signal.
FFT Properties
As with the swept-tuned analyzer, the input to the
analyzer is a continuous analog voltage. The
voltage might come directly from an electronic
circuit (for example, a local oscillator) or through a
transducer (for example, when measuring
vibration). Whatever the source of the input signal,
the FFT algorithm requires digital data. Therefore,
the analyzer must convert the analog voltage in to a
digital representation. So the first steps in building
an FFT analyzer are to build a sampler and an
analog-to-digital converter (ADC) in order to create
the digitized stream of samples that feeds the FFT
processor.

18 - 7
What Makes this Analyzer Different?

The Difference
The ideal analyzer would combine the advantages of both swept-tuned and FFT
analyzers while minimizing the disadvantages. To provide these advantages the
89400 series analyzers use FFT technology to provide wideband and high frequency
measurements as well as narrowband and low frequency measurements. The
nearby figure shows the difference between the modern microprocessor-controlled
swept-tuned analyzer and the 89400 series analyzers. The analog IF section is
replaced by a digital IF section which incorporates FFT technology and digital
signal processing to make very fast measurements.

Vector mode and zoom measurements

The 89400 series analyzers expand FFT technology to perform measurements on
narrow as well as relatively wide spans from baseband to the analyzer’s upper
frequency limits. Earlier FFT analyzers have been limited to narrow spans and low
frequencies. These limitations restricted the ability to measure phase and
to analyze time-variant, transient, and modulated signals. These measurements
often require real-time analysis with a non-zero start frequency and wide frequency
spans. Vector mode provides a single FFT measurement as wide as 7MHz (or 10
MHz in 0-10 MHz receiver mode). The zoom capability of Vector mode allows you
to analyze spans away from DC. You may place a selected narrower span at any
frequency within the analyzer’s frequency range.

18 - 8
 What Makes this Analyzer Different?

Stepped FFT measurements in Scalar mode

The 89400 series analyzers use FFT technology to provide enhanced wideband
measurements. Technological advances in designing analog-to-digital converters
and digital signal processors have been combined with FFT technology to provide
the same results as a swept-tuned analyzer but with additional capability. The
nearby diagrams reflect the differences between swept-tuned technology and
stepped FFT technology.

If you are familiar with swept-tuned analyzers for making high frequency
measurements you will find that the technology used in the 89400 series analyzers
has expanded the capability to make FFT measurements with excellent resolution
at higher frequency ranges. This has been accomplished by translating the highest
frequencies to a lower band, then performing FFTs on separate segments of the
spectrum. These segments are displayed contiguously so that the result appears as
it would with a swept-tuned spectrum analyzer.

If you have previously used FFT Dynamic Signal Analyzers, you will discover that
with the 89400 series analyzers you can select spans and resolution bandwidths
which have previously been unavailable in FFT analyzers.

For low resolution bandwidths, the stepped technology in the 89400 series
analyzers is much faster than swept-tuned analyzers. For each local oscillator step
in a swept-tuned analyzer, time is required for the IF filters to settle on each input
signal. With stepped technology, the largest time constraint is the time required to
collect the data and perform the FFT. This results in speeds up to 1000 times faster
with the new stepped technology.

18 - 9
 What Makes this Analyzer Different?

Differences between swept-tuned and stepped analyzer technologies in IF function

Simplified block diagram of a

modern swept-tuned analyzer

Simplified block diagram of the

HP 89400 series analyzer

18 - 10
 What Makes this Analyzer Different?

Signal flow diagram

A more complete block diagram shows the complex local oscillator which allows the
analyzer to use complex time analysis for phase, demodulation, and other time-related
measurements. The 89441A consists of the RF section (bottom box) and IF section (top
box). The IF section is the 89410A.

18 - 11
 What Makes this Analyzer Different?

The difference between swept-tuned and stepped analyzer technology in local oscillator
function

A swept-tuned analyzer
changes the local oscillator
frequency linearly over time
and measures one frequency
point at a time

A stepped analyzer like the HP

89400 series changes the
local oscillator frequency in
sequential steps over time and
measures multiple frequency
points at one time

18 - 12
19

Fundamental Measurement
Interactions

19 - 1
Fundamental Measurement Interactions

Measurement Resolution and Measurement Speed

You should understand the interactions and limits related to measurement
fundamentals in order to optimize your measurements . These fundamentals affect
measurement speed, measurement resolution, and display resolution. If you are
familiar with either swept-tuned or FFT spectrum analyzers, you will find both
similarities and differences in the 89400 series analyzers. In any case, you should
become acquainted with the material in this section in order to be comfortable with
measurements in the 89400 series analyzers.

Resolution bandwidth, frequency span, main length, and window selection are
closely related in determining measurement resolution and measurement speed.
The flexible display resolution also affects your perception of measurement results.

Resolution bandwidth
Resolution bandwidth—often called RBW—determines the analyzer’s frequency
resolution. It may also affect how fast the analyzer makes a measurement.
Normally, resolution bandwidth is adjusted automatically as you select different
frequency spans. Resolution bandwidth is one of the most important parameter
settings in a spectrum analyzer.

Because resolution bandwidth may also affect measurement time, manually

selecting a narrower resolution bandwidth can slow down a measurement more
than necessary. Selecting a resolution bandwidth that is too wide, on the other
hand, may not provide adequate frequency resolution and can obscure spectral
components that are close together.

Narrowing the resolution bandwidth lowers the noise floor because there is less
noise power within the bandwidth of a narrower filter. This occurs because noise
is equally distributed across the frequency spectrum, so the noise floor is lowered
as you progressively restrict the range of frequencies fed to the detector algorithm.

In most analyzers, the final IF filters determine the resolution bandwidth, but in the
89400 series analyzers the time record length and window shape determine the
resolution bandwidth—resolution bandwidth is largely independent of the span.
Resolution bandwidth is related to the number of frequency points and span only
when it approaches minimum and maximum limits. You will learn more about this
later in this chapter.

19 - 2
 Fundamental Measurement Interactions

Video filtering
In some analyzers a low-pass filter is included between the detector and the
display to smooth the noise level—in some cases revealing low-level signals that
might otherwise be obscured. Other analyzers are equipped with a “video
averaging” feature. This lets you average successive traces. Because video filtering
and video averaging both smooth the noise floor, the results of video averaging are
often similar to a single trace with video filtering.

Video averaging with the 89400 series vector signal analyzers is actually a better
approximation of noise than a video filtered-trace, since a series of averaged
measurements will reveal a complete frequency span much faster than the slower
progression of a single, video-filtered trace. However averaging, like video filtering,
does slow down a measurement.

Frequency span
Full-span measurements let you view the entire available frequency spectrum on
one display. With the 89410A, for example, full-span measurements extend from 0
Hz to 10 MHz. Measurements with spans that start at 0 Hz are often called
baseband measurements.

Alternatively, you may wish to view smaller slices of the frequency spectrum. You
can select any number of different spans and position these spans where you want
by specifying their start or center frequencies. This process of viewing smaller
spans is sometimes called zooming or band-selectable analysis. You can control
the frequency span examined by specifying a center frequency and a span size.
Alternatively, you can specify a start and a stop frequency to define a particular
frequency span.

In the 89400 series analyzers the sample rate is adjusted, based on the span, to
achieve the desired information bandwidth.

19 - 3
Fundamental Measurement Interactions

Bandwidth coupling
Bandwidth coupling may be used to link resolution bandwidth and frequency span
in ways which are important to understand when setting up your measurement.

The automatic adjustment of resolution bandwidth to frequency span is called

“bandwidth coupling.” It is an important feature and one common to most
spectrum analyzers.

For most measurement situations, the default bandwidth coupling type (auto)
provides the best compromise between frequency resolution and speed. And for
most measurements, bandwidth coupling is generally preferable since it simplifies
your measurement setup.

Changing bandwidth coupling

You can easily override the current resolution bandwidth selection by manually
entering a setting of your own. For example, you can specify a different resolution
bandwidth setting.

If you override a current resolution bandwidth setting, the analyzer remembers the
adjustment you made in terms of a resolution bandwidth to span ratio. It attempts
to maintain this ratio when calculating appropriate resolution bandwidth for
different spans. This is called offset coupling. For example, if you changed to a
narrower resolution bandwidth than the default (the resolution bandwidth selected
automatically), the analyzer maintains a narrower-than-normal resolution
bandwidth for subsequent spans. You can also specify this type of coupling by
selecting [offset coupled]. This type of coupling allows the response to appear the
same as the span is changed.

The other available coupling type is fixed coupling which maintains a resolution
bandwidth setting independent of other parameter changes. For example, this
allows you to change span without changing resolution bandwidth. Fixed coupling
is automatically selected if you explicitly alter the analyzer’s time record length.

Flexible bandwidth mode

The 89400 series analyzers offers two levels of flexibility in selecting a resolution
bandwidth setting: 1-3-10 and arb. The first, 1-3-10, is the more restrictive. In this
mode the analyzer always sets the resolution bandwidth at a 1 X 10x or 3 X 10x
value. In this mode the analyzer emulates many existing spectrum analyzers. The
arb mode allows arbitrary setting of resolution bandwidth. Your choice of
bandwidth mode in conjunction with your choice of rbw coupling mode affects
both resolution and speed. To get the best resolution to speed tradoff, choose auto
rbw coupling and arb rbw mode.

19 - 4
 Fundamental Measurement Interactions

Display resolution and frequency span

FFT analyzers have a finite record length, usually stated in number of “points,”
“lines,” or “bins.” We will use the term “frequency points.” Most FFT analyzers use
the same number of frequency points regardless of frequency span. However, with
the 89400 series analyzers you may select a variable number of frequency points.
The 89400 series analyzers have a default resolution of 401 frequency points but
you may select from 51 to 3201 points of display resolution. However, for a given
number of frequency points, narrower spans have finer frequency resolution. This
is because the same number of frequency points represents a smaller range of
frequencies.

For each frequency span, the analyzer assigns a discrete frequency value to each
frequency point:
Frequency span
Display resolution =
Number of frequency points−1
The analyzer then uses the specified start frequency to calculate nominal frequency
values for each of the remaining points.

To better understand the concept of display resolution, move the main marker from
display point to display point. Notice how the marker jumps to each point—you
cannot put the marker between points. As you move the marker, also notice how
the marker readout steps through a series of discrete frequencies that corresponds
to each display point.

Display resolution is different from frequency resolution. See “The relationship

between Frequency Resolution and Display Resolution” later in this chapter.

Hint For a given display resolution and window selection, the best measurement
resolution is achieved with the bandwidth mode set to arb and bandwidth coupling
set to auto.

19 - 5
Fundamental Measurement Interactions

Windowing
General
A window is a time-domain weighting function applied to the input signal. A
window is a filter used to compensate for the fact that most signals are not periodic
within the input time record. Depending on the window, the analyzer attenuates
the ends of the input time record, to prevent leakage—a smearing of energy across
the frequency spectrum—caused by transforming signals that are not periodic
within the time record.

FFT analyzers usually have several window types available. Each window offers
particular advantages. Because each window type produces different measurement
results (just how different depends on the characteristics of the input signal and
how you trigger on it), you should carefully select a window type appropriate for
the measurement you’re trying to make.

Windowing is a concept basic to understanding FFT spectrum analyzers. To learn

more, see Spectrum and Network Measurements by Robert A. Witte. Refer also to
online help for a description of each window type and guidelines on choosing a
window type.

Windows used with this analyzer

This analyzer functions as if the input signal were applied to a parallel bank of
narrow-band filters (the number corresponding to the number of frequency points).
The illustrations on the next page show the frequency-domain response of a single
filter when using uniform, Hanning, gaussian top, or flattop windows.

The left side of each illustration represents the center of each filter. Since the
filters are symmetrical, we’ve shown only one side of each filter response (the
other side is a mirror image). The horizontal axis is normalized to 1/ T.

Think of each drawing as a template. If you position a sine frequency at the exact
center of the filter, more of a sine wave’s energy appears in the center bin. Some of
its energy also appears in other bins. The amount of energy that spills into adjacent
bins depends on the type of window you use. Frequency resolution is determined
by the width of the main lobe. The Hanning window provides better frequency
resolution for a fixed time record length while the Flattop window has good
frequency resolution for a fixed resolution bandwidth.

19 - 6
 Fundamental Measurement Interactions

Frequency response of window filters

Uniform Flattop

Gaussian top Hanning

Window Comparison
(for a fixed time record) length)

Uniform

Hanning
Flattop

Gaussian top

19 - 7
Fundamental Measurement Interactions

Enhancing the Measurement Speed

You can make several choices in order to maximize measurement speed:
l Turn calculation off for unused traces under [Measurement Data] [more choices].
l Reduce the number of frequency points to the minimum required for your
measurement in order to reduce internal vector sizes. This applies to the
number of frequency points selected in [Res BW/Window] as well as the maximum
number of frequency points in memory configuration under [System Utilities].
107
l Choose a cardinal span: .
2n
l Turn the marker off.
l If averaging is used, select fast averaging.
l If you have a two channel analyzer turn off any unused channel under the [input]
key.
l If time calibrations are enabled, disable them.

To enhance real-time bandwidth as well as measurement speed, the following

suggestions apply:
l Follow the speed enhancing suggestions above.
l Set [rbw mode] to arb
l Select 1601 frequency points for a single channel measurement, or 801 frequency
points for a two channel measurement.
l Set [rbw coupling] to auto.

19 - 8
 Fundamental Measurement Interactions

Digital storage
All spectrum analyzers require some form of display storage to retain, on a CRT
screen, the relatively slow-moving results of a swept spectrum measurement. Early
spectrum analyzers used CRTs with long-persistence phosphors (or storage meshes
behind the CRT face) to maintain a visible trace throughout an entire frequency
sweep. Modern spectrum analyzers use digital technology to convert the analog
output from an analyzer’s video detector to binary numbers in an internal memory.
These values are then displayed on the analyzer’s CRT screen.

Although digital storage requires a display with a finite number of frequency points,
there are tremendous advantages to digitizing measurement results. Many
functions, such as trace math, were unobtainable with older spectrum analyzers.
Digitizing measurement results also makes it easy to save and recall traces and to
transfer measurement data to other instruments (for example, over the GPIB).

Zero response and DC measurements

What is zero response?

When viewing frequency spans that start at 0 Hz (or very close to 0 Hz), a spectral
line is usually visible at the extreme left of a spectrum analyzer’s display. This is
called zero response or LO feedthrough. In an FFT spectrum analyzer, zero
response is caused by residual dc that originates in the analyzer’s own input
amplifiers. Zero response gives the illusion of a dc offset, even if the input signal
has no dc component—and this occurs even if the analyzer has an ac-coupled input.

In the 89400 series vector signal analyzers, some degree of zero response is always
present in the 0 Hz bin (sometimes called the dc bin). The residual dc that causes
this offset may also leak into the first several bins as well. If you don’t want to see
any zero response on the analyzer’s display, simply start the frequency span several
bins above 0 Hz.

Can spectrum analyzers measure DC?

Most spectrum analyzers are not intended to measure dc. However, analyzers such
as the 89400 series analyzers can measure very low frequencies. This analyzer can,
in fact, measure dc, but not without including a dc offset of its own that can
contribute to (or obscure) a dc offset in the input signal. As we mentioned, this
internal offset is caused by residual dc that originates in the analyzer’s input
amplifiers. Thus, measurement performance at dc is not specified.

19 - 9
Fundamental Measurement Interactions

Special Considerations in Scalar Mode

Sweep time limitations
Sweep time is the time required for the analyzer to complete one full sweep on the
display. With the standard swept-tuned analyzer, sweep time is limited by analog
IF filter dynamics because the filter must have time to respond to the input signal
before accurately measuring a frequency.

In the 89400 series analyzers the digital IF and FFT act like a parallel filter bank so
the sweep speed is limited by data collection and digital processing time rather
than filters. For this reason, the 89400 series analyzers allow measurements for low
resolution bandwidths to run much faster than would be case with a swept
analyzer. The sweep time is adjusted automatically—the analyzer selects the
optimum sweep time, based on the frequency span, the resolution bandwidth, and
the maximum number of frequency points.

Stepped measurements
Traditional swept-tuned analyzers are subject to speed and frequency limits
imposed by the resolution bandwidth filters which determine the information
bandwidth. Traditional FFT analyzers have been subject to fixed resolution
bandwidth-to-span limitations. However, the information bandwidth of the 89400
series analyzers is determined by the span rather than the resolution bandwidth.
Therefore, the resolution bandwidth can be changed without affecting the span.

As discussed previously in the chapter “What Makes this Analyzer Different?”, the
scalar measurement mode uses a stepped local oscillator to perform multiple FFTs
which are then combined into a single trace. This technique, combined with digital
signal processing, allows the analyzer to maximize the sweep time and overcome
the fixed relationship between resolution bandwidth and span. In effect the
minimum resolution bandwidth is limited only by the maximum number of
frequency points and the minimum span.

19 - 10
 Fundamental Measurement Interactions

The relationship between frequency resolution and display resolution

For spectrum measurements, your ability to resolve two closely-spaced
components—that is, the analyzer’s ability to display each component as a separate
frequency point—may be limited by the display resolution. However, the maximum
frequency resolution obtainable is actually determined by the resolution bandwidth
you’ve selected. As you select increasingly narrower spans, the display resolution
improves until the point where you reach the maximum resolution available with
the current resolution bandwidth setting.

Resolution limited by current resolution bandwidth setting:

If the resolution bandwidth is insufficient to reveal two closely-spaced
components, the detector simply sees these two components as a single frequency
and displays them on as a single frequency point. The amplitude of the two
components is combined, though it varies from sweep to sweep as the phase
relationship between the two components changes.

Resolution limited by current display resolution:

If the display resolution is insufficient to reveal two closely-spaced
components—but the current resolution bandwidth setting does provide enough
resolution to distinguish two separate components—the detector uses an algorithm
(peak, normal, or sample) to select a frequency component and displays it as a
single frequency point. However, were you to narrow the span enough or increase
the number of frequency points—keeping the resolution bandwidth constant—you
would be able to view two discrete frequency components as the display resolution
improved.

Resolution bandwidth limitations

The standard swept analyzer uses the last in a series of analog IF filters to
determine resolution bandwidth. In contrast, the 89400 series analyzers use digital
IF and and DSP algorithm to determine resolution bandwidth. For this reason, the
narrowest practical resolution bandwidth in Scalar mode is constrained by the
length of time you are willing to wait for a measurement to be performed.

The widest resolution bandwidth in Scalar mode is 300 kHz for combinations of
span, resolution bandwidth, and maximum number of frequency points which
require that the local oscillator be stepped to more than one frequency. But for
measurements which do not require that the oscillator be stepped the widest
resolution bandwidth is determined by the resolution bandwidth to span ratio:
RBWmax = .3 × span

19 - 11
Fundamental Measurement Interactions

What is a detector and why is one needed

The measurement results that you view on the CRT screen are actually made up of
51 to 3201 evenly-spaced discrete frequency points. For spectrum measurements,
these points represent the range of frequencies that the analyzer evaluates during
each frequency sweep—extending from the currently-selected start frequency to
the currently-selected stop frequency. The analyzer always uses the number of
points you have selected to represent measurement data, regardless of selected
frequency span or measurement type.

During spectrum measurements, the analyzer evaluates all frequencies within its
displayed frequency span but it determines the number of points to display
depending on the number of frequency points you have selected. The interval
between frequency points represents one displayed point. Between two frequency
points the signal may rise and fall a number of times. A detector determines what
signal level to display for each interval.

For example, with narrow resolution bandwidths the stepped FFT measurements
can result in more measurement points than were requested. When this occurs a
data reduction must occur. The process of reducing the data is called detection.
Detection works by combining the results of several FFT bins into a single display
point. In the process of detection some information is lost. For this reason several
detectors are provided, each optimized for a different purpose.

The peak, normal, or sample detector uses its own algorithm to find a suitable
spectral point between the nominal frequencies that define a pair of frequency
points. It transfers the amplitude of this spectral point to a nearby display point.
The analyzer moves the detected point between two frequency points to the
leftmost point for each pair of frequency points. For more information on how
each detector type functions, display the relevant online help topics.

19 - 12
 Fundamental Measurement Interactions

Manual sweep
For swept spectrum measurements, sweep times can be very long for low
resolution bandwidth to span ratios. You can use manual sweep to disable the
analyzer’s automatic sweep. Manual sweep lets you tune the analyzer to a discrete
frequency. This lets you measure the amplitude of discrete frequencies without
waiting for the analyzer to sweep through an entire span—a considerable
advantage when using a narrow resolution bandwidth.

Manual sweep is also useful when making automated measurements. Using manual
sweep dramatically reduces measurement time since it’s much faster to transfer
amplitude data for a single frequency over the GPIB rather than sending data for
the entire display.

19 - 13
Fundamental Measurement Interactions

Special Considerations in Vector Mode

Related parameters
The characteristic time/resolution bandwidth relationship for the 89400 series
analyzers is a natural part of the FFT process and is common to all FFT analyzers,
not just this one. However, the 89400 series analyzer is unique in allowing you to
change span and resolution bandwidth independently.

The arrows indicate inversely related parameters.

Topics later in this chapter discuss these relationships in more detail but the most
important implictions are:
l Resolution bandwidth changes always affect time length for a given window.
l Time length changes always affect resolution bandwidth.
l Span changes always affect ∆t.
l ∆t is related to sample rate and is not user-accessible—it is only affected by
changing span.
l Span and resolution bandwidth may be adjusted independently.

19 - 14
 Fundamental Measurement Interactions

Time data
The 89400 series analyzers provide the ability to view and analyze the time domain
data. Time measurement data provides a time domain view of the input data before
FFT processing. The following points will help you interpret time data:
l Time data is available in Vector and Demodulation modes but not in Scalar
mode.
l Time data may look similar to an oscilloscope display in some cases but you
will see distortion, particularly at high frequencies. The distortion occurs
because the waveform is made up of the discrete samples from a time record
which is optimized for FFT measurements. FFT measurements require a lower
sample rate than would be required for an optimum time domain display.
l The analyzer must be in baseband mode to display data similar to an
oscilloscope. In other words the measurement must start at 0 Hz and the
center frequency must not be set explicitly. If you enter zoom mode either by
setting the center frequency explicitly or by pressing the zoom/baseband key, a
notice will flash on the screen which indicates that time data is in zoom mode.
In zoom (band-selectable) measurements the time data is displayed relative to
the center (local oscillator) frequency rather than relative to DC. Therefore,
zoom time data has a different appearance because baseband data is real while
zoom data is complex. Because the local oscillator is tuned to the center
frequency for zoom measurements, the time data is represented as the
difference frequency between the local oscillator and the input signal
frequencies.
l Time and time exponential averaging are the only types of averaging which
affect time data displays. Other average types only display the most recent
time record.
l In baseband mode with maximum span, some signals, particularly square
waves and transients, may appear to have excess distortion or ringing because
of the abrupt frequency cut-off of the anti-alias filters. In this case the time
data display can be improved by turning off the anti-alias filter under the [Input]
menu. Make sure you turn the filter back on before viewing frequency domain
data.

19 - 15
Fundamental Measurement Interactions

The time record

A time record is the amount of time-domain data the analyzer needs to perform one
FFT operation. The time record and its FFT are the building blocks the analyzer
needs for all subsequent measurements.

Why is a time record needed?

Essentially, the time record is a block of time-domain sample points. Since the
actual Fourier Transform does not have explicit time or frequency references (it
simply operates on a sequential collection of points), FFT analyzers must assign
arbitrary start and finish times for data to be transformed. These blocks of input
data are called time records.

For example, with the default display resolution of 401 frequency points, the 89400
series vector signal analyzer takes up to 1024 samples of time data to produce 512
points of frequency domain data. The analyzer usually displays the first 401 points
of this data and discards the rest (this accommodates the anti-aliasing filters, but
that’s beyond the scope of our current discussion).

The time record can be described by both a length and a size. The time record
length is the amount of time required to acquire a time record and is altered by
changing resolution bandwidth, window, main length, or gate length. The time
record size is the number of time points in the time record and is dictated by the
time record length in combination with the sample rate (and sample rate, in turn, is
directly related to span). More detail on these interactions can be found later in
this chapter.

Time record size refers to the number of time points (samples) in a time record. Time
record length refers to the amount of time needed to acquire a time record.

19 - 16
 Fundamental Measurement Interactions

Time record, span and resolution bandwidth

The 89400 series analyzers eliminate a restriction inherent in previous FFT
analyzers by allowing you to choose span and resolution bandwidth independent of
one another. This implies a different method of handling the time record before the
FFT.

The 89400 series analyzers take a number of samples (time points) directly related
to span. These samples are acquired in a time length inversely related to the
resolution bandwidth. The FFT converts this time record to the frequency domain
and displays the result across the number of frequency points (display points)
selected by the user.

Measurement speed and time record length

The length of a time record is inversely proportional to the resolution bandwidth.
For smaller resolution bandwidths, an FFT analyzer needs a longer time record and
therefore takes longer to make a measurement. For larger resolution bandwidths,
an FFT analyzer needs a shorter time record and can therefore make a
measurement much faster. The different time length required will become
noticeable as you start making measurements. Refer to the earlier discussion:
“Enhancing the Measurement Speed.”

19 - 17
Fundamental Measurement Interactions

How do the parameters interact?

Window bandwidth
In most analyzers the final IF filter determines the resolution bandwidth. In the
89400 series analyzers the window type you select shapes the resolution bandwidth
filter shape. And the window type, along with the time record length, determines
the width of the resolution bandwidth filter. Therefore, for a given window type, a
change in resolution bandwidth will directly affect the time record length.
Conversely, any change to time record length will cause a change in resolution
bandwidth.

Window Bandwidth
Flattop 3.8193596
Gaussian top 2.21234968
Hanning 1.5
Uniform 1.0

Important relationships
Window bandwidth represents the noise equivalent bandwidth of a window’s filter
shape and affects the resolution bandwidth and time record length:
WBW
T=
RBW
Some useful equations help explain the relationship between time record length
and time record size:
SR= 2.56 × span (in baseband mode)

SR= 1.28 × span (in zoom mode)

TP = SR × T

1
∆T =
SR

where:
RBW = resolution bandwidth (Hz)

WBW = window bandwidth

T = time record length (seconds)—refers to main length or gate length

TP = time record size (number of time points)

SR = sample rate (frequency)

19 - 18
 Fundamental Measurement Interactions

Time record length limitations

The maximum time record size (the number of points of time data) is limited by the
number of frequency points. As noted earlier, the sample rate (the rate at which
data is digitized and entered into the time record) is determined by span.
Therefore, the maximum time record length for a given span is determined by the
number of frequency points. Because of this we can see that for a given window
type, the minimum resolution bandwidth, while not directly affected by span and
number of frequency points, may be limited by these parameters.

Time record size (also called block size or number of time points) is determined by
window, span, and resolution bandwidth. But maximum and minimum time record
sizes are determined by memory and FFT constraints:

Maximum time record length

The time record length is limited by the number of frequency points.
(FP−1)
Tmax=
span
Where:
FP = number of frequency points

Maximum time record size

The time record size is also limited by the number of frequency points.
TPmax = 2.56 × (FP−1) (In baseband mode, real data)

TPmax = 1.28 × (FP−1) (In zoom mode, complex data)

Therefore:
Tmax = TPmax × ∆T

Minimum time record length

Minimum time record length is dependant on the resolution bandwidth to span
ratio:
WBW
Tmin =
RBWmax

RBWmax = .3 × span

Time record length and memory

To increase the time record length you can increase the number of frequency
points until you reach the maximum number of frequency points allocated in
memory (up to a maximum of 3201). To increase time record length you may have
to reallocate memory space under the [System Utility] key.

19 - 19
Fundamental Measurement Interactions

Time record processing

The relationship between a time record and the frequency data is relatively straight
forward if the minimum resolution bandwidth (and therefore the maximum time
record length) is used. In this case the time record size is equal to the time record
size required by the FFT (which is dictated by the number of frequency points and
must be a power of 2n). Therefore, the FFT converts the samples in the time record
directly to display points.

If a higher resolution bandwidth is selected, the time length necessary to acquire

the time samples is shorter and fewer samples are required. However, the FFT
must still operate on the same time record length for display purposes, so the extra
time record length is set to zero (zero padded) before the FFT process. This
enables the FFT to properly convert the time record to the required number of
display points without adversely affecting the frequency data.

The time record calculations

The length of the time record determines how long a given measurement will take
and the maximum frequency resolution you can measure. As described above the
FFT time record size may differ from the size dictated by the resolution bandwidth.
The FFT record size must be a power of 2 equal to or greater than the number of
frequency points.

TP = INTEGER 
T
∆T
 

TP (FFT) = 2n ≥ FP for zoom mode

TP (FFT) = 2n ≥ (FP × 2) for baseband mode

19 - 20
 Fundamental Measurement Interactions

The time record and zero padding

At the minimum resolution bandwidth in zoom mode, the time length necessary to collect
the time samples is equal to the time record length required by the FFT.

At any other resolution bandwidth, the number of frequency points remains the same, but
the time required to collect the time samples is shorter than the time record required by
the FFT. Therefore, the time record is extended by the addition of zeros.

19 - 21
20

Analog Demodulation
Concepts

20 - 1
Analog Demodulation Concepts

What is Analog Demodulation?

Analog demodulation demodulates amplitude modulated (AM), frequency
modulated (FM), and phase modulated (PM) signals.

Applications
AM, PM, and FM signals commonly consist of a high-frequency carrier signal
modulated by a lower-frequency audio, video, or data signal. For these signals,
analog demodulation removes the carrier frequency leaving the modulating signal
which is contained in the sidebands of the carrier signal. An example of this
application is included in “Demodulating an Analog Signal” earlier in this book.

The most obvious use for analog demodulation is analyzing intentionally modulated
signals as noted above. However, analog demodulation is also useful in analyzing
signals which are not intentionally modulated. For example, you may choose to:
l Apply phase and frequency demodulation to oscillator or synthesizer designs
to examine phase and frequency trajectories during frequency hops or to
establish phase-lock-loop lock-up time. For an example of this type of
application see “Characterizing a Transient Signal” earlier in this book.
l Demodulate a burst signal to evaluate wait time to determine when a signal
stabilizes.
l Amplitude and phase demodulate sidebands to determine the type of
modulation present in phase noise.
l Demodulate bearing noise in mechanical measurements to examine
frequency instability.

Using analog demodulation for zero span measurements

If you are accustomed to making zero span measurements, you can use AM
demodulation to make similar measurements. The information bandwidth of other
analyzers is limited by the resolution bandwidth filters, but in 89400 series
analyzers the information bandwidth is only limited by span. Therefore, AM
demodulated signals appear similar to zero span measurements with a wider choice
of information bandwidth.

20 - 2
 Analog Demodulation Concepts

How Does Analog Demodulation Work

in the Agilent 89400 Series Analyzer?
Analog demodulation alters the time record by translating the complex time data to
real (baseband) data. In the flow diagram below you can see the location of analog
demodulation in the measurement process. The analyzer performs all subsequent
measurement processes on demodulated data after this point. When you turn
demodulation off (press [Instrument Mode], [demodulation setup], [chx result], [demod off]), the
results are the same as Vector mode.

20 - 3
Analog Demodulation Concepts

Special Considerations for Analog Demodulation

Analog demodulation requires certain conditions and has certain constraints. For
this reason some default conditions for analog demodulation differ from other
instrument modes.
l The analyzer automatically selects zoom mode if you select analog
demodulation. Analog demodulation requires zoom mode in order to identify
the carrier. The center frequency should be set as close to the carrier
frequency as possible. The center frequency is programmed into the digital
local oscillator, and represents the initial estimate of the carrier frequency.
l Analog demodulation requires a positive start frequency.
l The visible demodulated span contains only half the number of points that
were contained in the original signal—the complex waveform is transformed
into a real waveform which cuts spectral information in half. Analog
demodulation operates only on double sideband signals. Therefore, the
information in one sideband is redundant and the analyzer mixes the signal
down to baseband (zero Hertz start), eliminates the carrier, and eliminates the
lower sideband.
l Selecting analog demodulation automatically enables time corrections. Time
corrections correct for amplitude and phase distortion caused by the anti-alias
and digital filters prior to demodulation. If you choose not to use time
correction, you must turn it off before selecting a demodulation type (AM, FM
or PM). The next section discusses the role of time correction in demodulated
measurements in more detail.
l The time record is altered by analog demodulation resulting in time domain
displays of amplitude versus time (AM demodulation), frequency versus time
(FM demodulation), and phase versus time (PM demodulation).
l The demodulator alters the time record, and the time record is the basis for
all other measured results. Therefore, all other measured results are computed
using the demodulated waveform. For example, the spectrum is the spectrum
of the modulating signal.
l Only one demodulator per channel can be selected at a time.
l The span must be set to include the carrier and all of its sidebands. Setting a
span which is too narrow will cause distortion.
l The bandwidth of the demodulator is approximately equal to the span.

20 - 4
 Analog Demodulation Concepts

Time Correction and Analog Demodulation

Time correction applies the correction factors which are derived during calibration.
Time calibrations are used to correct for magnitude and phase distortion in the
time record prior to the demodulation process. Left uncorrected these distortions,
which are introduced by the analog and digital filters, create errors in the
demodulated signal and also introduce cross modulation products such as
PM-to-AM conversions.

Time calibrations only correct for distortion over the selected span. Sidebands
which fall outside of the displayed span, but have not been completely attenuated
by the digital filter, are subject to severe distortion and can create errors in the
measurement. For this reason it is important that the signal reside completely
within the span and that all unwanted signals be well outside of the span. See
“Checking for interfering signals” later in this chapter.

Time correction is automatically enabled when you enter Analog Demodulation

mode, and is disabled when you return to Vector mode.

20 - 5
Analog Demodulation Concepts

The Importance of Span Selection

Including all important signal data
Analog demodulation is performed on the displayed span. Therefore, it is very
important to select your span carefully for the best results. If your display
truncates your signal’s sidelobes, distortion will result as you can see in the sample
below.

The selected span truncates some sidebands of this phase modulated signal,
resulting in distorted demodulation.

The selected span includes all the sidebands, resulting in clean demodulation.

20 - 6
 Analog Demodulation Concepts

Checking for interfering signals

Signals slightly outside the displayed span may interfere with demodulation. Due
to filter roll-off, the selected span may be influenced, during demodulation, by
signals within the guardband. You should check for such signals with a wider span
(1.28 times the desired span) prior to selecting your measurement span for a
demodulation measurement. The example below illustrates the results of
demodulating a signal with an interfering signal nearby.

The span shown to the left appears ideal, but the span to the right shows that
another signal (this one FM) appears in the guard band outside the desired span

The display to the left shows interference caused by demodulating 40 kHz span above,
which is affected by the signal present in the guard band. The demodulated signal to
the right shows the result if the interfering signal did not exist or existed outside the
guard band.

20 - 7
Analog Demodulation Concepts

The Importance of Carrier Identification

Accurate angle demodulation depends on precisely identifying the carrier
frequency. The demodulation algorithm for each demodulation type (AM, FM and
PM) assumes that the carrier is at the center frequency. The auto carrier feature
examines phase ramping to estimate the frequency error between the carrier and
the analyzer’s local oscillator. Auto carrier implementation varies depending on the
selected demodulator.

Auto carrier with AM demodulation

AM demodulation is the same with or without auto carrier.

Auto carrier with PM demodulation

The analyzer removes frequency error and phase offset between its center
frequency and the detected carrier frequency. The carrier estimate is based on data
within a single time record and is updated on a record-by-record basis. This
method of carrier identification does not involve carrier locking with a
phase-locked loop.

Auto carrier with FM demodulation

FM is obtained by differentiating PM, so the auto carrier calculation, in removing
the phase ramp, eliminates frequency offsets in FM as well.

Special considerations for auto carrier use

Turning off auto carrier may yield the results you want if:
l You want to evaluate frequency or phase variations above or below the
carrier frequency. Examples of this situation would occur in examining a
hopping carrier, transients, or oscillations varying around 0 degrees. In these
cases you must turn off auto carrier to make the instability visible.
l Your signal has a suppressed carrier.
l You have locked the analyzer and device-under-test to a common reference
and carefully set the center frequency.
l Your signal’s sidelobes are truncated.
l Your signal contains low frequency modulation (fewer than 5 cycles over the
time record) or transients.
l Your carrier is not stationary. In this case auto carrier may cause phase
wrapping.

20 - 8
 Analog Demodulation Concepts

AM Demodulation Specifics
The algorithm
Envelope = √

real 2 + imag 2

_______________
AM =

√real 2 + imag 2 −Carr Amplitude
_______________
Carr Amplitude

AM = modulation
where _______________ index
Carr Amplitude = carrier amplitude averaged across the entire time record.

AM demodulation averages the carrier amplitude to derive the modulation index from the
envelope

20 - 9
Analog Demodulation Concepts

PM Demodulation Specifics
The algorithm

Auto carrier off

PM = tan−1
imag 

 real 
Auto carrier on
PM = tan−1
imag 

 real 

If you select both phase and frequency auto carrier type the frequency offset of the
carrier from the local oscillator is estimated and removed:
___________
tan−1
imag 
=0
 real 

20 - 10
 Analog Demodulation Concepts

Detecting the carrier phase and frequency

Example of a carrier PM modulated by a sine wave

20 - 11
Analog Demodulation Concepts

FM Demodulation Specifics
The algorithm

Auto carrier off

FM is the derivative of PM.

Auto carrier on
Since FM is the derivative of PM, phase offset differentiates to 0 and auto carrier
removes only the frequency error, ω∆.

Example of a carrier FM modulated by a sine wave

20 - 12
 Analog Demodulation Concepts

Interactions with other features

Choosing trigger type with analog demodulation
The analog demodulator is located after the triggering circuits so it is not possible
to trigger on the demodulated signal. However, if the signal is AM modulated or
pulsed, IF trigger can be used to obtain a stationary time waveform.

External triggering (or triggering on the second input channel) can be used if the
original modulating signal is available.

Using gating and averaging with analog demodulation

When using analog demodulation, features such as gating, band power markers,
and averaging work as described for vector mode except that they work on the
demodulated signal, not the original signal. When demodulating, the entire time
record is demodulated.

Two-channel measurements and analog demodulation

Analog demodulation is performed on a channel-by-channel basis. This
architecture lets you independently select the type of demodulation, and, more
importantly, lets you demodulate one channel while measuring baseband signals on
the other. For example, you can use a PM demodulator to measure unintentional
phase modulation of an oscillator on channel one. At the same time you can
measure the signal thought to be causing the modulation by using channel two for a
baseband measurement. You can compare the two measurements visually by
observing their respective time and spectrum displays, or you can compare them
mathematically by using the coherence, cross-spectrum, frequency response, or
cross-correlation functions. You can use this two-channel capability to perform
tasks such as: tracking down the sources of phase-noise in an oscillator, or
computing the frequency response (magnitude and phase) of a modulator or
demodulator.

There are special considerations for using analog demodulation with two channels:
l Both channels share the same center frequency when demodulators are
enabled for both channels.
l When one channel has a demodulator enabled and the second does not, the
second channel makes baseband measurements with a span which is half of the
selected measurement span. This span matches the span of the demodulated
spectrum of the first channel.

20 - 13
21

Gating Concepts

21 - 1
 Gating Concepts

What is Time Gating?

Time gating is only available in Vector, Analog Demodulation, and Wideband CDMA
instrument modes.

Note This chapter describes time gating for Vector and Analog Demodulation. To learn
how time gating works for Wideband CDMA, see the Wideband CDMA Concepts
chapter.

Gating allows the instrument to isolate a portion of a time record for further
viewing and analysis. Time gating is often used to analyze non-stationary signals or
portions of stationary signals such as burst signals from devices such as videotape
recorders, computer disk drives, TDMA commmunication bursts, or ultrasonic
tranducers.

Time gating lets you see a large time domain signal, only part of which you want to
analyze. From the larger displayed signal you can easily isolate a portion for
further analysis.

You can see an example of how to set up a basic burst measurement in “Using
Gating to Characterize a Burst Signal” in the Getting Started Guide. You can see
examples of how to use gating for specific types of measurements in “Making
On/Off Ratio Measurements” and “Characterizing a Transient Signal” in this book.
These examples use baseband analysis, but gating works equally well for zoomed
measurements.

For details on the use of the various keys related to time gating, display Online Help
for the specific keys.

21 - 2
 Gating Concepts

Time gating isolates a portion of a time domain signal.

In this example the gated portion contains two of the four spectral
components included within the total (main) test signal.
Note that gating has minimized leakage, resulting in more
distinct signals in the frequency domain.

21 - 3
Gating Concepts

How Does it Work?

The analyzer collects a record of data the duration of which is defined by main
length, and the position of which, relative to the trigger, is defined by [main delay].
After the main time data has been acquired, the analyzer isolates a portion of the
main data as the gate record. The duration and position (relative to the start of the
main record) of the gate are determined by [gate length] and [gate delay ] respectively.

The gate record is then processed as the analyzer’s time record. That is, spectrum,
frequency response, coherence, and correlation displays are based on the gate
record.

To control time gating, you define the following parameters:

l Trigger setup (if any)
l [main delay] (this corresponds to trigger delay)
l [main length]
l [gate delay]
l [gate length]

After you have acquired main time data, you may reposition or resize the gate
without obtaining new main time data. The analyzer will automatically recompute
and redisplay the data contained in the modified gate record.

21 - 4
 Gating Concepts

Important Concepts
l Main delay and gate delays are independant for each channel—you can select
different delay times for both main and gate regions for each channel—though
lengths must be the same for both channels.
l You can smoothly scroll with the knob to change data position and length.
l You can use averaging on gated measurements. Rms and peak averaging are
applied to the spectrum display of the gate, while time averaging is applied to
the main time record.
l You can step the gate delay through the main time data in predetermined
increments by setting [step size] to the desired increment, then selecting [gate delay]
and using the up and down arrow keys.
l In some applications the overhead required to acquire a new main record may
be undesirable. In this case you may expand the gate region to cover the entire
display with the [gate->main] key. You may then continue to analyze the previous
gate region without collecting a large main record.
l The maximum length of the main record is determined by the span and the
number of frequency points. When working in the time domain you may not
want to be limited by the selected span. In this case you would like the span to
be adjusted to meet your main record length needs. To do this you can set [time
resltn] to [auto] then you may adjust main length while span is automatically
adjusted to keep the number of points in the main record within the limits of
the maximum number of frequency points in memory.
l You can increase the main record length by increasing the number of
frequency points. You will then be able to select a narrower resolution
bandwidth, longer time length, or wider frequency span.
l Resolution bandwidth cannot be independantly adjusted in time gating
because the resolution bandwidth is determined by the length of the time
record. Since you set the length of the time record in gated measurements, you
cannot set resolution bandwidth.
l If you are using gating to estimate the frequency or phase of a signal at a
specific point in time you may want to consider using demodulation instead.
With demodulation you can measure the instantaneous frequency or phase at a
point in time, rather the estimate over the gate interval.

21 - 5
Gating Concepts

Parameter Interactions
There are some inherent parameter interactions:

The arrows indicate inversely related parameters.

l When the gate is off, rbw and time length refer to main length. When the gate
is on these parameters refer to gate length—when you change gate length you
change the resolution bandwidth. This also implies that you cannot explicitly
set resolution bandwidth without changing the gate length.
l When the gate or main length are explicitly set, rbw coupling is set to fixed,
and rbw mode is set to arbitrary. This allows time length to be unrestricted and
can therefore be set to whatever you enter.
l When rms or peak averaging is on, changing main or gate delay or length will
abort any averaging in progress and a new one will start. Changing the gate
when time average is on will not abort the average.

21 - 6
22

Digital Demodulation
Concepts (Opt. AYA)

22 - 1
Digital Demodulation Concepts (Opt. AYA)

Overview
What you learn in this chapter
This chapter contains an overview of digital demodulation (option AYA) and
presents concepts on how the analyzer demodulates digitally modulated signals.
For examples on setting up a digital demodulation measurement and analyzing
demodulated data, see the chapters titled “Using Digital Demodulation” and
“Analyzing Digitally Demodulated Signals’’. For examples on creating an ideal
digitally modulated signal and creating a user-defined filter, see the chapter titled
‘’Creating User-defined Signals’’. For key-specific information, refer to online Help,
which is accessed by pressing the [Help] key on the analyzer.

If you need background references

The information covered in this chapter assumes a good basic knowledge of digital
demodulation. If you need general information on the subject of digital
demodulation ask your sales representative about the following Agilent
Technologies publications:
l Application Note 355—Digital Radio Theory and Measurements (Part Number
5091-4777E)
l Application Note 1298—Digital Modulation in Communication Systems-An
Introduction (Part Number 5965-7160E)
l 89400 Product Note 8—Vector Modulation Analysis (Part Number 5091-8687E)
l 89400 Product Note 14—Using Error Vector Magnitude Measurements to
Analyze and Troubleshoot Vector-Modulated Signals (Part Number 5965-2898E)
A number of good text books on the subject of digital communication are available,
for example:
l Edward A. Lee and Daniel G. Messerschmitt, Digital Communication, Kluwer
Academic Publishers, Boston, 1992

What this analyzer does

This analyzer provides various analysis techniques for several standard and
non-standard digital modulation formats. Analysis capabilities include several
display formats as well as numeric error measurements. Flexible scaling and
markers enhance these capabilities.

This analyzer requires no external filters, coherent carriers, or symbol clock timing
signals. It allows you to demodulate pulsed or continuous carriers and it locks to
the carrier and to a defined symbol rate. The analyzer uses the demodulated signal
to generate an ideal reference signal. The ideal reference signal is compared to the
demodulated signal to provide a quantitative measure of system errors. It has
built-in filters which may be applied to the measured signal as well as to the
reference signal. This allows you maximum flexibility in comparing your signal to
an ideal signal. Additionally, this allows complete flexibility to probe any analog
point in a communication system. An optional second baseband channel allows IQ
baseband measurement capability.

22 - 2
 Digital Demodulation Concepts (Opt. AYA)

Measurement Flow
General block diagram
The following block diagram illustrates the location of digital demodulation in the
analyzer’s measurement path.

22 - 3
Digital Demodulation Concepts (Opt. AYA)

Digital Demodulator Block diagram (except FSK)

The following is the block diagram for the analyzer’s digital demodulator (the
Digital Demod block shown in the general block diagram). The block diagram is
for all demodulation formats except FSK.

22 - 4
 Digital Demodulation Concepts (Opt. AYA)

Digital Demodulator Block diagram: FSK

The following is the block diagram for the analyzer’s digital demodulator (the
Digital Demod block shown in the general block diagram) when FSK is selected.

22 - 5
Digital Demodulation Concepts (Opt. AYA)

Another way of viewing the digital demodulation process is illustrated in the

following table:

Measurement Path for Digital Demodulation

Parameters
Time Sequence Effect Result Notes
Involved
Signal setup Range Sets up basic signal Raw data Any changes made
Trigger parameters required in the signal setup
Frequency/span for digital parameters later in
Input demodulation the time sequence
Time Corrections require that a new
Window measurement be
Memory performed
Pre-demodulation Pulse search Determines what Input to This signal is the
Search length signal enters the demodulator maximum possible
Result length demodulator by length, limited only
trimming the raw by memory
data constraints.
Demodulation Setup parameters Determines the Demodulated, The result length
Format type of symbol-locked data derived previously
Symbol rate demodulation in the default may be shortened if
Filter performed on the display format desired
Result length input time data
Sync search
Points/symbol
Clock adjust
Post-demodulation Measurement data Formats the data Formatted, Changes may be
Display format for display demodulated data made in the display
Scale parameters without
Marker acquiring a new
Normalize measurement

22 - 6
 Digital Demodulation Concepts (Opt. AYA)

Measurement management
Measurement and display choices
The flexibility of this analyzer provides numerous possible ways of viewing digitally
demodulated signals. You may demodulate signals of various types then view
aspects of those signals in several ways. Modulation format, measurement data,
and data format may be combined in different ways for specific measurement
needs.

Carrier locking

22 - 7
Digital Demodulation Concepts (Opt. AYA)

The analyzer’s center frequency must be close to the transmitted carrier frequency
to achieve carrier lock. The required proximity of the center frequency to the
carrier frequency varies depending on the signal type, symbol rate, and system
noise. If symbol locking appears poor, you may achieve better carrier locking by
observing the “Freq Err” value in the symbol table of any successfully locked
measurement and adding that amount to the center frequency.

To obtain reliable carrier locking, the difference between the analyzer’s center
frequency and carrier frequency should be within 3% of the symbol rate, as follows:

(center frequency) − (carrier frequency) ≤ ± .03 (symbol rate)

For QAM signals, the result length also affects the analyzer’s ability to achieve
reliable carrier lock, as shown in the following table. Result lengths less than those
shown may result in unreliable carrier lock. Optimal carrier locking occurs when
the result length is 1000 or greater.

The minimum result length

If the modulation type is:
for reliable carrier lock is:
16 QAM 50 symbols
32 QAM 75 symbols

The following may also affect carrier locking:

l A frequency span that is too narrow.

l An incorrect range setting.
l Using the wrong measured or reference filter.

22 - 8
 Digital Demodulation Concepts (Opt. AYA)

I-Q measured signal

The IQ measured signal is the result of resampling the data to an integer number of
points per symbol and applying IQ origin offset compensation (FSK and MSK type 1
do not compensate for IQ origin offset or droop), system gain normalization, carrier
locking, and filtering to the incoming signal. The filtering is user-selectable.

I-Q reference signal

A powerful analysis technique involves comparing a demodulated signal with an
ideal signal generated from detected bits. The analyzer detects bits from the
measured IQ signal and reconstructs a sequence of ideal I and Q states. These are
then treated as ideal impulses and are baseband filtered according to a reference
filter selected by the user. The resultant IQ reference can be overlaid or compared
side-by-side with the IQ measured signal.

The reference filter can be selected independently from the measured signal
(although the alpha/BT is the same for both filters). Therefore, you can apply
different filters to the measured and reference signals to accommodate special test
requirements. See the section on filtering later in this chapter.

Special considerations for FSK demodulation

Separate I and Q signals do not exist for FSK demodulated signals. There is no
phase information (the imaginary data is zero) because FSK demodulation
performs FM demodulation on the input signal. Consequently, the FSK
demodulator produces real rather than complex data. The display result is a
baseband, single-sided spectrum and a time display representing frequency versus
time.

22 - 9
 Digital Demodulation Concepts (Opt. AYA)

Parameter interactions
Changing one parameter may necessitate changes in other parameters to achieve
the desired measurement. The following topics may help you optimize your
measurements and explain the reasons for changes made automatically by the
analyzer.

Span considerations
When selecting a frequency span, select the narrowest span that includes all of your
signal components (select a span that is larger than the bandwidth of your signal).
If you select a span that is too narrow, your measurement may have excessive
errors or the analyzer may lose carrier lock. If you select a span that is too wide,
your measurement may be affected by excessive noise and slower speed. You may
want to select the Vector instrument mode and set the span while viewing the
spectrum of your signal.

The analyzer displays a warning message if the frequency span is less than the
symbol rate. If you see this warning message, increase the frequency span until it
includes all components of your signal.

When using digital demodulation, the symbol rate determines the maximum
frequency span that you can select, as follows:

(SymbolRate)
Maximum Span = 20
1.28

This formula can be rewritten to obtain the maximum span-to-symbol-rate ratio.

For FSK demodulation formats, you can increase this ratio:++

Maximum Span 20 Maximum Span

= OR = 15.625
Symbol Rate 1.28 Symbol Rate

++ The FSK demodulation format presents an exception because the value [max span/symR] in the [configure meas memory] menu
can be increased in value up to 100 if sufficient memory is available. This accommodates the wider spans often required by FSK
demodulation.

22 - 10
 Digital Demodulation Concepts (Opt. AYA)

You may notice that the frequency span for digitally demodulated spectrums
(the IQ measured spectrum, IQ reference spectrum, and error vector spectrum) is
different than that set with the [span] softkey. The reason is that the digital
demodulation process derives its own frequency span and sample rate, as follows:

Demod Sample Rate = (Symbol Rate) × (points−per−symbol)

$ E Demod~Span~=~(Demod~Sample~Rate)~/~1.28

When viewing digitally demodulated spectrums, aliasing may occur if the frequency
span is too large. Aliasing only affects the validity of the spectral displays, not the
accuracy of the demodulation. The analyzer displays DATA? if it suspects aliasing
in digitally demodulated spectrums. To prevent aliasing, be sure that:

(Symbol Rate) × (points−per−symbol)

Frequency Span ≤
1.28

where: Frequency span is set by pressing [Frequency] [span].

symbol rate is set by pressing [Instrument Mode], [demodulation setup], [symbol rate].
points-per-symbol is set by pressing [Time], [points/symbol].

Data size considerations

Maximum data size for measured and reference IQ data is generally determined by
max time pts++ which is user selectable, 64-2048 (up to 4096 samples with
extended RAM option UFG or UTH). A given measurement will generally use all
available memory as defined by max time pts:

max time pts≥ result length × points ⁄ symbol

Resolution bandwidth
Resolution bandwidth in digital demodulation is determined automatically by the
analyzer, and cannot be set independently. Resolution bandwidth is determined by
the time record length in the same manner as in Vector mode (see Fundamental
Measurement Interactions):

RBW = WBW ⁄ T

where RBW = resolution bandwidth

T = time record length
WBW = window bandwidth

However, in Digital Demodulation the time record length is determined as follows:

T ≈ Result Length (in symbols) ⁄ Symbol Rate

Display limitations

++ See [max time points] under [System Utility], [memory usage], [configure meas memory]

22 - 11
 Digital Demodulation Concepts (Opt. AYA)

Points per symbol affects all displayed results by controlling result resolution but is
not coupled to span. More points per symbol improves the resolution of the vector
diagram, but there is still only one point at the symbol clock.

# time points = result length × points ⁄ symbol

Result length takes precedence over points per symbol—if you try to set points per
symbol such that the number of time points would exceed the memory size++, an
error indicates that the requested number of points per symbol cannot be set. In
this case you may choose to reduce the result length to allow an increased number
of points per symbol. Conversely, if you increase result length such that the
number of time points exceeds the memory limit, points per symbol automatically
decreases to allow the result length to increase.

Feature Availability in Digital Demod

Most features which are available with other instrument modes are also available
with digital demodulation with the following exceptions:
l Time gating is not available
l Averaging is not available for any trace data in digital demod but may be
applied to numeric error data in the symbol table. Three types of averaging are
available:
– rms (video)—computes an rms average of each rms error update in
the table

– rms exponential—is like rms (video) except that averaging continues past
the average count with an exponential weighting

– continuous peak hold—keeps track of the peak rms error of each update
in the table

++ See [max time points] under [System Utility]

22 - 12
 Digital Demodulation Concepts (Opt. AYA)

Special considerations for sync search

Synchronization words (or patterns) are often used to resolve carrier phase
ambiguity on non-differential modulation formats. It is important to realize that
synchronization words are optional and are not necessary in order to achieve
carrier locking.

Sync search lets you use a synchronization pattern to isolate a portion of your
signal for display and analysis. The analyzer searches through demodulated data to
find your sync pattern, and then uses the [result length] to determine how much data to
display, and the [offset] to display data relative to the sync pattern.

Note The sync pattern must be a multiple of the number of bits-per-symbol. For
example, if the number of bits-per-symbol is 4 (as with 16 QAM), the number of bits
in the sync pattern must be a multiple of four. Sync search lets you specify any
number of bits for the sync pattern, however, bits that aren’t a multiple of the
bits-per-symbol are truncated. In this example, if you entered 6 bits for the sync
pattern, sync search would only use the first four bits.

Triggering determines when the analyzer starts demodulating data and search
length determines when the analyzer stops demodulating data. Sync search locates
only the first match and ignores any subsequent matches within the search length.
The sync pattern and the offset must fit within the search length.

You may use sync search with or without pulse search, although many
measurements require both features.

See “Using Digital Demodulation” earlier in the book for an example of using sync
search. See online help for the following softkeys for additional information (all
softkeys are under the [Time] hardkey):
l [sync search]
l [search length]
l [sync setup] → [pattern]
l [sync setup] → [offset]

22 - 13
Digital Demodulation Concepts (Opt. AYA)

Special considerations for pulsed signals

Pulsed transmissions are generated by on/off (or burst) carriers such as those used
in mobile units. For this type of signal the analyzer searches within the defined
search length to locate the first complete pulse which can occur anywhere in the
search length. Pulses beyond the first complete pulse are disregarded and are not
demodulated. In order for the pulse search algorithm to detect a pulse edge, the
pulse must be at least 15 dB above the noise floor. The resultant pulse is then
demodulated. If sync search is selected in addition to pulse search, the analyzer
performs pulse search first then performs sync search on the resultant
demodulated pulse as illustrated below.

See “Using Digital Demodulation” earlier in the book for an example of using pulse
search and related features. See online help for key-specific information related to
pulse search.

The result length, which includes both the offset and the sync word, must not violate the
boundaries of the demodulated pulse.

22 - 14
 Digital Demodulation Concepts (Opt. AYA)

Speed and resolution considerations

Maximizing speed - measurement and display
l Use fewer points per symbol
l Use fewer symbols
l Decrease span

Maximizing resolution
l Use more points per symbol
l Increase max time points++

++ This parameter is under [System Utility] [memory usage] [config meas memory]. See the online help on this key for more
information on this topic

22 - 15
Digital Demodulation Concepts (Opt. AYA)

Filtering
General information
Data filtering limits bandwidth and reduces intersymbol interference. This analyzer
includes several commonly used filter types and has the ability to apply
user-defined filters. All filters are computed to 20 symbols in length.

The shape and width of the filter is defined by the alpha (for cosine filters) or the
BT (for gaussian filters). The alpha or BT indicates the filter roll-off (or excess
bandwidth) of the selected filter which occurs due to the practical inability of filter
technology to build a perfectly square (“brick-wall”) filter which would have an
alpha of 0 (no excess bandwidth). For example, a typical filter with an alpha of 0.3
has a bandwidth 30% greater than the theoretical minimum.

Filter choices for the measured and reference signals

Filtering for various communication systems may occur either at the transmitter or
the receiver; or the filtering may be distributed between the transmitter and the
receiver. This is an important concept which affects your filter selection for the
measured and reference signals. The analyzer’s measured filter represents filtering
in the system’s receiver while the reference filter represents filtering in the entire
system. Both filters share the same alpha/BT.

As an example, if the communication system under test filters the signal partially at
the transmitter and partially at the receiver (as with NADC signals), you choose a
root-raised-cosine filter for the measured signal to simulate receiver filtering in a
system which includes a similar filter at the transmitter. However, the reference
signal is a calculated ideal signal which is derived from the measured signal and
reflects only the filtering applicable to the measured signal. Therefore, you must
choose raised-cosine filtering for the reference in order to represent all the filtering
included throughout the simulated system. The following table shows some
examples of filter selection:

If the transmitter filter is: The measure filter should be: The reference filter should be:
root raised cosine root raised cosine raised cosine
raised cosine none raised cosine
gaussian none gaussian
any type user defined any type

22 - 16
 Digital Demodulation Concepts (Opt. AYA)

Square-root raised cosine filters

Many communication systems use distributed filtering, that is, filtering is
performed partially in the transmitter, to limit bandwidth, and partially in the
receiver, to limit interference. To achieve the overall desired frequency response
each filter’s transfer function is based on the square root of the desired response.
For these systems matched square-root raised cosine filters are used in the
transmitter and the receiver sections of the system to achieve optimum signal to
noise ratio. This implies that you must select similar filter characteristics in the
analyzer (which simulates the receiver) to the filter characteristics of the
transmitter. The standard NADC and JDC demodulation types offered in this
analyzer provide this type of filter.

The equation for the square-root raised cosine (root Nyquist) filter follows:++

 (1−α)
1 when 0 ≤ f ≤
 2T
H(f) =  1 

√ π(2fT−1) 
 2 1−sin 2 α  when
1− α
2T
≤f≤
1+ α
2T

0 when otherwise

Raised cosine filters
Raised cosine filters are used in systems which perform all the filtering in the
transmitter. This is typical of some mobile communication systems.

The equation for the raised cosine (Nyquist) filter follows:++

π t α π t
sin  cos  
T
   T . 1
h(t) =
πt 2
2 α 
1− t
 T 

++ Where T is the symbol interval

22 - 17
 Digital Demodulation Concepts (Opt. AYA)

Gaussian filter
Gaussian filters are typically used for MSK and FSK signals. Typically MSK
transmission performs all filtering in the transmitter.++

g(t) = h(t) ∗ rect 

t
T
 

−t2
e
2 σ2 T 2
h(t) =

√ 2π σT


√ ln(2)
σ=
2 π BT

1 T
 when t <
rect  =
t T 2

T 
0 otherwise

Low pass filter (for FSK)
FSK measurements often require spans which are much larger than the symbol rate
because of large deviations inherent in many FSK signals. The wide span may
result in a noisy demodulated signal. The low pass filter is a type of narrow,
gaussian filter with a cutoff frequency greater than the symbol rate. Applying the
low pass filter to the measured signal may smooth the result.

User defined filters

You may define your own filter by creating an impulse response and storing it in a
data register. The impulse response must have 20 points per symbol and have an
overall length of less than or equal to 20 symbols (401 points). Also, the impulse
response must be real and should have an odd number of points so that the center
of the impulse is on a symbol.

Accuracy of user defined filters is undefined. The alpha and BT menu choice is not
applicable because user defined filters includes alpha/BT as part of the definition.

You must select reference filtering based on the purpose and type of user defined
filter. In some circumstances you may have to create a special user defined filter
for the reference signal.

++ Where T is the symbol interval

22 - 18
 Digital Demodulation Concepts (Opt. AYA)

Typical applications for user defined filters include:

l Custom filters other than those provided. In this case both measured and
reference filters are user defined.
l Modified filters which are based on the raised cosine filter but add channel
equalization. In this case the measure filter is a user defined filter and the
reference filter is standard.

See “Creating User Defined Signals” earlier in this book for an example of how to
create, view, and apply a user defined filter.

The documentation for the Standard Data Format Utilities which accompanies this
analyzer also includes a section (Mathcad Examples) describing how to create
waveforms and filters for use with this analyzer.

Special consideration for MSK

Matched filtering of a demodulated signal is not available with the MSK
demodulators. The measured filter is normally off. A user defined filter can be
selected. Its primary use is for additional band-limiting and channel equalization.

IS-95 Filters
The analyzer includes two filters that conform to US CDMA digital cellular standard
IS-95:
l IS-95 base-phase EQ filter.
l IS-95 Chebyshev filter.

These filters let you measure base stations or mobile phones that conform to US
CDMA digital cellular standard IS-95. The IS-95 base-phase EQ filter is the inverse
of the forward all-pass phase-compensation filter that is specified for IS-95
base-station transmission. The IS-95 Chebyshev filter is the filter specified in IS-95
that provides overall channel filtering for both base-to-mobile and mobile-to-base
links.

The IS-95 base-phase EQ filter is applied to the I/Q measured signal. The
IS-95 Chebyshev filter is applied to the I/Q reference signal. For additional details,
see online help (press [Help], [1] to display the online help index, then scroll through
the index and select IS-95 base phase EQ Softkey or IS-95 Chebyshev Softkey).

Hint Select [CDMA BASE] or [CDMA MOBILE] to automatically configure the analyzer for
base-station or mobile measurements. Both softkeys are located under
[Instrument Mode], [demodulation setup], [demod format], [standard setups].

22 - 19
 Digital Demodulation Concepts (Opt. AYA)

EDGE Filter
The EDGE filter is available as the reference filter and is used for demodulating
EDGE PSK signals. One of two filters may be used depending on the selection of
the measured filter.

Note The EDGE reference filter is available only in analyzers that have option B7A
(Enhanced Data rates for GSM Evolution—EDGE)

If the measured filter is OFF and you select the EDGE reference filter, the EDGE
reference filter is a linearized Gaussian filter as defined by ETSI (European
Telecommunications Standard Institute) GSM 05.04.

If the measured filter is EDGE (winRC) and you select the EDGE reference filter,
the EDGE reference filter is the GSM 05.04 filter described above convolved with
the EDGE (winRC) filter (the winRC filter is described below).

The BT of the EDGE reference filter is fixed. The value of BT (set with the [alpha/BT]
softkey) does not affect this filter.

EDGE (winRC) Filter

The EDGE (winRC) filter is available as the measured filter and is used for
demodulating EDGE PSK signals. This filter is a windowed, raised-cosine filter that
has an approximate 180 kHz, 6 dB bandwidth, as defined by ETSI (European
Telecommunications Standard Institute) GSM 05.05.

The BT of the EDGE (winRC) filter is fixed. The value of BT (set with the [alpha/BT]
softkey) does not affect this filter.

22 - 20
23

Video Demodulation
Concepts (Opt. AYH)

23 - 1
 Video Demodulation Concepts (Opt. AYH)

Overview
What you learn in this chapter
This chapter presents an overview of video demodulation and includes concepts to
help you understand how the analyzer demodulates video signals. For examples on
setting up a video demodulation measurement and analyzing demodulated data, see
the chapters titled “Using Video Demodulation” and “Analyzing Digitally
Demodulated Signals’’. For examples on creating an ideal digitally modulated
signal and creating a user-defined filter, see the chapter titled ‘’Creating
User-defined Signals’’. For key-specific information, refer to online Help, which is
accessed by pressing the [Help] key on the analyzer.

What option AYH does

Option AYH (Video Modulation Analysis) provides the following capabilities:

l 8 and 16 VSB (Vestigual Side Band) Demodulation

l 16, 32, and 64 DVB QAM (Digital Video Broadcast QAM) Demodulation
l 16, 32, 64 and 256 QAM Demodulation
l Mirrored (flipped) frequency spectrums so you can remove the effects of hi-side
mixing.
l For 89441A analyzers, option AYH lets you select a maximum frequency span of
7 MHz or 8 MHz for RF measurements.

Note Option AYH requires options AYA (Digital Demodulation Analysis) with
UFG or UTH (4 MB or 20 MB Extended RAM and Additional I/O).

Option AYH (Video Demodulation) is similar to option AYA (Digital Demodulation).

If you know how to use digital modulation, learning how to use video demodulation
will be easy. Both options provide the same analysis capabilities (measurement
data and display formats) and, for the most part, the same error parameters.
Option AYH adds a few new error parameters for VSB signals, such as SNR and
pilot level.

Like digital demodulation, video demodulation does not require external filters,
coherent carriers, or symbol clock timing signals. Instead, the analyzer locks to the
carrier and to the symbol-clock phase. The analyzer uses the demodulated signal to
generate an ideal reference signal. The ideal reference signal is compared to the
demodulated signal to provide a quantitative measure of system errors. The
analyzer has built-in filters which may be applied to the measured signal as well as
to the reference signal. This allows you maximum flexibility in comparing your
signal to an ideal signal. Additionally, this allows complete flexibility to probe any
analog point in your communication system. An optional second baseband channel
allows IQ baseband measurement capability.

23 - 2
 Video Demodulation Concepts (Opt. AYH)

Measurement Flow
General block diagram
The Video Demodulation instrument mode uses the analyzers digital demodulator
to demodulate video signals. The following block diagram shows the location of
the digital demodulator in the analyzer’s block diagram.

Location of Digital Demodulator in Analyzer’s Block Diagram

23 - 3
Video Demodulation Concepts (Opt. AYH)

Digital demodulator block diagram: QAM and DVB QAM

Selecting Video Demodulation reconfigures the analyzer’s digital demodulator
to demodulate video signals.

The following illustration shows the digital-demodulator block diagram when

QAM or DVB QAM is selected. The block diagram is identical for both
demodulation formats, with the following exceptions:

l Bits are decoded differently.

l I/Q origin offset is removed from QAM measurements. I/Q origin offset is not
removed from DVB QAM measurements (but it is reported in the symbol table).

If you’re familiar with digital demodulation, you may notice that 16 QAM and 32
QAM are available in both digital demodulation and video demodulation modes
(see table below). You can use either mode to demodulate these QAM formats.
However, note that DVB QAM measurements do not remove I/Q origin offset
whereas QAM measurements do remove I/Q origin offset.

Digital Demodulation Video Demodulation

(option AYA) (option AYH)
16 QAM ü ü

32 QAM ü ü

64 QAM ü

256 QAM ü

23 - 4
 Video Demodulation Concepts (Opt. AYH)

Digital Demodulator Block Diagram: QAM and DVB QAM

23 - 5
Video Demodulation Concepts (Opt. AYH)

Digital demodulator block diagram: VSB

The following shows the digital-demodulator block diagram when 8 VSB or 16 VSB
is selected.

Digital Demodulator Block Diagram: VSB

23 - 6
 Video Demodulation Concepts (Opt. AYH)

Measurement management
Measurement and display choices
The flexibility of this analyzer provides numerous possible ways of viewing digitally
demodulated signals. You may demodulate signals of various types then view
aspects of those signals in several ways. Modulation format, measurement data,
and data format may be combined in different ways for specific measurement
needs.

Carrier locking (all except VSB)

23 - 7
Video Demodulation Concepts (Opt. AYH)

For QAM signals, the analyzer’s center frequency must be close to the transmitted
carrier frequency to achieve carrier lock. The required proximity of the center
frequency to the carrier frequency varies depending on the signal type, symbol rate,
and system noise. If symbol locking appears poor, you may achieve better carrier
locking by observing the “Freq Err” value in the symbol table of any successfully
locked measurement and adding that amount to the center frequency.

To obtain reliable carrier locking, the difference between the analyzer’s center
frequency and carrier frequency should be within 3% of the symbol rate for 16 and
32 QAM. For 64 and 256 QAM, the difference should be within 1% and 0.2%:

16 or 32 QAM/DVB QAM: (center frequency) − (carrier frequency) ≤ ± .03 (symbol rate)

64 QAM/DVB QAM: (center frequency) − (carrier frequency) ≤ ± .01 (symbol rate)
256 QAM/DVB QAM: (center frequency) − (carrier frequency) ≤ ± .002 (symbol rate)

For QAM signals, the result length also affects the analyzer’s ability to achieve
reliable carrier lock, as shown in the following table. Result lengths less than those
shown may result in unreliable carrier lock. Optimal carrier locking occurs when
the result length is 1000 or greater.

The minimum result length

If the modulation type is:
for reliable carrier lock is:
16 QAM or 16 DVB QAM 50 symbols
32 QAM or 32 DVB QAM 75 symbols
64 QAM or 64 DVB QAM 150 symbols
256 QAM 400 symbols

The following may also affect carrier locking:

l A frequency span that is too narrow.

l An incorrect range setting.
l Using the wrong measured or reference filter.

23 - 8
 Video Demodulation Concepts (Opt. AYH)

Carrier locking and pilot search: VSB

The same parameters that affect carrier locking for QAM signals also affect carrier
locking for VSB signals. For VSB signals, the analyzer must also find the pilot
signal to achieve carrier lock.

When you start a VSB measurement, the analyzer displays SEARCHING FOR
PILOT. This message appears for a fixed amount of time while the analyzer
searches for the pilot signal. The measurement begins when the pilot search ends
(when the SEARCHING FOR PILOT message disappears).

If your VSB signal uses a high-side pilot, you must configure the analyzer to
demodulate a high-side pilot by selecting the Video Demodulation instrument mode
and pressing [Instrument Mode], [demodulation setup], [more], [freq spectrum mirror]. If you don’t
do this, the analyzer cannot find the pilot signal and cannot lock to your signal.

If the analyzer cannot lock to your signal, it displays CARRIER LOCK?. There are
several conditions that may cause loss of carrier lock, one of which is an
unsuccessful pilot search. If carrier lock is not obtained after several
measurements, the analyzer assumes the pilot search failed and repeats the pilot
search.

CARRIER LOCK? may appear even when the data appears to be locked. If this
occurs, the results may be inaccurate. Normally, this condition clears in a few
measurements as the instrument obtains better estimates of the carrier frequency.

The following paragraphs describe other conditions that may cause loss of carrier
lock. The analyzer always assumes pilot search failed if there are carrier lock
problems, even if the cause is one of the conditions below.

Like QAM signals, the analyzer’s center frequency must be close to the transmitted
carrier frequency to achieve carrier lock for VSB signals. The following formulas
show you how to calculate the ideal center frequency. Use the formula for low-side
pilot if your pilot is on the left (low) side of your spectrum; use high-side pilot if
your pilot is on the right (high) side of your spectrum.

Symbol Rate
Center Frequency (LOW SIDE PILOT) = + (Pilot Frequency)
4

Symbol Rate
Center Frequency (HIGH SIDE PILOT) = (Pilot Frequency) −
4

Hint The ‘’Using Video Demodulation’’ chapter contains a task that shows you how to
use the analyzer’s frequency counter to measure your pilot frequency.

The result length also affects the analyzer’s carrier locking. For VSB
measurements, the result length should be at least 800 symbols. Smaller result
lengths may cause unreliable carrier locking.

23 - 9
Video Demodulation Concepts (Opt. AYH)

The following may also cause loss of carrier lock:

l A frequency span that is too narrow.

l An incorrect range setting.
l Using the wrong measured or reference filter.
l Using [freq spectrum normal] when you should be using [freq spectrum mirror].
l A pilot that is not in phase with the transmitted data. You may be able to
compensate for pilot-phase problems by using [clock adjust] to change when the
analyzer samples the I/Q trajectory (press [Instrument Mode], [demodulation setup],
[more], [clock adjust]).

Input Range
The input range must be set correctly to obtain accurate measurements. Input
ranges that are too low overload the analyzer’s ADC. Input ranges that are too high
increase noise, which increases errors reported in error parameters, such as EVM.

To select the optimum input range when using the 89441A RF section (2-2650 MHz)
receiver, press [Range], [ch1 range] and decrease the range (using the down-arrow key)
until OV1 appears in the active trace. Then increase the range one step at a time
(by pressing the up-arrow key) until OV1 disappears.

To select the optimum input range when using the 89410A or the 89441A RF section
(0-10 MHz) or IF section (0-10 MHz) receiver, press [Range], [ch1 range] and decrease
the range (using the down-arrow key) until the Channel 1 Over and Half LEDs turn
on. Then increase the range one step at a time (by pressing the up-arrow key) until
the Over LED turns off.

I-Q measured signal

The IQ measured signal is the result of resampling the data to an integer number of
points per symbol and applying system gain normalization, carrier locking, and
filtering to the incoming signal. The filtering is user-selectable. For QAM
measurements, the analyzer also applies I/Q origin offset compensation (similar to
pilot removal for 8 VSB and 16 VSB). I/Q origin offset compensation is not applied
to DVB QAM measurements.

I-Q reference signal

A powerful analysis technique involves comparing a demodulated signal with an
ideal signal generated from detected bits. The analyzer detects bits from the
measured IQ signal and reconstructs a sequence of ideal I and Q states. These are
then treated as ideal impulses and are baseband filtered according to a reference
filter selected by the user. The resultant IQ reference can be overlaid or compared
side-by-side with the IQ measured signal.

The reference filter can be selected independently from the measured signal
(although the alpha is the same for both filters). Therefore, you can apply different
filters to the measured and reference signals to accommodate special test
requirements. For details, see the section on filtering later in this chapter.

23 - 10
 Video Demodulation Concepts (Opt. AYH)

Parameter interactions
Changing one parameter may necessitate changes in other parameters to achieve
the desired measurement. The following topics may help you optimize your
measurements and explain the reasons for changes made automatically by the
analyzer.

Data size considerations

Maximum data size for measured and reference IQ data is generally determined by
max time pts++ which is user selectable, 64-2048 (up to 4096 samples with
extended RAM option UFG or UTH). A given measurement will generally use all
available memory as defined by max time pts:

max time pts≥ result length × points ⁄ symbol

Resolution bandwidth
Resolution bandwidth in video demodulation is determined automatically by the
analyzer, and cannot be set independently. Resolution bandwidth is determined by
the time record length in the same manner as in Vector mode (see Fundamental
Measurement Interactions):

RBW = WBW ⁄ T

where RBW = resolution bandwidth

T = time record length
WBW = window bandwidth

However, in Video Demodulation the time record length is determined as follows:

T ≈ Result Length (in symbols) ⁄ Symbol Rate

++ See [max time points] under [System Utility], [memory usage], [configure meas memory]

23 - 11
Video Demodulation Concepts (Opt. AYH)

Span considerations
When selecting a frequency span, select the narrowest span that includes all of your
signal components (select a span that is larger than the bandwidth of your signal).
If you select a span that is too narrow, your measurement may have excessive
errors or the analyzer may lose carrier lock. If you select a span that is too wide,
your measurement may be affected by excessive noise and slower speed. You may
want to select the Vector instrument mode and set the span while viewing the
spectrum of your signal.

The analyzer displays a warning message if the frequency span is less than the
symbol rate for QAM and DVB QAM formats, and less than half the symbol rate for
VSB formats. If you see this warning message, increase the frequency span until it
includes all components of your signal.

When using video demodulation, the symbol rate determines the maximum
frequency span that you can select, as follows:

(SymbolRate)
Maximum Span (VSB) = 20
2.56

(SymbolRate)
Maximum Span (All Other Formats) = 20
1.28

These formulas can be rewritten to obtain the maximum-span-to-symbol-rate ratio.

These ratios are fixed and cannot be changed.

Maximum Span 20 Maximum Span

VSB: = OR = 7.8125
Symbol Rate 2.56 Symbol Rate

Maximum Span 20 Maximum Span

All Other Formats: = OR = 15.625
Symbol Rate 1.28 Symbol Rate

For VSB measurements, you must use arbitrary spans; you cannot use cardinal
spans (cardinal spans are spans that are powers of two, such as 2.5 MHz, 5 MHz,
and 10 MHz). The analyzer displays an error message if you select a cardinal span
when using the VSB demodulation format.

You may notice that the frequency span for digitally demodulated spectrums
(the IQ measured spectrum, IQ reference spectrum, and error vector spectrum) is
different than that set with the [span] softkey. The reason is that the digital
demodulation process derives its own frequency span and sample rate, as follows:

Demod Sample Rate = (symbol rate) × (points−per−symbol)

Demod Span = (Demod Sample Rate) ⁄ 1.28

23 - 12
 Video Demodulation Concepts (Opt. AYH)

When viewing digitally demodulated spectrums, aliasing may occur if the frequency
span is too large. The aliasing does not affect the accuracy of the demodulation,
only the validity of the spectral displays. The analyzer displays DATA? if it
suspects aliasing in digitally demodulated spectrums. To prevent aliasing, be sure
that:

Symbol Rate × (points−per−symbol)

Frequency Span ≤
1.28

where: Frequency span is set by pressing [Frequency] [span].

symbol rate is set by pressing [Instrument Mode], [demodulation setup], [symbol rate].
points-per-symbol is set by pressing [Time], [points/symbol].

Display limitations
Points per symbol affects all displayed results by controlling result resolution but is
not coupled to span. More points per symbol improves the resolution of the vector
diagram, but there is still only one point at the symbol clock.

(number of time points) = (result length) × (points−per−symbol)

Result length takes precedence over points per symbol—if you try to set points per
symbol such that the number of time points would exceed the memory size++, an
error indicates that the requested number of points per symbol cannot be set. In
this case you may choose to reduce the result length to allow an increased number
of points per symbol. Conversely, if you increase result length such that the
number of time points exceeds the memory limit, points per symbol automatically
decreases to allow the result length to increase.

++ See [max time points] under [System Utility]

23 - 13
Video Demodulation Concepts (Opt. AYH)

Feature Availability in Video Demodulation

Most features which are available with other instrument modes are also available
with video demodulation with the following exceptions:
l Time gating is not available
l Averaging is not available for any trace data in video demodulation but may
be applied to numeric error data in the symbol table. Three types of averaging
are available:
– rms (video)—computes an rms average of each rms error update in
the table

– rms exponential—is like rms (video) except that averaging continues past
the average count with an exponential weighting

– continuous peak hold—keeps track of the peak rms error of each update
in the table

23 - 14
 Video Demodulation Concepts (Opt. AYH)

Special considerations for sync search

Synchronization words (or patterns) are often used to resolve carrier phase
ambiguity on non-differential modulation formats. It is important to realize that
synchronization words are optional and are not necessary in order to achieve
carrier locking.

Note You cannot use sync search for VSB measurements. The [sync search on/off] softkey
(under the [Time] hardkey) is ghosted (inactive) if VSB is selected.

Sync search lets you use a synchronization pattern to isolate a portion of your
signal for display and analysis. The analyzer searches through demodulated data to
find your sync pattern, and then uses the [result length] to determine how much data to
display, and the [offset] to display data relative to the sync pattern.

Note The sync pattern must be a multiple of the number of bits-per-symbol. For
example, if the number of bits-per-symbol is 4 (as with 16 QAM), the number of bits
in the sync pattern must be a multiple of four. Sync search lets you specify any
number of bits for the sync pattern, however, bits that aren’t a multiple of the
bits-per-symbol are truncated. In this example, if you entered 6 bits for the sync
pattern, sync search would only use the first four bits.

Triggering determines when the analyzer starts demodulating data and search
length determines when the analyzer stops demodulating data. Sync search locates
only the first match and ignores any subsequent matches within the search length.
The sync pattern and the offset must fit within the search length.

You may use sync search with or without pulse search, although many
measurements require both features.

See “Using Digital Demodulation” earlier in the book for an example of using sync
search. See online help for the following softkeys for additional information (all
softkeys are under the [Time] hardkey):
l [sync search]
l [search length]
l [sync setup] → [pattern]
l [sync setup] → [offset]

23 - 15
 Video Demodulation Concepts (Opt. AYH)

Special considerations for pulsed signals

You can use pulse search to demodulate pulsed (burst) transmissions. Pulsed
transmissions are generated by on/off carriers such as those used in mobile units.

Pulse search is identical for both video and digital demodulation. To learn how to
use pulse search, see ‘’Special considerations for pulsed signals’’ in the Digital
Demodulation Concepts chapter. For an example using pulse search, see
‘’To set up pulse search’’ in the Using Digital Demodulation chapter.

Note You cannot use pulse search for VSB measurements. The [pulse search on/off] softkey
(under the [Time] hardkey) is ghosted (inactive) if VSB is selected.

Speed and resolution considerations

Maximizing speed - measurement and display
l Use fewer points per symbol
l Use fewer symbols
l Decrease span

Maximizing resolution
l Use more points per symbol
l Increase max time points++

++ This parameter is under [System Utility] [memory usage] [config meas memory]. See the online help on this key for more
information on this topic

23 - 16
 Video Demodulation Concepts (Opt. AYH)

Filtering
General information
The analyzer uses the same filters for digital demodulation and video
demodulation. For details, see ‘’Filtering’’ in the Digital Demodulation Concepts
chapter, and Note the following differences:
l You cannot use user-defined filters when VSB is selected.
l To use user-defined filters with QAM and DVB QAM, the impulse response that
defines the filter must have 20 points-per-symbol and be no longer than 40
symbols, which means the overall length must be less-than-or-equal-to 801
points.

Mirrored Spectrums
Option AYH provides a feature that lets you configure the analyzer’s
demodulator to conjugate the complex time-domain waveform. This has the
effect of flipping the spectrum around the analyzer’s center frequency. To
enable this feature, select the Video Demodulation instrument mode, then press
[Instrument Mode], [demodulation setup], [freq spectrum mirror].

This feature is often used with VSB measurements that use a high-side pilot.
For VSB signals, the pilot must be on the low-side of the spectrum to achieve
carrier lock. For details, see ‘’Carrier locking and pilot search: VSB’’ earlier in
this chapter.

23 - 17
Video Demodulation Concepts (Opt. AYH)

RF receiver and 8 MHz spans

Option AYH lets you extend the maximum frequency span of the
H89441A RF receiver from 7 MHz to 8 MHz. Press [Instrument Mode] [receiver]
[RF 2-2650 MHz)] to display the normal (7 MHz) and wide (8 MHz) softkey menu. Press
[RF (2-2650 MHz) normal] to configure the RF section for a maximum frequency span of
7 MHz; press [RF (2-2650 MHz) wide] to extend the maximum frequency span to 8 MHz.

Option AYH replaces this softkey: With these softkeys:

[RF section (2-2650 MHz)] [RF (2-2650 MHz) normal]
[RF (2-2650 MHz) wide]

Remember that, in two-channel analyzers, channel 1 determines the frequency span

of channel 2. Therefore, selecting wide (8 MHz) also extends the frequency span
of channel 2 to 8 MHz (channel 2 is always baseband, so the frequency span is
0 to 8 MHz).

Normal (7 MHz) provides the best specifications. Specifications are located in the
Installation & Verification Guide.

23 - 18
24

Wideband CDMA Concepts

(Options B73, B79, and 080)

24 - 1
 Wideband CDMA Concepts (Options B73, B79, and 080)

Overview
What you learn in this chapter
This chapter presents an overview of wideband CDMA (W-CDMA) and includes
concepts to help you understand how the analyzer measures W-CDMA signals. For
examples on setting up a W-CDMA measurement, see the chapter titled “Using
Wideband CDMA”’. For key-specific information, refer to online Help, which is
accessed by pressing the [Help] key on the analyzer.

What option B73 does

Option B73 (Digital Wideband CDMA Analysis) lets you measure base-station
W-CDMA signals that conform to the Japanese Experimental System and that use
Perch 1. With option B73, the analyzer despreads and decodes W-CDMA signals
and automatically identifies all the active channels regardless of the code rate or
the short code. Option B73 provides the following features:

Note Option B73 requires options AYA (Digital Demodulation Analysis), AY9 (Extended
Time Capture), and UTH (20 Mb Extended RAM and Additional I/O) .

l Single or composite code-domain power displays (the composite display shows

all code layers simultaneously). You can normalize code-domain power to
display the code-domain power relative to the total signal power in the code
domain.
l Time-domain displays, such as IQ measured, IQ reference, IQ magnitude or
phase error, and error vector traces.
l Composite time-domain display, which is the detected, filtered time-domain
waveform (useful for VOX on/off and transmission power control measurements.
l Symbol table display that lets you see error parameters (such as RHO) and
demodulated bits for a code channel.
l Adjustable filter alpha.
l 4 MHz, 8 MHz, and 16 MHz, variable chip rates (the chip rate is variable from
slightly below 4 MHz to slightly above 16 MHz).
l 32 MHz chip rate when using the ch1 + j*ch2 receiver (the I component is input
on channel 1; the Q component on channel 2).
l A standard setup that automatically configures the analyzer for 5 MHz, 10 MHz,
and 20 MHz W-CDMA measurements.

24 - 2
 Wideband CDMA Concepts (Options B73, B79, and 080)

l Scramble-code group or manual scramble-code detection. You can enter the

scramble code manually or configure the analyzer to determine the scramble
code within a scramble-code group.
l Mirrored (flipped) frequency spectrums so you can remove the effects of hi-side
mixing.
l Time gating, which lets you select one or more slots for analysis.

What option B79 does

Option B79 extends option B73 (Digital Wideband CDMA Analysis) to also measure
W-CDMA signals that conform to ARIB (Association of Radio Industries and
Businesses) revision 1.0-1.2. Option B79 adds additional code layers, code
channels, code groups, and scramble codes required for ARIB 1.0-1.2 signals.

Note Option B79 requires option B73.

What option 080 does

Option 080 (Digital 3GPP W-CDMA Analysis) lets you measure 3GPP, version 3.1
(1999) forward link (base-station) W-CDMA signals. Like option B73 and B79, this
option despreads and decodes W-CDMA signals and automatically identifies all the
active channels regardless of the code rate or the short code. Option 080 provides
the same features as option B73, with the following exceptions:

Note Option 080 requires options AYA (Digital Demodulation Analysis), AY9 (Extended
Time Capture), and UTH (20 Mb Extended RAM and Additional I/O) .

l 3.84 MHz variable chip rate (the chip rate is variable from slightly below 3.84
MHz to slightly above 3.84 MHz).
A standard setup that automatically configures the analyzer for 5 MHz W-CDMA
measurements.

24 - 3
Wideband CDMA Concepts (Options B73, B79, and 080)

Measurement Flow
The Wideband CDMA instrument mode inserts the W-CDMA digital demodulator
into the signal path, as shown in the following illustration. The next page shows
the block diagram for the W-CDMA digital demodulator.

Location of W-CDMA Digital Demodulator in Analyzer’s Block Diagram

24 - 4
 Wideband CDMA Concepts (Options B73, B79, and 080)

Digital Demodulator Block Diagram: W-CDMA

24 - 5
 Wideband CDMA Concepts (Options B73, B79, and 080)

Setting up a W-CDMA Measurement

To make a W-CDMA measurement, you select the Wideband CDMA instrument
mode and then select the type of W-CDMA signal that you want to demodulate. If
you have option B73 only, you can only demodulate W-CDMA signals that conform
to the 1998 Japanese Experimental System. With option B79, you can also
demodulate signals that conform to ARIB rev 1.0-1.2. With option 080, you can
demodulate 3GPP version 3.1 (1999) forward link W-CDMA signals.

In the W-CDMA instrument mode, the analyzer accumulates and demodulates one
or more slots of data and displays the results.
Important Japanese Experimental System W-CDMA signals must use Perch 1. ARIB rev
1.0-1.2 signals must use Perch with the first search code. 3GPP signals must use
PSCH and CPICH. The analyzer can only demodulate W-CDMA signals that meet
these criteria.

You can configure the analyzer to demodulate standard and non-standard W-CDMA
signals. To demodulate a standard W-CDMA signal, first set the demodulation
format by pressing [Instrument Mode], [demodulation setup], [demod format], [trial 1998 forward link]
or [ARIB 1.0-1.2 forward link]. or [3GPP 1999 forward link]. The [trial 1998 forward link], [ARIB 1.0-1.2
forward link], and [3GPP 1999 forward link] softkeys display the standard-setups menu that
automatically configures most W-CDMA measurement parameters (such as chip
rate, center frequency, main length, filter alpha, and scramble code).

Signal Connections and Input Range

You connect your signal to the CHANNEL 1 input and configure the input (via the
[Input] hardkey) as you do for all other measurements. If you have a two-channel
analyzer, channel 2 is only enabled when the receiver is (ch1 + j*ch2). The (ch1 +
j*ch2) receiver lets you measure the I and Q components of your W-CDMA
signal—connect I to the CHANNEL 1 input; connect Q to the CHANNEL 2 input.

The input range must be set correctly to obtain accurate measurements. Input
ranges that are too low overload the analyzer’s ADC. Input ranges that are too high
increase noise, which increases errors reported in error parameters, such as EVM.

To select the optimum input range when using the 89441A RF section (2-2650 MHz)
receiver, press [Range], [ch1 range] and decrease the range (using the down-arrow key)
until OV1 appears in the active trace. Then increase the range one step at a time
(by pressing the up-arrow key) until OV1 disappears.

To select the optimum input range when using the 89410A or the 89441A RF section
(0-10 MHz) or IF section (0-10 MHz) receiver, press [Range], [ch1 range] and decrease
the range (using the down-arrow key) until the Channel 1 Over and Half LEDs turn
on. Then increase the range one step at a time (by pressing the up-arrow key) until
the Over LED turns off.

Frequency Span

24 - 6
 Wideband CDMA Concepts (Options B73, B79, and 080)

When selecting a frequency span, select the narrowest span that includes all of your
signal components (select a span that is slightly larger than the bandwidth of your
signal). If you select a span that is too narrow, your measurement may have
excessive errors or the analyzer may lose carrier lock. If the span is less than the
chip rate, the analyzer displays DATA? to indicate that the results may not be
accurate.

For best results, the frequency span should be set according to the following
formula (filter_alpha is the filter characteristic of the Nyquist filter used to
demodulate the W-CDMA signal):

FrequencySpan ≥ (1 + filter_alpha) Chip_Rate

If you select a span that is too wide, your measurement may be affected by
excessive noise and slower speed. You may want to select the Vector instrument
mode and set the span while viewing the spectrum of your signal.

Note For W-CDMA measurements, the [mx WCDMA span] softkey (press [System Utility],
[memory], [configure meas memory]) limits the analyzer’s frequency span.. The value
of this softkey determines the amount of memory reserved for W-CDMA
measurements. For optimum memory usage, set the value of [mx WCDMA span] no
larger than the maximum span that you will measure.

Center Frequency
The analyzer’s center frequency must be set within 500 Hz of the center frequency
of your W-CDMA signal, otherwise the analyzer may not be able to lock onto your
signal.

Scramble Code
The analyzer must know the scramble code to demodulate your W-CDMA signal.
You can manually enter the scramble code or you can configure the analyzer to use
an algorithm to find the scramble code from within a scramble-code group. The
algorithm increases measurement time. To minimize measurement time, the
algorithm should only be used when you don’t know the scramble code. To setup
the scramble code, press [Instrument Mode], [demodulation setup], [scramble code]. For
additional details about scramble codes, see online help for the [scramble code] softkey.

24 - 7
 Wideband CDMA Concepts (Options B73, B79, and 080)

Chip Rates, Code Layers, and Symbol Rates

To successfully demodulate W-CDMA signals, the analyzer’s chip rate must match
the chip rate of your W-CDMA signal. The chip rate determines which code layer is
assigned to each symbol rate , as shown in the following table.

Note Selecting the ch1 + j*ch2 receiver (the I component is input on channel 1; the Q
component on channel 2) doubles the maximum chip rate .

Symbol Rate (kSymbols/second)

Code Layer 3.84 MHz 4 MHz 8 MHz 16 MHz

(code class) Chip Rate †† Chip Rate Chip Rate Chip Rate
C0
C1 2048 † 4096 †
C2 960 †† 1024 2048 † 4096 †
C3 480 †† 512 1024 2048 †
C4 240 †† 256 512 1024
C5 120 †† 128 256 512
C6 60 †† 64 128 256
C7 30 †† 32 64 128
C8 15 †† 16 32 64
C9 7.5 †† 8† 16 32
C10 8† 16
C11 8†

† These rates are only available for ARIB 1.0-1.2 W-CDMA signals.
†† These rates are only available for 3GPP 1999 forward link W-CDMA signals.

24 - 8
 Wideband CDMA Concepts (Options B73, B79, and 080)

Main Length
The number of slots that the analyzer demodulates is set by the [Instrument Mode],
[demodulation setup], [main length] softkey. This softkey is also located under the [Time]
hardkey for convenience.

Main length must be entered as an integer number of slots. You can enter from 8 to
20 slots. You can also enter main length using units of seconds. If you use seconds,
the analyzer automatically increments your entry as necessary to obtain an integer
number of slots.

Filtering
The analyzer always uses a root-raised cosine filter to demodulate W-CDMA
signals. Unlike Digital or Video Demodulation, you cannot change the filter—you
can only control the filter’s alpha.

To change the filter’s alpha, press [Instrument Mode], [demodulation setup], [filter alpha]. You
can select a filter alpha from 0.05 to 1.0 in increments of 0.01.

Mirrored Spectrums
The Wideband CDMA instrument mode provides a feature that lets you configure
the analyzer’s demodulator to conjugate the complex time-domain waveform. This
has the effect of flipping the spectrum around the analyzer’s center frequency. This
feature is useful when the W-CDMA signal is inverted, or flipped due to an IF
downconversion. To enable this feature, select the Wideband CDMA instrument
mode, then press [Instrument Mode], [demodulation setup], [freq spectrum mirror].

Time-Domain Corrections
As shown in the general block diagram earlier in this chapter, the analyzer’s digital
demodulator uses time-domain data. Because of this, the analyzer automatically
enables time-domain corrections when you select the Wideband CDMA instrument
mode. For details about time-domain corrections, see online help for [System
Utility], [time domain cal].

24 - 9
 Wideband CDMA Concepts (Options B73, B79, and 080)

Trigger Signal
If you use a trigger signal, the first slot in the measurement is the slot after the first
complete LMS (Long-Code Masked Symbol) for trial 1998 signals or the slot with
the first SCS (Search Code Symbol) for ARIB 1.0-1.2 signals, or the slot with the
first PSCH (Primary Sync Channel) for 3GPP signals.

Note For trial 1998 signals, LMS is at the end of the slot; for ARIB 1.0-1.2 signals, SCS is
at the beginning of the slot; for 3GPP signals, PSCH is at the beginning of the slot.

Setting up a trigger for ARIB 1.0-1.2 and 3GPP signals is different than setting up a
trigger for trial 1998 signals. For ARIB 1.0-1.2 and 3GPP signals, if you trigger at the
beginning of a slot, that slot becomes the first slot in the measurement (because the
SCS or PSCH is at the beginning of the slot). For trial 1998 signals, the slot
immediately following that slot becomes the first slot in the measurement (because
the LMS is at the end of the slot).

The following illustration shows you how to use trigger delay to select the first slot
in a measurement for trial 1998 signals.

TIP For trial 1998 signals, to make slot 1 the first slot in the measurement when you
trigger at the start of a frame, set a pre-trigger delay (negative trigger delay) such
that the signal sampling begins before the LMS preceeding slot 1. For a 4.096 MHz
system, the duration of the LMS is –625µs. In this case, slot 1 will be the first slot in
the measurement if you trigger at the start of the frame (the start of slot 1) and set
the trigger delay to –625µs.

Main length is set to 16 slots. Without a

trigger signal, the analyzer may begin the
measurement using any slot. In this
example, slot 2 is the first slot in the
measurement (as shown by the shaded
portion).

Main length is again set to 16 slots. This

example uses a trigger signal with trigger
delay to make slot 1 the first slot in the
measurement. The trigger delay ensures
that the trigger point occurs before the
LMS in slot 16.

24 - 10
 Wideband CDMA Concepts (Options B73, B79, and 080)

Viewing Measurement Results

The flexibility of the analyzer provides numerous ways of viewing W-CDMA signals.
Measurement data and data format may be combined in different ways for specific
measurement needs, as shown below. Press [Measurement Data] to select the
measurement data; press [Data Format] to select a display format for the data.

24 - 11
 Wideband CDMA Concepts (Options B73, B79, and 080)

Code-Domain Power Displays

The analyzer lets you view measurement data in the code domain or time domain.
In the code domain, you can view the code-domain power for all layers—called
composite code-domain power—or code-domain power for a single layer.

Note The analyzer uses data from the first two slots in the measurement to generate
code-domain power displays. Remaining slots do not affect code-domain power. If
time gating is on, the analyzer uses the first two slots in the gate time-record unless
the gate length is one slot. If the gate length is set to one slot, the analyzer only
uses the first slot in the gate time-record to compute code-domain power.

The analyzer introduces a unique way of viewing composite code-domain power.

The composite display uses a different color for each code layer. For example, if
channel 11 in code layer 64 ksym/s contains power, the channel is colored orange.
If channel 3 in code layer 16 ksym/s contains power, the channel is colored red. If a
channel is inactive (contains no power), the channel is not colored. At a single
glance you can determine the channels and layers that contain power and the
channels that are inactive (contain no power).

The analyzer also lets you view code-domain power for a single layer. You can
view code-domain power for any active layer (if a layer is inactive, the softkey that
selects it is ghosted).

Tip For examples of composite and single-layer code-domain power, see the chapter
titled “Using Wideband CDMA”’.

By default, code-domain power is normalized and is relative to the total signal

power in the code domain. To turn normalization off, press [Instrument Mode],
[demodulation setup], [normalize off].

24 - 12
 Wideband CDMA Concepts (Options B73, B79, and 080)

Time-Domain Displays
The analyzer provides several time-domain displays, including a symbol table that
shows the demodulated bits with error parameters such as RHO, Frequency Error,
and EVM.
Composite Time Domain
The composite time-domain display is the filtered, detected time-domain signal and
shows power (in dB) versus time for all channels. This display is useful for VOX
on/off and transmission power control measurements.
Other Time-Domain Displays
The remaining time-domain displays (including the symbol table) show IQ data for
a single layer. The [code layer] and [code channel] softkeys (under the [Time] hardkey)
determine which layer is displayed. You can view the symbol table or the IQ
measured or IQ reference signal for the selected layer.

Tip To quickly set the [code layer] and [code channel] softkeys, display composite
code-domain power, place the marker on the desired layer and channel, then press
[Marker →], [mkr → layer/channel].

The symbol table shows the final product of digital demodulation—the binary bits
for each symbol in the selected channel and layer. In addition, the symbol table
displays information about the W-CDMA signal and error parameters, such as ( For
more details about these error parameters, see the analyzer’s online help):
l The scramble code used to demodulate the W-CDMA signal.
l The slot number of the first slot in the measurement.
l T frame—how far the slot is offset (in slots) from the Perch channel (not used
for 3GPP 1999 forward link W-CDMA signals).
l T slot— how far the slot is offset (in symbols) from the Perch channel (not used
for 3GPP 1999 forward link W-CDMA signals).
l T Trigger—the amount of time, in chips, from the trigger to the start of the sync
symbol.
l tDPCH—the tDPCH time value for the DPCH on the selected code channel (only
used for 3GPP signals).
l pilots—the number of pilot bits detected in the DPCH for the selected code
channel (only used for 3GPP signals).
l RHO, EVM, frequency error, magnitude error, and phase error.

The IQ measured signal is the result of resampling the data to an integer number of
points per symbol and applying system gain normalization, carrier locking, and
root-cosine filtering to the incoming signal.

The IQ reference signal provides a powerful analysis technique and involves

comparing the demodulated signal with an ideal signal generated from detected
bits. The analyzer detects bits from the measured IQ signal and reconstructs a
sequence of ideal I and Q states. These are then treated as ideal impulses. The
resultant IQ reference can be overlaid or compared side-by-side with the IQ
measured signal.

24 - 13
 Wideband CDMA Concepts (Options B73, B79, and 080)

You can also display the IQ magnitude error, IQ phase error, or the IQ error vector
trace. These error displays show the error between the IQ measured and IQ
reference signals. IQ magnitude error is the magnitude error between these two
signals; IQ phase error is the phase error between these two signals. The IQ error
vector trace is created by computing the error vector from corresponding points in
the IQ measured and IQ reference signals.

Note For further details about time-domain displays and error parameters, see online
help (press [Help], then press [Measurement Data] and the appropriate softkey).

Time Gating
The Wideband CDMA instrument mode lets you perform time-gating analysis on
demodulated results. Time gating lets you view selected slots within the
measurement. In other words, time gating lets you view a slice of the measurement
results, called the gate.

You use the [gate length] softkey to set the width of the gate (in slots). You use the
[gate delay] softkey to determine which slot to use as the first slot in the gated
measurement. A gate delay of zero selects the first slot in the measurement; a gate
delay of one selects the second slot, and so forth.

When time gating is on, time gating applies to all code domain, time domain, and
symbol table/error summary displays. For an example of time gating, see To view
data for one or more slots in the chapter titled ‘’Using Wideband CDMA’’.

Tip When time gating is on, the analyzer adds the word Gate to the trace title to remind
you that the trace does not include all slots in the measurement but, instead, only
includes the slots selected by the gate.

24 - 14
 Wideband CDMA Concepts (Options B73, B79, and 080)

Parameter interactions
Changing one parameter may necessitate changes in other parameters to achieve
the desired measurement. The following topics may help you optimize your
measurements and explain the reasons for changes made automatically by the
analyzer.

Data size considerations

The value of [main length] determines the amount of data, in slots, that the analyzer
acquires. The value of [max time pts]++ det ermines the maximum data size for IQ
measured and IQ reference data, and can be set from 256 to 4096 points.

Tip Even with max time pts set to 4096 time points, there are signals that contain more
data than you can view at one time (for example, if the signal is composite time or
IQ measured or IQ reference for higher symbol rates). For these signals, turn on
time gating and view the measurement data in slices.

Resolution bandwidth
Resolution bandwidth in W-CDMA is determined automatically by the analyzer, and
cannot be set independently. Resolution bandwidth is determined by the time
record length in the same manner as in Vector mode (see Fundamental
Measurement Interactions):

RBW = WBW ⁄ T

where RBW = resolution bandwidth

T = time record length
WBW = window bandwidth

In Wideband CDMA, the time record length is determined as follows:

T ≈ Result Length (in symbols) ⁄ Symbol Rate

Points Per Symbol

All W-CDMA measurement data, except composite time, is displayed at 1
point-per-symbol. Composite time is displayed at 16 points-per-symbol.

++ See [max time points] under [System Utility], [memory usage], [configure meas memory]

24 - 15
Wideband CDMA Concepts (Options B73, B79, and 080)

Feature Availability in W-CDMA

Most features which are available with other instrument modes are also available
with W-CDMA analysis, with the following exceptions:
l You cannot include Wideband CDMA measurement data in math functions.
In addition, when the instrument mode is Wideband CDMA, you cannot display
math functions.
l Averaging is not available for trace data in W-CDMA but may be applied to
numeric error data in the symbol table. Three types of averaging are available:
– rms (video)—computes an rms average of each rms error update in
the table

– rms exponential—is like rms (video) except that averaging continues past
the average count with an exponential weighting

– continuous peak hold—keeps track of the peak rms error of each update in
the table

Increasing Measurement Speed

Here are some tips for maximizing measurement speed.
l Use fewer slots
l Use the smallest frequency span
l Use cardinal frequency spans
l Enter the scramble code instead of specifying a scramble-code group.
l Use the smallest possible value for mx WCDMA span (press [System Utility],
[memory], [configure meas memory]).

24 - 16
 Wideband CDMA Concepts (Options B73, B79, and 080)

Troubleshooting W-CDMA Measurements

This section lists common problems that you may encounter when making
W-CDMA measurements, plus the cause of and fix for each problem.

Problem/Cause Solution

Unable to demodulate W-CDMA Signal

ü For [trial 1998 forward link], Perch 1 signal Verify that signal includes Perch 1 and that the Perch
not present or signal level too low. signal level is adequate.
ü For [ARIB rev 1.0-1.2], Perch signal with Verify that signal includes Perch with the first search
first search code not present or signal level code and that the signal level of each is adequate.
too low.
ü For [3GPP 1999 forward link], PSCH or Verify that signal includes PSCH and CPICH and that
CPICH are missing or too low. the signal level of each is adequate.
ü W-CDMA signal is frequency inverted. Press [Instrument Mode], [demodulation setup] and
select [freq spectrum mirror].
ü Frequency span is too narrow. See Frequency Span earlier in this chapter.
ü Analyzer’s center frequency is too far from See Center Frequency earlier in this chapter.
the signal center frequency.
ü Too many active channels. Reduce the number of active channels in your
W-CDMA signal.
ü Incorrect scramble code specified. Press [Instrument Mode], [demodulation setup],
[scramble code] and set the scramble code to that used
by your signal.
ü Several base stations are transmitting and Increase the Perch signal level for the station that you
the analyzer is using the Perch channel want to lock to.
from the wrong base station (the analyzer
selects the base station with the highest
Perch signal level).
ü Code group selected and wrong group Press [Instrument Mode], [demodulation setup],
number specified. [scramble code] and enter the correct group number.

24 - 17
Wideband CDMA Concepts (Options B73, B79, and 080)

Problem/Cause Solution

Noise floor increases at channel turn on

ü The channel that you are turning on uses a Change the scramble code used to transmit the
different scramble code than that set in channel to match that used by the Perch channel.
analyzer.

Wrong channels marked as active

ü Poor signal-to-noise ratio. Increase signal power or remove cause of
noise/distortion.
ü All bits not transmitted within a physical Transmit all bits within a physical channel, or select
channel. code-domain power for the layer associated with the
physical channel of interest (single-layer code-domain
power displays are accurate even if the analyzer is
unable to identify active channels in the composite
code-domain power display) .
ü Data pattern induces large disparity in Verify that your transmitter is transmitting properly.
power between slower channels overlaid
by the channel of interest.
Tslot and Tframe or tDPCH incorrect
(DATA? displayed in the symbol table)
ü Time gating is on and the gate length is Select 3 or more slots or turn off time gating.
less than 3 slots (fewer slots introduce a
higher degree of false correlation on pilot
and sync signals).

24 - 18
 Index

INDEX

! band-selectable measurements 19-3

080, option 24-3 SEE ALSO zoom measurements
16QAM demodulation, example 10-1 bandwidth
1c2 (Agilent IBASIC option) a-v SEE information bandwidth
2-channels SEE noise equivalent bandwidth
digital demod 8-16 SEE resolution bandwidth
video demod 9-15 SEE window bandwidth
32QAM signal, example 11-2 bandwidth coupling 19-4
3GPP, W-CDMA analysis 24-3 baseband measurements 19-3
3GPP, W-CDMA signals 13-2 baseband signals
digital demod 8-16
A video demod 9-15
accessories a-vi, a-ix bins
adaptive equalization 12-1 SEE frequency points
aliasing bit patterns, digital demodulation
digital demod 22-12 SEE symbol states
video demod 23-11, 24-15 block diagrams
AM demodulation analog demodulation 20-3
algorithm 20-9 digital demodulation 22-4
analyzers general 18-11
types of 18-4 scalar 17-3
arbitrary waveforms, using 6-1 vector 17-5
assistance (by phone) 2-i video demodulation 23-3
average power statistics W-CDMA 24-5
displaying 5-4 block size
averaging SEE time record size
in analog demodulation 20-13 burst modulation
in digital demodulation 22-13 SEE pulsed signals
video demodulation 23-14, 24-16
AY9 (1 Megasample extended time C
capture option) a-v capture
AYA (vector modulation analysis) a-v SEE time capture
AYB (spectrogram/waterfall) a-v carrier locking
AYH (video modulation analysis) a-v analog demod 20-8
AYJ (adaptive equalization) 12-1 video demod 22-9, 23-8
carrier, auto 20-8
B CCDF
B73 (W-CDMA option) a-v displaying 5-2
B79 (ARIB rev 1.0-1.2 W-CDMA CDMA 22-20
option) a-v channel
B7A (EDGE option) a-v optional IF input a-viii
Index

code domain power (W-CDMA) 24-12 VSB, example 9-6

composite time domain (W-CDMA) detection 19-12
24-13 digital demodulation
constellation diagram, example 8-5, setup 9-2
9-12 SEE demodulation, digital
coupling disk drive
bandwidth 19-4 flexible disks for a-x
display resolution 19-11 - 19-12
D display state definitions 10-10
dc measurements 19-9 display, digital 19-9
dc response 19-9 DVB QAM
demodulation, analog block diagram 23-4
SEE ALSO AM, FM, PM demodulation dynamic signal analyzers 18-6
averaging 20-13
block diagram 20-3 E
carrier locking 20-8 EDGE signal
examples 1-1 SEE ALSO demodulating
gating 20-13, 21-5 SEE ALSO viewing error vector
overview 17-6 error analysis
triggering 20-13 digital demod 10-12
demodulation, digital video demod 10-12
block diagrams 22-4 error parameters, W-CDMA 24-13
constellation points, viewing 10-5 error summaries
example 10-1 video demod 10-8
filtering 22-17 error vector, example 10-12
filters, user defined 11-6 eye diagram, example 8-5, 9-12
formatting displays 22-8
power calculation 10-4 F
pulse search 8-6 FFT 18-6
setup 8-2, 22-7 properties of 18-7
signal creation 11-2 SEE ALSO spectrum analyzers
standard setups, example 8-4 filters
sync search 8-8 digital demodulation 22-17
demodulation, video gaussian 22-19
block diagram 23-3 IF 19-18
capabilities 23-2 raised cosine 22-18
carrier locking 22-9, 23-8 square-root raised cosine 22-18
center frequency, setting 9-4 user defined 11-6
constellation points, viewing 10-5 user defined 22-19
DVB QAM, example 9-10 video 19-3
example 10-1 flexible disks a-x
filters, user defined 11-6 FM demodulation
formatting and displaying 23-7 algorithm 20-12
formatting displays 9-12 example 3-5
mirrored spectrums 23-17, 24-9 fourier transform 18-7
range, setting 23-10, 24-6 frequency domain overview 18-2
setup 9-2 frequency points 19-5
setup, 32 DVB QAM 9-8 and display 19-12
signal creation 11-2 frequency range, extending 16-2
symbol locking 22-9, 23-8 frequency resolution 19-11
sync search 9-13
 Index

frequency span M
and display 19-5 manual sweep 19-13
and information bandwidth 19-3 markers
and resolution bandwidth 19-14, 19-19 X-axis, example 10-7
video demodulation 22-12, 23-12 markers, polar
FSK SEE polar markers
block diagram 22-6 measured signal
FTP, and LAN 14-10 video demod 23-10
measured signal, in digital demod 22-10
G measurement resolution 19-11
gating
measurement time 19-2, 19-10, 19-17
description 21-2
memory limitations 19-19
example 3-4, 4-2
mirrored spectrums 16-6, 23-17, 24-9
in analog demodulation 20-13
gaussian filters 22-19 N
GPIB NADC demodulation, example 8-4
LAN use 14-6 narrowband measurements 18-8
guardbands, analog demodulation 20-7 Network, Local Area
SEE LAN
H noise equivalent bandwidth 19-18
help (by phone) 2-i
noise, measuring 4-1
horizontal axis
normalization 10-4
SEE X-axis
HP/Agilent 71910A wideband receiver O
16-3 offset
video demod 9-13
I offset, in digital demod 8-8
pulse search 23-16
Option 080, 3GPP W-CDMA Analysis
information bandwidth 19-3, 19-10
24-3
IQ measured signal 22-10, 23-10
options, available a-v
IQ reference signal 22-10, 23-10
IS-95 22-20 P
peak power statistics
L displaying 5-4
LAN
peak tracking, example 3-4
activating 14-5
peak-to-average statistics
addressing 14-4
displaying 5-4
and FTP 14-10
phase noise, analysis example 2-1
GPIB with 14-6
phone assistance 2-i
interface, connecting 14-1
PM demodulation
X-Windows with 14-7
algorithm 20-10
leakage 19-6
example 2-3
linear averaging
points
SEE time averaging
SEE frequency points
lines
points-per-symbol, setting 10-4
SEE frequency points
polar markers
lo feedthrough
example 10-4
SEE zero response
units, example 10-4
Local Area Network
power calculation, digital
SEE LAN
demodulation 10-4
local oscillator 18-12
power measurements
displaying 5-1
Index

pulse length, in digital demod 8-6 SEE sweep time

pulse search search length
in digital demod 22-15 digital demod 8-8
setup example 8-6 video demod 9-13
Demodulation, video 23-16 search length, in digital demod 22-15
pulsed signals, in digital demod 22-15 service assistance (by phone) 2-i
slots (W-CDMA)
Q viewing 13-10
QAM source
block diagram, video demod 23-4 optional RF a-viii
QAM demodulation, example 10-1 span
SEE ALSO frequency span
R in analog demodulation 20-6
raised cosine filters 22-18
recalling video demodulation 22-12, 23-12
spectrogram displays 7-16 span, 8 MHz (opt. AYH) 23-18
waterfall displays 7-16 spectral displays 7-1
reference signal spectrogram
digital demod 22-10 displaying 7-1
video demod 23-10 spectrum analyzers
references comparisons 18-9
digital demod 22-2 types of 18-4
resolution square-root raised cosine filters 22-18
display 19-12 standard setups
frequency 19-11 digital demodulation, example 8-4
measurement 19-5 state definitions
resolution bandwidth SEE display state definitions
and gating 21-5 statistical power mesaurements
SEE ALSO bandwidth coupling displaying 5-1
digital demod 22-12 supplied accessories a-vi, a-ix
effect on noise 19-2 sweep speed
scalar limitations 19-11 SEE sweep time
vector limitations 19-18 sweep time
video demod 23-11, 24-15 described 19-10
resolution bandwidth (W-CDMA) 24-15 SEE ALSO measurement time
result length sweep, manual 19-13
digital demod 8-8, 22-15 symbol conventions
video demod 9-13 SEE display state definitions
root raised cosine filters 22-18 symbol rate 22-12, 23-12
symbol states, example 10-8
S symbol/error table, example 10-8
sample frequency sync pattern
SEE sample rate video demod, creating, saving,
sample rate 19-14 recalling 9-14
video demodulation 22-12, 23-12 video demod, user-defined 9-14
scalar measurements sync pattern, in digital demod
block diagrams 17-3 creating, saving, recalling 8-9
limitations 19-11 user-defined 8-9
overview 17-2 sync search
scaling example 8-8, 9-13
X-axis, example 10-7 in digital demod 22-14
scan time video demod 23-15
 Index

sync word block diagrams 17-5

digital demod 8-8 overview 17-4
video demod 9-13 video averaging 19-3
video demodulation
T SEE demodulation, video
T Frame 24-13 video filtering 19-3
T Slot 24-13 VSB
T Trigger 24-13 block diagram 23-6
telephone assistance 2-i center frequency, setting 9-4
time averaging 19-15
time capture W
playback example 2-2 W-CMDA signals, demodulating 13-2
recalling data from disk 2-2 waterfall
time corrections, analog demodulation displaying 7-1
20-5 wideband cdma
time data 19-15 about W-CDMA 24-1 - 24-18
time domain displays (W-CDMA) 24-13 block diagram 24-5
time domain overview 18-2 code channel, viewing 13-8
time gating code domain power 24-12
SEE gating code layer, viewing 13-6
time gating (W-CDMA) 24-14 demodulating 13-2, 13-4
time record 18-7, 19-16 error parameters 24-13
time record length feature availability 24-16
and memory 19-19 making wcdma measurements 13-1,
defined 19-16 24-6
limitations 19-19 max time pts, setting 24-15
time record size RBW 24-15
defined 19-16 slots, viewing 13-10
traces symbol table/error parameters 13-12
overlaid 8-14 time domain displays 24-13
trigger time gating 24-14
analog demodulation 20-13 troubleshooting W-CDMA
troubleshooting, W-CDMA 24-17 measurements 24-17
two-channels x-scale markers 13-14
digital demod 8-16 window bandwidth
video demod 9-15 defined 19-18
values 19-18
U windows
UFG (LAN and 4 MB memory option) window functions 19-6
14-2
UFG (LAN and 4 MB memory X
option)a-v a-v X-axis, scaling and markers 10-7
UG7 (advanced LAN option) a-v X-Windows, LAN use 14-7
user defined filters 22-19
UTH (LAN and 20 MB memory option) Z
a-v zero padding 19-20
zero response 19-9
V zero span measurements 20-2
vector averaging zoom measurements 18-8, 19-3
SEE time averaging
vector diagram, example 8-5, 9-12
vector measurements 18-8
Need Assistance?
If you need assistance, contact your nearest Agilent Technologies Sales and
Service Office listed in the Agilent Catalog. You can also find a list of local
service representatives on the Web at:
http://www.agilent.com/services/English/index.html or contact your nearest
regional office listed below.

If you are contacting Agilent Technologies about a problem with your Agilent
89410A Vector Signal Analyzer, please provide the following information:
q Model number: Agilent 89410A
q Serial number:
q Options:
q Date the problem was first encountered:
q Circumstances in which the problem was encountered:
q Can you reproduce the problem?
q What effect does this problem have on you?
You may find the serial number and options from the front panel of your
analyzer by executing the following:

Press [System Utility], [more], [serial number].

Press [System Utility], [options setup].

If you do not have access to the Internet, one of these centers can direct you to
your nearest representative:

United States Test and Measurement Call Center

(800) 452-4844 (Toll free in US)
Canada (905) 206-4725
Europe (31 20) 547 9900
Japan Measurement Assistance Center
(81) 426 56 7832
(81) 426 56 7840 (FAX)
Latin America (305) 267 4245
(305) 267 4288 (FAX)
Australia/New Zealand 1 800 629 485 (Australia)
0800 738 378 (New Zealand)
Asia-Pacific (852) 2599 7777
(FAX) (852) 2506 9285
Agilent 89400-Series Documentation Roadmap

If you are thinking about... And you want to... Then read the analyzer’s...

Unpacking and installing Install the analyzer, or do operation Installation and Verification Guide
the analyzer verification or performance verification
tests

Getting started Make your first measurements with Getting Started Guide
your new analyzer

Review measurement concepts Operator’s Guide

Learn what each key does Online Help (press the [ Help ] key)

Making measurements Learn how to make typical Getting Started Guide

measurements and Operator’s Guide

Creating automated Learn the Agilent Instrument BASIC Agilent 89400-Series

measurements interface Using Agilent Instrument BASIC

(To receive Instrument BASIC Program with Agilent Instrument Agilent Instrument BASIC User’s
and Agilent Instrument BASIC BASIC Handbook
manuals, order option 1C2)

Remote operation Learn about the GPIB and SCPI GPIB Programmer’s Guide

Find specific GPIB commands quickly Agilent 89400-Series

GPIB Commands: Quick Reference

Find GPIB command details Agilent 89400-Series

GPIB Command Reference

Using analyzer data with a PC Transfer analyzer data to or from a PC Standard Data Format Utilities:
application (Personal Computer) application User’s Guide

Display analyzer data on a PC, or

display PC data on the analyzer

Servicing the analyzer Adjust, troubleshoot, or repair the Service Guide

(To receive service information, analyzer
order option OB3)

89441-90067