No part of this manual may be
reproduced in any form or by any
means (including electronic storage
and retrieval or translation into a
foreign language) without prior
agreement and written consent from
Keysight Technologies, Inc. as
governed by United States and
international copyright laws.
Trademark Acknowledgments
Manual Part Number
N9040-90002
Edition
Edition 1, December 2020
S
upersedes: August 2020
Published by:
Keysight Technologies
1400 Fountaingrove Parkway
Santa Rosa, CA 95403
Warranty
THE MATERIAL CONTAINED IN THIS
DOCUMENT IS PROVIDED “AS IS,”
AND IS SUBJECT TO BEING
CHANGED, WITHOUT NOTICE, IN
FUTURE EDITIONS. FURTHER, TO
THE MAXIMUM EXTENT PERMITTED
BY APPLICABLE LAW, KEYSIGHT
DISCLAIMS ALL WARRANTIES,
EITHER EXPRESS OR IMPLIED WITH
REGARD TO THIS MANUAL AND
ANY INFORMATION CONTAINED
HEREIN, INCLUDING BUT NOT
LIMITED TO THE IMPLIED
WARRANTIES OF
MERCHANTABILITY AND FITNESS
FOR A PARTICULAR PURPOSE.
KEYSIGHT SHALL NOT BE LIABLE
FOR ERRORS OR FOR INCIDENTAL
OR CONSEQUENTIAL DAMAGES IN
CONNECTION WITH THE
FURNISHING, USE, OR
PERFORMANCE OF THIS
DOCUMENT OR ANY INFORMATION
CONTAINED HEREIN. SHOULD
KEYSIGHT AND THE USER HAVE A
SEPARATE WRITTEN AGREEMENT
WITH WARRANTY TERMS
COVERING THE MATERIAL IN THIS
DOCUMENT THAT CONFLICT WITH
THESE TERMS, THE WARRANTY
TERMS IN THE SEPARATE
AGREEMENT WILL CONTROL.
Technology Licenses
The hardware and/or software
described in this document are
furnished under a license and may be
used or copied only in accordance
with the terms of such license.
U.S. Government Rights
The Software is “commercial
computer software,” as defined
by Federal Acquisition Regulation
(“FAR”) 2.101. Pursuant to FAR
12.212 and 27.405-3 and
Department of Defense FAR
Supplement (“DFARS”) 227.7202,
the U.S. government acquires
commercial computer software
under the same terms by which
the software is customarily
provided to the public.
Accordingly, Keysight provides
the Software to U.S. government
customers under its standard
commercial license, which is
embodied in its End User License
Agreement (EULA), a copy of
which can be found at
http://www.keysight.com/find/sweula
The license set forth in the EULA
represents the exclusive authority
by which the U.S. government
may use, modify, distribute, or
disclose the Software. The EULA
and the license set forth therein,
does not require or permit,
among other things, that
Keysight: (1) Furnish technical
information related to
commercial computer software
or commercial computer
software documentation that is
not customarily provided to the
public; or (2) Relinquish to, or
otherwise provide, the
government rights in excess of
these rights customarily provided
to the public to use, modify,
reproduce, release, perform,
display, or disclose commercial
computer software or
commercial computer software
documentation. No additional
government requirements
beyond those set forth in the
EULA shall apply, except to the
extent that those terms, rights, or
licenses are explicitly required
from all providers of commercial
computer software pursuant to
the FAR and the DFARS and are
set forth specifically in writing
elsewhere in the EULA. Keysight
shall be under no obligation to
update, revise or otherwise
modify the Software. With
respect to any technical data as
defined by FAR 2.101, pursuant
to FAR 12.211 and 27.404.2 and
DFARS 227.7102, the U.S.
government acquires no greater
than Limited Rights as defined in
FAR 27.401 or DFAR 227.7103-5
(c), as applicable in any technical
data.
Safety Notices
A CAUTION notice denotes a hazard. It
calls attention to an operating
procedure, practice, or the like that,
if not correctly performed or adhered
to, could result in damage to the
product or loss of important data. Do
not proceed beyond a CAUTION
notice until the indicated conditions
are fully understood and met.
A WARNING notice denotes a hazard.
It calls attention to an operating
procedure, practice, or the like that,
if not correctly performed or adhered
to, could result in personal injury or
death. Do not proceed beyond a
WARNING notice until the indicated
conditions are fully understood and
met.
Page 3
Where to Find the Latest Information
Documentation is updated periodically. For the latest information about these products, including instrument software
upgrades, application information, and product information, browse to one of the following URLs, according to the name
of your product:
http://www.keysight.com/find/uxa
To receive the latest updates by email, subscribe to Keysight Email Updates at the following URL:
http://www.keysight.com/find/MyKeysight
Information on preventing instrument damage can be found at:
www.keysight.com/find/PreventingInstrumentRepair
Is your product software up-to-date?
Periodically, Keysight releases software updates to fix known defects and incorporate product enhancements. To search
for software updates for your product, go to the Keysight Technical Support website at:
This chapter contains the specifications for the core signal analyzer. The specifications and
characteristics for the measurement applications and options are covered in the chapters that
follow.
15
Page 16
UXA Signal Analyzer
Definitions and Requirements
Definitions and Requirements
This book contains signal analyzer specifications and supplemental information. The distinction
among specifications, typical performance, and nominal values are described as follows.
Definitions
— Specifications describe the performance of parameters covered by the product warranty
(temperature = 0 to 55°C also referred to as "Full temperature range" or "Full range", unless
otherwise noted).
— 95th percentile values indicate the breadth of the population (≈2σ) of performance tolerances
expected to be met in 95% of the cases with a 95% confidence, for any ambient temperature in
the range of 20 to 30°C. In addition to the statistical observations of a sample of instruments,
these values include the effects of the uncertainties of external calibration references. These
values are not warranted. These values are updated occasionally if a significant change in the
statistically observed behavior of production instruments is observed.
— Typical describes additional product performance information that is not covered by the product
warranty. It is performance beyond specification that 80% of the units exhibit with a 95%
confidence level over the temperature range 20 to 30°C. Typical performance does not include
measurement uncertainty.
— Nominal values indicate expected performance, or describe product performance that is useful
in the application of the product, but is not covered by the product warranty.
Conditions Required to Meet Specifications
The following conditions must be met for the analyzer to meet its specifications.
— The analyzer is within its calibration cycle. See the General section of this chapter.
— Under auto couple control, except that Auto Sweep Time Rules = Accy.
— For signal frequencies <10 MHz, DC coupling applied.
— Any analyzer that has been stored at a temperature range inside the allowed storage range but
outside the allowed operating range must be stored at an ambient temperature within the
allowed operating range for at least two hours before being turned on.
— The analyzer has been turned on at least 30 minutes with Auto Align set to Normal, or if Auto
Align is set to Off or Partial, alignments must have been run recently enough to prevent an Alert
message. If the Alert condition is changed from “Time and Temperature” to one of the disabled
duration choices, the analyzer may fail to meet specifications without informing the user. If Auto
Align is set to Light, performance is not warranted, and nominal performance will degrade to
become a factor of 1.4 wider for any specification subject to alignment, such as amplitude
tolerances.
Certification
Keysight Technologies certifies that this product met its published specifications at the time of
shipment from the factory. Keysight Technologies further certifies that its calibration
measurements are traceable to the International System of Units (SI) via national metrology
institutes (www.keysight.com/find/NMI) that are signatories to the CIPM Mutual Recognition
Arrangement.
16
Page 17
UXA Signal Analyzer
Frequency and Time
Frequency and Time
DescriptionSpecificationsSupplemental Information
Frequency Range
Maximum Frequency
Option 5088.4 GHz
Option 51313.6 GHz
Option 52626.5 GHz
Option 54444 GHz
Option 55050 GHz
Preamp Option P088.4 GHz
Preamp Option P1313.6 GHz
Preamp Option P2626.5 GHz
Preamp Option P4444 GHz
Preamp Option P5050 GHz
Minimum Frequency
PreampAC CoupledDC Coupled
Off10 MHz2 Hz
On10 MHz9 kHz
Band
0 (2 Hz to 3.6 GHz)
1 (3.5 to 8.4 GHz)1—1Options 508, 513, 526, 544, 550
2 (8.3 to 13.6 GHz)1—2Options 513, 526, 544, 550
3 (13.5 to 17.1 GHz)2—2Options 526, 544, 550
c
Harmonic
Mixing Mode
1—1Options 508, 513, 526, 544, 550
LO Multiple (N
a
)
Band Overlaps
b
4 (17.0 to 26.5 GHz)2—4Options 526, 544, 550
5 (26.4 to 34.5 GHz)2—4Options 544, 550
6 (34.4 to 50 GHz)4—8Options 544, 550
a. N is the LO multiplication factor. For negative mixing modes (as indicated by the “—” in the “Harmonic Mixing
Mode” column), the desired 1st LO harmonic is higher than the tuned frequency by the 1st IF.
17
Page 18
UXA Signal Analyzer
Frequency and Time
b. In the band overlap regions, for example, 3.5 to 3.6 GHz, the analyzer may use either band for measurements, in
this example Band 0 or Band 1. The analyzer gives preference to the band with the better overall specifications
(which is the lower numbered band for all frequencies below 26 GHz), but will choose the other band if doing so is
necessary to achieve a sweep having minimum band crossings. For example, with CF = 3.58 GHz, with a span of
40 MHz or less, the analyzer uses Band 0, because the stop frequency is 3.6 GHz or less, allowing a span without
band crossings in the preferred band. If the span is between 40 and 160 MHz, the analyzer uses Band 1, because
the start frequency is above 3.5 GHz, allowing the sweep to be done without a band crossing in Band 1, though
the stop frequency is above 3.6 GHz, preventing a Band 0 sweep without band crossing. With a span greater than
160 MHz, a band crossing will be required: the analyzer sweeps up to 3.6 GHz in Band 0; then executes a band
crossing and continues the sweep in Band 1.
Specifications are given separately for each band in the band overlap regions. One of these specifications is for the
preferred band, and one for the alternate band. Continuing with the example from the previous paragraph
(3.58 GHz), the preferred band is band 0 (indicated as frequencies under 3.6 GHz) and the alternate band is band
1 (3.5 to 8.4 GHz). The specifications for the preferred band are warranted. The specifications for the alternate
band are not warranted in the band overlap region, but performance is nominally the same as those warranted
specifications in the rest of the band. Again, in this example, consider a signal at 3.58 GHz. If the sweep has been
configured so that the signal at 3.58 GHz is measured in Band 1, the analysis behavior is nominally as stated in the
Band 1 specification line (3.5 to 8.4 GHz) but is not warranted. If warranted performance is necessary for this signal, the sweep should be reconfigured so that analysis occurs in Band 0. Another way to express this situation in
this example Band 0/Band 1 crossing is this: The specifications given in the “Specifications” column which are
described as “3.5 to 8.4 GHz” represent nominal performance from 3.5 to 3.6 GHz, and warranted performance
from 3.6 to 8.4 GHz.
c. Band 0 is extendable (set “Extend Low Band” to On) to 3.7 GHz instead of 3.6 GHz in instruments with frequency
option 508, 513 or 526 and with firmware of version A.16.17 or later.
18
Page 19
UXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Precision Frequency Reference
Accuracy±[(time since last adjustment x
aging rate) + temperature
stability + calibration accuracy
a]b
Temperature Stability
Full temperature range
Aging Rate
±4.5 × 10
−9
±2.5 × 10
−10
/day (nominal)
Total Aging
−8
1 Year
Settability
Warm-up and Retrace
c
300 s after turn on
600 s after turn on
Achievable Initial Calibration Accuracy
d
±3 × 10
±4 × 10
±3.1 × 10
−11
−8
Nominal
±1 × 10
±1 × 10
−7
of final frequency
−8
of final frequency
Standby powerStandby power is supplied to both
the CPU and the frequency
reference oscillator.
Residual FM
(Center Frequency = 1 GHz
≤0.25 Hz × N
(nominal)
e
p-p in 20 ms
10 Hz RBW, 10 Hz VBW)
a. Calibration accuracy depends on how accurately the frequency standard was adjusted to 10 MHz. If the adjust-
ment procedure is followed, the calibration accuracy is given by the specification “Achievable Initial Calibration
Accuracy.”
b. The specification applies after the analyzer has been powered on for four hours.
c. Standby mode applies power to the oscillator. Therefore warm-up and retrace only apply if the power connec-
tion is lost and restored. The warm-up reference is one hour after turning the power on. The effect of retracing
is included within the “Achievable Initial Calibration Accuracy” term of the Accuracy equation.
d. The achievable calibration accuracy at the beginning of the calibration cycle includes these effects:
1) Temperature difference between the calibration environment and the use environment
2) Orientation relative to the gravitation field changing between the calibration environment and the use envi-
ronment
3) Retrace effects in both the calibration environment and the use environment due to turning the instrument
power off.
4) Settability
e. N is the LO multiplication factor.
19
Page 20
UXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Frequency Readout Accuracy±(marker freq × freq ref accy +
a
+ 2 Hz +
b
)
Example for EMC
0.10%×span + 5% × RBW
0.5 × horizontal resolution
d
Single detector only
±0.0032% (nominal)
a. The warranted performance is only the sum of all errors under autocoupled conditions. Under non-autocoupled
conditions, the frequency readout accuracy will nominally meet the specification equation, except for conditions
in which the RBW term dominates, as explained in examples below. The nominal RBW contribution to frequency
readout accuracy is 2% of RBW for RBWs from 1 Hz to 390 kHz, 4% of RBW from 430 kHz through 3 MHz (the
widest autocoupled RBW), and 30% of RBW for the (manually selected) 4, 5, 6 and 8 MHz RBWs.
First example: a 120 MHz span, with autocoupled RBW. The autocoupled ratio of span to RBW is 106:1, so
the RBW selected is 1.1 MHz. The 5% × RBW term contributes only 55 kHz to the total frequency readout accuracy, compared to 120 kHz for the 0.10% × span term, for a total of 175 kHz.
Second example: a 20 MHz span, with a 4 MHz RBW. The specification equation does not apply because the
Span: RBW ratio is not autocoupled. If the equation did apply, it would allow 20 kHz of error (0.10%) due to the
span and 200 kHz error (5%) due to the RBW. For this non-autocoupled RBW, the RBW error is nominally 30%,
or 1200 kHz.
b. Horizontal resolution is due to the marker reading out one of the trace points. The points are spaced by
span/(Npts – 1), where Npts is the number of sweep points. For example, with the factory preset value of 1001
sweep points, the horizontal resolution is span/1000. However, there is an exception: When both the detector
mode is “normal” and the span > 0.25 × (Npts – 1) × RBW, peaks can occur only in even-numbered points, so
the effective horizontal resolution becomes doubled, or span/500 for the factory preset case. When the RBW is
autocoupled and there are 1001 sweep points, that exception occurs only for spans >750 MHz
c. Specifications apply to traces in most cases, but there are exceptions. Specifications always apply to the peak
detector. Specifications apply when only one detector is in use and all active traces are set to Clear Write. Specifications also apply when only one detector is in use in all active traces and the "Restart" key has been pressed
since any change from the use of multiple detectors to a single detector. In other cases, such as when multiple
simultaneous detectors are in use, additional errors of 0.5, 1.0 or 1.5 sweep points will occur in some detectors,
depending on the combination of detectors in use.
d. In most cases, the frequency readout accuracy of the analyzer can be exceptionally good. As an example, Key-
sight has characterized the accuracy of a span commonly used for Electro-Magnetic Compatibility (EMC) testing
using a source frequency locked to the analyzer. Ideally, this sweep would include EMC bands C and D and thus
sweep from 30 to 1000 MHz. Ideally, the analysis bandwidth would be 120 kHz at −6 dB, and the spacing of the
points would be half of this (60 kHz). With a start frequency of 30 MHz and a stop frequency of 1000.2 MHz and
a total of 16168 points, the spacing of points is ideal. The detector used was the Peak detector. The accuracy of
frequency readout of all the points tested in this span was with ±0.0032% of the span. A perfect analyzer with
this many points would have an accuracy of ±0.0031%
of span. Thus, even with this large number of display
points, the errors in excess of the bucket quantization limitation were negligible.
a. Instrument conditions: RBW = 1 kHz, gate time = auto (100 ms), S/N ≥ 50 dB, frequency = 1 GHz
b. If the signal being measured is locked to the same frequency reference as the analyzer, the specified count
accuracy is ±0.100 Hz under the test conditions of footnote a. This error is a noisiness of the result. It will
increase with noisy sources, wider RBWs, lower S/N ratios, and source frequencies > 1 GHz.
DescriptionSpecificationsSupplemental Information
Frequency Span
Range
Option 5080 Hz, 10 Hz to 8.4 GHz
Option 5130 Hz, 10 Hz to 13.6 GHz
Option 5260 Hz, 10 Hz to 26.5 GHz
Option 5440 Hz, 10 Hz to 44 GHz
Option 5500 Hz, 10 Hz to 50 GHz
Resolution2 Hz
Span Accuracy
Swept
FFT
±(0.1% × span + horizontal resolution
±(0.1% × span + horizontal resolution
a
)
a
)
a. Horizontal resolution is due to the marker reading out one of the sweep points. The points are spaced by
span/(Npts − 1), where Npts is the number of sweep points. For example, with the factory preset value of 1001
sweep points, the horizontal resolution is span/1000. However, there is an exception: When both the detector
mode is “normal” and the span > 0.25 × (Npts − 1) × RBW, peaks can occur only in even-numbered points, so
the effective horizontal resolution becomes doubled, or span/500 for the factory preset case. When the RBW is
auto coupled and there are 1001 sweep points, that exception occurs only for spans >750 MHz.
a. Delayed trigger is available with line, video, RF burst and external triggers.
b. Prior to A.19.28 software, zero span trigger delay was limited to -150 ms to 500 ms.
22
Page 23
UXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
TriggersAdditional information on some of the triggers and
gate sources
VideoIndependent of Display Scaling and Reference Level
Minimum settable level−170 dBmUseful range limited by noise
Maximum usable level
Highest allowed mixer level
a
+ 2 dB (nominal)
Detector and Sweep Type relationships
Sweep Type = Swept
Detector = Normal, Peak, Sample or
Negative Peak
Triggers on the signal before detection, which is
similar to the displayed signal
Detector = AverageTriggers on the signal before detection, but with a
single-pole filter added to give similar smoothing to
that of the average detector
Sweep Type = FFTTriggers on the signal envelope in a bandwidth
wider than the FFT width
RF Burst
Level Range
Level Accuracy
c
−40 to −10 dBm plus attenuation (nominal)
Absolute±2 dB + Absolute Amplitude Accuracy (nominal)
Relative±2 dB (nominal)
Bandwidth (−10 dB)
b
Most cases
d
>80 MHz (nominal)
Start Freq <300 MHz,
RF Burst Level Type = Absolute
a. The highest allowed mixer level depends on the IF Gain. It is nominally –10 dBm for Preamp Off and IF Gain =
Low.
b. Noise will limit trigger level range at high frequencies, such as above 15 GHz.
c. With positive slope trigger. Trigger level with negative slope is nominally 1 to 4 dB lower than positive slope.
d. Include RF Burst Level Type = Relative.
24
Page 25
UXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Gated Sweep
Gate MethodsGated LO
Gated Video
Gated FFT
Span RangeAny span
Gate Delay Range0 to 100.0 s
Gate Delay Settability4 digits, ≥100 ns
Gate Delay Jitter33.3 ns p-p (nominal)
Gate Length Range
(Except Method = FFT)
Gated FFT and Gated Video Frequency
and Amplitude Errors
1 μs to 5.0 sGate length for the FFT method is fixed at
1.83/RBW, with nominally 2% tolerance.
Nominally no additional error for gated
measurements when the Gate Delay is greater
than the MIN FAST setting
Gated LO Frequency Errors
Gate ≥ 10 μsNominally no additional error when the Gate
Delay is greater than the MIN FAST setting
1.0 μs ≤ Gate < 10 μsNominal error given by 100 ns × N × (Span/ST) × √(SpanPosition × ST / GateLength); see footnote
Gated LO Amplitude ErrorsNominally no additional error when the Gate
Delay is greater than the MIN FAST setting
Phase Noise EffectsGated LO method overrides the loop
configuration to force single loop in place of dual
loop.
Gate SourcesExternal 1
Pos or neg edge triggered
External 2
Line
RF Burst
Periodic
a
a. ST is sweep time; SpanPosition is the location of the on-screen signal, 0 being the left edge of the screen and 1
being the right edge. N is the harmonic mixing number.
DescriptionSpecificationsSupplemental Information
Number of Frequency Sweep Points
(buckets)
Factory preset1001
Range1 to 100,001Zero and non-zero spans
25
Page 26
UXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Resolution Bandwidth (RBW)
Range (−3.01 dB bandwidth)
Standard
1 Hz to 8 MHz
Bandwidths above 3 MHz are 4, 5, 6, and
8 MHz.
Bandwidths 1 Hz to 3 MHz are spaced at
10% spacing using the E24 series (24
per decade): 1.0, 1.1, 1.2, 1.3, 1.5, 1.6,
1.8, 2.0, 2.2, 2.4, 2.7, 3.0, 3.3, 3.6, 3.9,
4.3, 4.7, 5.1, 5.6, 6.2, 6.8, 7.5, 8.2, 9.1 in
each decade.
With Option B2X, B5X, or H1G and
Option RBE
a
10, 15, 20, 25, 30, 40, 50, 60, 70, 80,
100, 133, 150, 200, and 212 MHz, in
Spectrum Analyzer mode and zero span.
Power bandwidth accuracy
b
RBW RangeCF Range
1 Hz to 100 kHzAll±0.5% (0.022 dB)
110 kHz to 1.0 MHz< 3.6 GHz±1.0% (0.044 dB)
1.1 to 2.0 MHz< 3.6 GHz±0.07 dB (nominal)
2.2 to 3 MHz< 3.6 GHz0 to −0.2 dB (nominal)
4 to 8 MHz< 3.6 GHz0 to −0.4 dB (nominal)
Noise BW to RBW ratio
Accuracy (−3.01 dB bandwidth)
c
d
1.056 ±2% (nominal)
1 Hz to 1.3 MHz RBW±2% (nominal)
1.5 MHz to 3 MHz RBW
CF ≤ 3.6 GHz
CF > 3.6 GHz
±7% (nominal)
±8% (nominal)
4 MHz to 8 MHz RBW
CF ≤ 3.6 GHz
CF > 3.6 GHz
±15% (nominal)
±20% (nominal)
Selectivity (−60 dB/−3 dB)4.1:1 (nominal)
a. Option RBE enables wider bandwidth filters in zero span in the Signal Analyzer mode. Available detectors are
Peak+ and Average. VBW filtering is disabled. Minimum sweep time is the greater of 200 μS or 200ns/pt. The
filter shape is approximately square. Support for Average detector was first added in SW Version A.23.05.
26
Page 27
UXA Signal Analyzer
Frequency and Time
b. The noise marker, band power marker, channel power and ACP all compute their results using the power band-
width of the RBW used for the measurement. Power bandwidth accuracy is the power uncertainty in the results
of these measurements due only to bandwidth-related errors. (The analyzer knows this power bandwidth for
each RBW with greater accuracy than the RBW width itself, and can therefore achieve lower errors.) The warranted specifications shown apply to the Gaussian RBW filters used in swept and zero span analysis. There are
four different kinds of filters used in the spectrum analyzer: Swept Gaussian, Swept Flattop, FFT Gaussian and
FFT Flattop. While the warranted performance only applies to the swept Gaussian filters, because only they are
kept under statistical process control, the other filters nominally have the same performance.
c. The ratio of the noise bandwidth (also known as the power bandwidth) to the RBW has the nominal value and
tolerance shown. The RBW can also be annotated by its noise bandwidth instead of this 3 dB bandwidth. The
accuracy of this annotated value is similar to that shown in the power bandwidth
accuracy specification.
d. Resolution Bandwidth Accuracy can be observed at slower sweep times than auto-coupled conditions. Normal
sweep rates cause the shape of the RBW filter displayed on the analyzer screen to widen significantly. The true
bandwidth, which determines the response to impulsive signals and noise-like signals, is not affected by the
sweep rate.
DescriptionSpecificationSupplemental information
Analysis Bandwidth
a
With Option B25 (standard)25 MHz
With Option B4040 MHz
With Option B2X255 MHz
With Option B5X510 MHz
a. Analysis bandwidth is the instantaneous bandwidth available about a center frequency over which the input sig-
nal can be digitized for further analysis or processing in the time, frequency, or modulation domain.
27
Page 28
UXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Preselector Bandwidth
Mean Bandwidth at CF
a
Freq Option ≤ 526Freq Option>526
5 GHz58 MHz46 MHz
10 GHz57 MHz52 MHz
15 GHz59 MHz53MHz
20 GHz64 MHz55 MHz
25 GHz74 MHz56 MHz
35 GHz62 MHz
44 GHz70 MHz
Standard Deviation9% 7%
−3 dB Bandwidth−7.5% relative to −4 dB bandwidth, nominal
a. The preselector can have a significant passband ripple. To avoid ambiguous results, the −4 dB bandwidth is
characterized.
DescriptionSpecificationsSupplemental Information
Video Bandwidth (VBW)
RangeSame as Resolution Bandwidth range
plus wide-open VBW (labeled 50 MHz)
Accuracy±6% (nominal)
in swept mode and zero span
a. For FFT processing, the selected VBW is used to determine a number of averages for FFT results. That number is
chosen to give roughly equivalent display smoothing to VBW filtering in a swept measurement. For example, if
VBW = 0.1 × RBW, four FFTs are averaged to generate one result.
a
28
Page 29
UXA Signal Analyzer
Amplitude Accuracy and Range
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
Measurement Range
Preamp OffDisplayed Average Noise Level to +30 dBm
Preamp OnDisplayed Average Noise Level to +24 dBmOptions P08, P13, P26, P44, P50
Input Attenuation Range0 to 70 dB, in 2 dB steps
DescriptionSpecificationsSupplemental Information
Maximum Safe Input LevelApplies with or without preamp
(Options P08, P13, P26, P44, P50)
Average Total Power+30 dBm (1 W)
Peak Pulse Power
(≤10 μs pulse width,
≤1% duty cycle,
input attenuation ≥ 30 dB)
DC voltage
DC Coupled±0.2 Vdc
AC Coupled±100 Vdc
DescriptionSpecificationsSupplemental Information
Display Range
Log ScaleTen divisions displayed;
Linear ScaleTen divisions
DescriptionSpecificationsSupplemental Information
Marker Readout
+50 dBm (100 W)
0.1 to 1.0 dB/division in 0.1 dB steps, and
1 to 20 dB/division in 1 dB steps
Resolution
Log (decibel) units
Trace Averaging Off, on-screen0.01 dB
Trace Averaging On or remote0.001 dB
Linear units resolution≤1% of signal level (nominal)
29
Page 30
UXA Signal Analyzer
Amplitude Accuracy and Range
Frequency Response
DescriptionSpecificationsSupplemental Information
Frequency ResponseRefer to the footnote for
(Maximum error relative to
reference condition (50 MHz)
b
Mechanical attenuator only
Swept operation
c
, LNP offd,
,
Attenuation 10 dB)
Option 544 or 550 (mmW)
Option 508, 513, or 526 (μW)
20 to 30°CFull range95th Percentile (≈2σ)
3 Hz to 10 MHzxx±0.46 dB±0.54 dB
10 to 20 MHzx±0.35 dB±0.44 dB±0.19 dB
Band Overlaps on page 17.
Freq Option 526 only: Modes
above 18 GHz
a
10 to 20 MHz
20 to 50 MHz
20 to 50 MHz
e
x±0.46 dB±0.54 dB±0.20 dB
x±0.35 dB±0.44 dB±0.19 dB
x±0.35 dB±0.44 dB±0.20 dB
50 MHz to 3.6 GHzx±0.35 dB±0.44 dB±0.14 dB
50 MHz to 3.6 GHz
3.6 to 3.7 GHz (Band 0)x
3.5 to 5.2 GHz
3.5 to 5.2 GHz
5.2 to 8.4 GHz
5.2 to 8.4 GHz
8.3 to 13.6 GHz
8.3 to 13.6 GHz
13.5 to 17.1 GHz
13.5 to 17.1 GHz
gh
gh
gh
gh
gh
gh
gh
gh
x±0.35 dB±0.47 dB±0.16 dB
See note
x±1.5 dB±2.5 dB±0.50 dB
x±1.7 dB±3.5 dB±0.69 dB
x±1.5 dB±2.5 dB±0.42 dB
x±1.5 dB±2.5 dB±0.42 dB
x±2.0 dB±2.7 dB±0.51 dB
x±2.0 dB±2.5 dB±0.39 dB
x±2.0 dB±2.7 dB±0.57 dB
x±2.0 dB±2.7 dB±0.54 dB
f
17.0 to 22 GHz
17.0 to 22 GHz
gh
gh
x±2.0 dB±2.7 dB±0.65 dB
x±2.0 dB±2.8 dB±0.62 dB
30
Page 31
UXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
22.0 to 26.5 GHz
22.0 to 26.5 GHz
26.4 to 34.5 GHz
34.4 to 50 GHz
gh
gh
gh
gh
x±2.5 dB±3.7 dB±0.87 dB
x±2.5 dB±3.5 dB±0.59 dB
x±2.5 dB±3.6 dB±0.93 dB
x±3.2 dB±4.9 dB±1.28 dB
a. Signal frequencies above 18 GHz are prone to additional response errors due to modes in the Type-N connector
used. Only analyzers with frequency Option 526 that do not also have input connector Option C35 will have
these modes.With the use of Type-N to APC 3.5 mm adapter part number 1250-1744, there are nominally six
such modes. The effect of these modes with this connector are included within these specifications.
b. See the Electronic Attenuator (Option EA3) chapter for Frequency Response using the electronic attenuator.
c. For Sweep Type = FFT, add the RF flatness errors of this table to the IF Frequency Response errors. An additional
error source, the error in switching between swept and FFT sweep types, is nominally ±0.01 dB and is included
within the “Absolute Amplitude Error” specifications.
d. Refer to LNP Chapter for the frequency response specifications with LNP on.
e. Specifications apply with DC coupling at all frequencies. With AC coupling, specifications apply at frequencies of
50 MHz and higher. Statistical observations at 10 MHz and lower show that most instruments meet the specifica-
tions, but a few percent of instruments can be expected to have errors that, while within the specified limits, are
closer to those limits than the measurement uncertainty guardband, and thus are not warranted. The AC coupling
effect at 20 to 50 MH is negligible, but not warranted.
f. Band 0 is extendable (set “Extend Low Band” to On) to 3.7 GHz instead of 3.6 GHz in instruments with frequency
Option 508, 513 or 526 and with firmware of version A.16.17 or later. Subject to these conditions, statistical
observations show that performance nominally fits within the same range within the 3.6 to 3.7 GHz frequencies as
within the next lower specified frequency range, but is not warranted.
g. Specifications for frequencies >3.5 GHz apply for sweep rates ≤100 MHz/ms.
h. Preselector centering applied.
31
Page 32
UXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
IF Frequency Response
a
Freq Option 526 only: Modes above 18 GHz
b
(Demodulation and FFT
response relative to the
center frequency)
Center
Freq (GHz)
Span
(MHz)
c
Preselector
Max Errord Midwidth Error
(95th
Percentile)
Slope (dB/MHz)
(95th
Percentile)
RMSe
(nominal)
<3.6≤10±0.20 dB±0.12 dB±0.100.02 dB
≥3.6, ≤26.5≤10 On0.23 dB
≥3.6, ≤26.5≤10
Off
±0.25 dB±0.12 dB±0.100.02 dB
f
>26.5, ≤50≤10 On0.12 dB
f
>26.5, ≤50≤10
Off
±0.30 dB±0.12 dB±0.100.024 dB
a. The IF frequency response includes effects due to RF circuits such as input filters, that are a function of RF fre-
quency, in addition to the IF passband effects.
b. Signal frequencies above 18 GHz are prone to additional response errors due to modes in the Type-N connector
used. Only analyzers with frequency Option 526 that do not also have input connector Option C35 will have
these modes.With the use of Type-N to APC 3.5 mm adapter part number 1250-1744, there are nominally six such
modes. These modes cause nominally up to −0.35 dB amplitude change, with phase errors of nominally up to
±1.2°.
c. This column applies to the instantaneous analysis bandwidth in use. In the Spectrum Analyzer Mode, this would be
the FFT width.
d. The maximum error at an offset (f) from the center of the FFT width is given by the expression
± [Midwidth Error + (f × Slope)], but never exceeds ±Max Error. Here the Midwidth Error is the error at the center
frequency for a given FFT span. Usually, the span is no larger than the FFT width in which case the center of the
FFT width is the center frequency of the analyzer. When using the Spectrum Analyzer mode with an analyzer span
is wider than the FFT width, the span is made up of multiple concatenated FFT results, and thus has multiple centers of FFT widths; in this case the f in the equation is the offset from the nearest center. Performance is nominally
three times better at most center frequencies.
e. The “rms” nominal performance is the standard deviation of the response relative to the center frequency, inte-
grated across the span. This performance measure was observed at a center frequency in each harmonic mixing
band, which is representative of all center frequencies; it is not the worst case frequency.
f. Standard Option MPB is enabled.
32
Page 33
UXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
IF Phase LinearityDeviation from mean phase linearity
Freq Option 526 only: Modes above
a
18 GHz
Center Freq
(GHz)
Span
(MHz)
Preselector
Peak-to-peak
(nominal)
RMS (nominal)
≥0.02, <3.6≤10n/a0.14°0.032°
c
≥3.6≤10
Off
0.27°0.057°
≥3.6≤10On0.93°0.22°
a. Signal frequencies above 18 GHz are prone to additional response errors due to modes in the Type-N connector
used. Only analyzers with frequency Option 526 that do not also have input connector Option C35 will have
these modes.With the use of Type-N to APC 3.5 mm adapter part number 1250-1744, there are nominally six
such modes. These modes cause nominally up to −0.35 dB amplitude change, with phase errors of nominally up
to ±1.2°.
b. The listed performance is the standard deviation of the phase deviation relative to the mean phase deviation
from a linear phase condition, where the rms is computed across the span shown and over the range of center
frequencies shown.
c. Standard Option MPB is enabled.
b
33
Page 34
UXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
Absolute Amplitude Accuracy
At 50 MHz
20 to 30°C
Full temperature range
At all frequencies
20 to 30°C
Full temperature range
a
±0.24 dB
±0.13 dB
(95th percentile)
±0.28 dB
a
±(0.24 dB + frequency response)
±(0.28 dB + frequency response)
95th Percentile Absolute
Amplitude Accuracy
b
(Wide range of signal levels,
RBWs, RLs, etc.,
0.01 to 3.6 GHz)
Atten = 10 dB
Atten = 10, 20, 30, or 40 dB
±0.16 dB
±0.18 dB
Amplitude Reference Accuracy±0.05 dB (nominal)
Preamp On
c
±(0.36 dB + frequency response)
(P08, P13, P26, P44, P50)
a. Absolute amplitude accuracy is the total of all amplitude measurement errors, and applies over the following
subset of settings and conditions: 1 Hz ≤ RBW ≤ 1MHz; Input signal −10 to −50 dBm (details below); Input
attenuation 10 dB; span < 5 MHz (nominal additional error for span ≥ 5 MHz is 0.02 dB); all settings auto-coupled except Swp Time Rules = Accuracy; combinations of low signal level and wide RBW use VBW ≤ 30 kHz to
reduce noise. When using FFT sweeps, the signal must be at the center frequency.
This absolute amplitude accuracy specification includes the sum of the following individual specifications under
the conditions listed above: Scale Fidelity, Reference Level Accuracy, Display Scale Switching Uncertainty, Resolution Bandwidth Switching Uncertainty, 50 MHz Amplitude Reference Accuracy, and the accuracy with which
the instrument aligns its internal gains to the 50 MHz Amplitude Reference.
The only difference between signals within the range above –50 dBm and those signals below that level is the
scale fidelity. Our specifications and experience show no difference between signals above and below this level.
The only reason our Absolute Amplitude Uncertainty specification does not go below this level is that noise
detracts from our ability to verify the performance at all levels with acceptable test times and yields. So the performance is not warranted at lower levels, but we fully expect it to be the same.
34
Page 35
UXA Signal Analyzer
Amplitude Accuracy and Range
b. Absolute Amplitude Accuracy for a wide range of signal and measurement settings, covers the 95th percentile
proportion with 95% confidence. Here are the details of what is covered and how the computation is made:
The wide range of conditions of RBW, signal level, VBW, reference level and display scale are discussed in footnote a. There are 44 quasi-random combinations used, tested at a 50 MHz signal frequency. We compute the
95th percentile proportion with 95% confidence for this set observed over a statistically significant number of
instruments. Also, the frequency response relative to the 50 MHz response is characterized by varying the signal
across a large number of quasi-random verification frequencies that are chosen to not correspond with the frequency response adjustment frequencies. We again compute the 95th percentile proportion with 95% confidence for this set observed over a statistically significant number of instruments. We also compute the 95th
percentile accuracy of tracing the calibration of the 50 MHz absolute amplitude accuracy to a national standards organization. We also compute the 95th percentile accuracy of tracing the calibration of the relative frequency response to a national standards organization. We take the root-sum-square of these four independent
Gaussian parameters. To that rss we add the environmental effects of temperature variations across the 20 to
30°C range. These computations and measurements are made with the mechanical attenuator only in circuit,
set to the reference state of 10 dB.
A similar process is used for computing the result when using the electronic attenuator under a wide range of
settings: all even settings from 4 through 24 dB inclusive, with the mechanical attenuator set to 10 dB. The 95th
percentile result was 0.21 dB.
c. Same settings as footnote a, except that the signal level at the preamp input is −40 to −80 dBm. Total power at
preamp (dBm) = total power at input (dBm) minus input attenuation (dB). This specification applies for signal
frequencies above 100 kHz.
DescriptionSpecificationsSupplemental Information
Input Attenuation Switching UncertaintyRefer to the footnote for
Band Overlaps on page 17
(Relative to 10 dB (reference setting))
50 MHz (reference frequency), preamp off
Attenuation 12 to 40 dB±0.14 dB±0.04 dB (typical)
Attenuation 2 to 8 dB, or > 40 dB±0.18 dB±0.06 dB (typical)
Attenuation 0 dB±0.05 dB (nominal)
Attenuation >2 dB, preamp off
3 Hz to 3.6 GHz±0.3 dB (nominal)
3.5 to 8.4 GHz±0.5 dB (nominal)
8.3 to 13.6 GHz±0.7 dB (nominal)
13.5 to 26.5 GHz±0.7 dB (nominal)
26.5 to 50 GHz±1.0 dB (nominal)
35
Page 36
UXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
RF Input VSWR
(at tuned frequency, DC coupled)
10 dB atten. 50 MHz (ref condition)1.07:1 (nominal)
0 dB atten. 0.01 to 3.6 GHz< 2.2:1 (nominal)
95th Percentile
a
RF/μWmmW
Band 0 (0.01 to 3.6 GHz,10 dB atten)1.1011.116
Band 1 (3.5 to 8.4 GHz,10 dB atten)1.2781.144
Band 2 (8.3 to 13.6 GHz,10 dB atten)1.3411.158
Band 3 (13.5 to 17.1 GHz,10 dB atten)1.581.258
Band 4 (17.0 to 26.5 GHz,10 dB atten)1.5601.233
Band 5 (26.4 to 34.5 GHz,10 dB atten)1.363
Band 6 (34.4 to 50 GHz,10 dB atten)1.55
Nominal VSWR vs. Freq, 10 dBSee plots following
Atten. > 10 dBSimilar to atten. = 10
dB
RF Calibrator (e.g. 50 MHz) is OnOpen input
Alignments runningOpen input for some, unless "All but RF" is
selected
Preselector centeringOpen input
a. X-Series analyzers have a reflection coefficient that is excellently modeled with a Rayleigh probability distribu-
tion. Keysight recommends using the methods outlined in Application Note 1449-3 and companion Average
Power Sensor Measurement Uncertainty Calculator to compute mismatch uncertainty. Use this 95th percentile
VSWR information and the Rayleigh model (Case C or E in the application note) with that process.
36
Page 37
Nominal VSWR Band [Plot]
UXA Signal Analyzer
Amplitude Accuracy and Range
37
Page 38
Nominal VSWR, above 3.5 GHz [Plot]
UXA Signal Analyzer
Amplitude Accuracy and Range
38
Page 39
UXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
Resolution Bandwidth Switching UncertaintyRelative to reference BW of 30 kHz,
verified in low band
a
1.0 Hz to 1.5 MHz RBW±0.03 dB
1.6 MHz to 2.7 MHz RBW±0.05 dB
3.0 MHz RBW±0.10 dB
Manually selected wide RBWs: 4, 5, 6, 8 MHz±0.30 dB
a. RBW switching uncertainty is verified at 50 MHz. It is consistent for all measurements made without the prese-
lector, thus in Band 0 and also in higher bands with the Preselector Bypass option. In preselected bands, the
slope of the preselector passband can interact with the RBW shape to make an apparent additional RBW
switching uncertainty of nominally ±0.05 dB/MHz times the RBW.
DescriptionSpecificationsSupplemental Information
Reference Level
Range
Log Units −170 to +30 dBm, in 0.01 dB steps
Linear Units 707 pV to 7.07 V, with 0.01 dB resolution (0.11%)
Accuracy
0 dB
a
a. Because reference level affects only the display, not the measurement, it causes no additional error in measure-
ment results from trace data or markers.
DescriptionSpecificationsSupplemental Information
Display Scale Switching Uncertainty
a
Switching between Linear and Log
Log Scale Switching
0 dB
0 dB
a
a. Because Log/Lin and Log Scale Switching affect only the display, not the measurement, they cause no addi-
tional error in measurement results from trace data or markers.
39
Page 40
UXA Signal Analyzer
3
σ
320dB()110
SN⁄3dB+()20dB⁄()–
+log=
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
Display Scale Fidelity
ab
Absolute Log-Linear Fidelity
(Relative to the reference condition: −25 dBm
input through 10 dB attenuation, thus
−35 dBm at the input mixer)
Input mixer level
c
Linearity
Typ ical
−18 dBm ≤ ML ≤−10 dBm±0.10 dB±0.04 dB
ML < −18 dBm±0.07 dB±0.02 dB
Relative Fidelity
d
Applies for mixer levelc range from −10 to
−80 dBm, mechanical attenuator only,
preamp off, and dither on.
Sum of the following terms:Nominal
high level term
instability term
slope term
prefilter term
Up to ±0.015 dB
0.0019 dBrms
From equation
Up to ±0.005 dB
e
f
g
h
a. Supplemental information: The amplitude detection linearity specification applies at all levels below −10 dBm at
the input mixer; however, noise will reduce the accuracy of low level measurements. The amplitude error due to
noise is determined by the signal-to-noise ratio, S/N. If the S/N is large (20 dB or better), the amplitude error
due to noise can be estimated from the equation below, given for the 3-sigma (three standard deviations) level.
The errors due to S/N ratio can be further reduced by averaging results. For large S/N (20 dB or better), the
3-sigma level can be reduced proportional to the square root of the number of averages taken.
b. The scale fidelity is warranted with ADC dither set to Medium. Dither increases the noise level by nominally only
0.28 dB for the most sensitive case (preamp Off, best DANL frequencies). With dither Off, scale fidelity for low
level signals, around −60 dBm or lower, will nominally degrade by 0.2 dB. Dither High will give exceptional linear relative scale fidelity, but increase DANL by 0.63 dB instead of 0.28 dB.
c. Mixer level = Input Level − Input Attenuation
d. The relative fidelity is the error in the measured difference between two signal levels. It is so small in many cases
that it cannot be verified without being dominated by measurement uncertainty of the verification. Because of
this verification difficulty, this specification gives nominal performance, based on numbers that are as conservatively determined as those used in warranted specifications. We will consider one example of the use of the error
equation to compute the nominal performance.
Example: the accuracy of the relative level of a sideband around −60 dBm, with a carrier at −5dBm, using attenuation = 10 dB, RBW = 3 kHz, evaluated with swept analysis. The high level term is evaluated with P1 =
−15 dBm and P2 = −70 dBm at the mixer. This gives a maximum error within ±0.008 dB. The instability term is
±0.0019 dB if the measurement is completed within a minute. The slope term evaluates to ±0.022 dB. The pre-filter term applies and evaluates to the limit of ±0.005 dB. The sum of all these terms is ±0.037 dB.
40
Page 41
UXA Signal Analyzer
Amplitude Accuracy and Range
e. Errors at high mixer levels will nominally be well within the range of ±0.015 dB × {exp[(P1 − Pref)/(8.69 dB)] −
exp[(P2 − Pref)/(8.69 dB)]} (exp is the natural exponent function, e
x
). In this expression, P1 and P2 are the pow-
ers of the two signals, in decibel units, whose relative power is being measured. Pref is −10 dBm (−10 dBm is
the highest power for which linearity is specified). All these levels are referred to the mixer level.
f. The stability of the analyzer gain can be an error term of importance when no settings have changed. These have
been studied carefully in the UXA. One source of instability is the variation in analyzer response with time when
fully warmed up in a stable lab environment. This has been observed to be well modeled as a random walk process, where the difference in two measurements spaced by time t is given by a × sqrt(t), where a is
0.0019 dBrms per root minute. The other source of instability is updated alignments from running full or partial
alignments in the background or invoking an alignment. Invoked alignments (Align Now, All) have a standard
deviation of 0.0018 dB, and performing these will restart the random walk behavior. Partial alignments (Auto
Align set to "Partial") have a standard deviation that is, coincidentally, also 0.0018 dBrms, and only occurs once
every ten minutes. The standard deviation from full background alignment (Auto Align set to "Normal") is 0.015
dBrms; with these alignments on, there is no additional random walk behavior. (Keysight recommends setting
alignments (Auto Align) to Normal in order to make the best measurements over long periods of time or in environments without very high temperature stability. For short term measurements in highly stable environments,
setting alignments to Partial can give the best stability. Setting Alignments to Off is not recommended where
stability matters.)
g. Slope error will nominally be well within the range of ±0.0004 × (P1 − P2). P1 and P2 are defined in footnote e.
h. A small additional error is possible. In FFT sweeps, this error is possible for spans under 4.01 kHz. For non-FFT
measurements, it is possible for RBWs of 3.9 kHz or less. The error is well within the range of ±0.0021 × (P1 P2) subject to a maximum of ±0.005 dB. (The maximum dominates for all but very small differences.) P1 and P2
are defined in footnote e.
DescriptionSpecificationsSupplemental Information
Available DetectorsNormal, Peak, Sample, Negative Peak,
Average
Average detector works on RMS,
Voltage and Logarithmic scales
41
Page 42
UXA Signal Analyzer
Dynamic Range
Dynamic Range
Gain Compression
DescriptionSpecificationsSupplemental Information
1 dB Gain Compression Point
(Two-tone)
20 to 40 MHz+2 dBm (nominal)
40 to 3.6 GHz+5 dBm (nominal)
3.6 to 26.5 GHz+10 dBm (nominal)
26.5 to 50 GHz0 dBm (nominal)
Clipping (ADC Over-range)
Any signal offset−10 dBm
Signal offset > 5 times IF prefilter bandwidth and IF
Gain set to Low
IF Prefilter Bandwidth
Zero Span orSweep Type = FFT,–3 dB Bandwidth
abc
Maximum power at mixer
LNP off
Low frequency exceptions
+12 dBm (nominal)
d
,
e
f
Swept
≤3.9 kHz<4.01 kHz8.9 kHz
4.3 to 27 kHz<28.81 kHz79 kHz
30 to 160 kHz<167.4 kHz303 kHz
180 to 390 kHz<411.9 kHz966 kHz
430 kHz to 8 MHz<7.99 MHz10.9 MHz
, RBW =
a. Large signals, even at frequencies not shown on the screen, can cause the analyzer to incorrectly measure
on-screen signals because of two-tone gain compression. This specification tells how large an interfering signal
must be in order to cause a 1 dB change in an on-screen signal.
b. Specified at 1 kHz RBW with 100 kHz tone spacing. The compression point will nominally equal the specification
for tone spacing greater than 5 times the prefilter bandwidth. At smaller spacings, ADC clipping may occur at a
level lower than the 1 dB compression point.
FFT Width =(nominal)
42
Page 43
UXA Signal Analyzer
Dynamic Range
c. Reference level and off-screen performance: The reference level (RL) behavior differs from some earlier analyz-
ers in a way that makes this analyzer more flexible. In other analyzers, the RL controlled how the measurement
was performed as well as how it was displayed. Because the logarithmic amplifier in these analyzers had both
range and resolution limitations, this behavior was necessary for optimum measurement accuracy. The logarithmic amplifier in this signal analyzer, however, is implemented digitally such that the range and resolution greatly
exceed other instrument limitations. Because of this, the analyzer can make measurements largely independent
of the setting of the RL without compromising accuracy. Because the RL becomes a display function, not a measurement function, a marker can read out results that are off-screen, either above or below, without any change
in accuracy. The only exception to the independence of RL and the way in which the measurement is performed
is in the input attenuation setting: When the input attenuation is set to auto, the rules for the determination of
the input attenuation include dependence on the reference level. Because the input attenuation setting controls
the tradeoff between large signal behaviors (third-order intermodulation, compression, and display scale fidelity) and small signal effects (noise), the measurement results can change with RL changes when the input attenuation is set to auto.
d. Mixer power level (dBm) = input power (dBm) − input attenuation (dB).
e. The ADC clipping level declines at low frequencies (below 50 MHz) when the LO feedthrough (the signal that
appears at 0 Hz) is within 5 times the prefilter bandwidth (see table) and must be handled by the ADC. For
example, with a 300 kHz RBW and prefilter bandwidth at 966 kHz, the clipping level reduces for signal frequencies below 4.83 MHz. For signal frequencies below 2.5 times the prefilter bandwidth, there will be additional
reduction due to the presence of the image signal (the signal that appears at the negative of the input signal frequency) at the ADC.
f. This table applies without Option FS1 or FS2, fast sweep. With Option FS1or FS2, which is a standard
option in the UXA, this table applies for sweep rates that are manually chosen to be the same as or slower than
"traditional" sweep rates, instead of the much faster sweep rates, such as autocoupled sweep rates, available
with FS1or FS2. Sweep rate is defined to be span divided by sweep time. If the sweep rate is ≤ 1.1 times
RBW-squared, the table applies. Otherwise, compute an "effective RBW" = Span / (SweepTime × RBW). To
determine the IF Prefilter Bandwidth, look up this effective RBW in the table instead of the actual RBW. For
example, for RBW = 3 kHz, Span = 300 kHz, and Sweep time = 42 ms, we compute that Sweep Rate = 7.1
MHz/s, while RBW-squared is 9 MHz/s. So the Sweep Rate is < 1.1 times RBW-squared and the table applies;
row 1 shows the IF Prefilter Bandwidth is nominally 8.9 kHz. If the sweep time is 1 ms, then the effective RBW
computes to 100 kHz. This would result in an IF Prefilter Bandwidth from the third row, nominally 303 kHz.
43
Page 44
UXA Signal Analyzer
Dynamic Range
Displayed Average Noise Level
DescriptionSpecificationsSupplemental Information
Displayed Average Noise
Level (DANL) without
Noise Floor Extension
(mmW)
Option 544 or 550LNP
3 to 10 Hzx–95 dBm (nominal)
10 to 100 Hzx–114 dBm (nominal)
100 Hz to 1 kHzx–128 dBm (nominal)
1 to 9 kHzx–136 dBm (nominal)
9 to 100 kHzx−141 dBm−141 dBm−144 dBm
100 kHz to 1 MHzx−150 dBm−150 dBm−154 dBm
1 to 10 MHz
10 MHz to 1.2 GHzx−153 dBm−152 dBm−155 dBm
1.2 to 2.1 GHzx−151 dBm−150 dBm−153 dBm
a
LNP on20 to 30°CFull range Typical
off
b
x−154 dBm−153 dBm−156 dBm
Input terminated
Sample or Average detector
Averaging type = Log
0 dB input attenuation
IF Gain = High
1 Hz Resolution Bandwidth
Refer to the footnote for
Band Overlaps on
page 17.
2.1 to 3 GHzx−150 dBm−149 dBm−152 dBm
3.0 to 3.6 GHzx−149 dBm−148 dBm−151 dBm
3.5 to 4.2 GHzx−145 dBm−142 dBm−148 dBm
3.5 to 4.2 GHzx−151 dBm−149 dBm−154 dBm
4.2 to 6.6 GHzx−144 dBm−142 dBm−148 dBm
4.2 to 6.6 GHzx−152 dBm−150 dBm−154 dBm
6.6 to 8.4 GHzx−147 dBm−145 dBm−149 dBm
6.6 to 8.4 GHzx−153 dBm−151 dBm−155 dBm
8.3 to 13.6 GHzx−147 dBm−145 dBm−149 dBm
8.3 to 13.6 GHzx−153 dBm−151 dBm−155 dBm
13.5 to 14 GHzx−144 dBm−142 dBm−148 dBm
13.5 to 14 GHzx−150 dBm
−148 dBm−153 dBm
44
Page 45
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
14 to 17 GHzx−145 dBm−143 dBm−148 dBm
14 to 17 GHzx−151 dBm−149 dBm−153 dBm
17 to 22.5 GHzx−141 dBm−139 dBm−146 dBm
17 to 22.5 GHzx−149 dBm−147 dBm−152 dBm
22.5 to 26.5 GHzx−139 dBm−137 dBm−143 dBm
22.5 to 26.5 GHzx−146 dBm−145 dBm−150 dBm
26.4 to 30 GHzx−138 dBm−136 dBm−143 dBm
26.4 to 30 GHzx−146 dBm−144 dBm−150 dBm
30 to 34 GHzx−138 dBm−135 dBm−143 dBm
30 to 34 GHzx−146 dBm−144 dBm−150 dBm
33.9 to 37 GHzx−134 dBm−131 dBm−140 dBm
33.9 to 37 GHzx−142 dBm−139 dBm−148 dBm
37 to 40 GHzx−132 dBm−129 dBm−139 dBm
37 to 46 GHzx−141 dBm−138 dBm−146 dBm
40 to 49 GHzx−130 dBm−126 dBm−137 dBm
46 to 50 GHzx−139 dBm−136 dBm−145 dBm
49 to 50 GHzx−128 dBm−124 dBm−135 dBm
b
Additional DANL, IF Gain = Low
xx–164.5 dBm (nominal)
a. DANL for zero span and swept is measured in a 1 kHz RBW and normalized to the narrowest available RBW,
because the noise figure does not depend on RBW and 1 kHz measurements are faster.
b. Setting the IF Gain to Low is often desirable in order to allow higher power into the mixer without overload, better
compression and better third-order intermodulation. When the Swept IF Gain is set to Low, either by auto coupling or manual coupling, there is noise added above that specified in this table for the IF Gain = High case. That
excess noise appears as an additional noise at the input mixer. This level has sub-decibel dependence on center
frequency. To find the total displayed average noise at the mixer for Swept IF Gain = Low, sum the powers of the
DANL for IF Gain = High with this additional DANL. To do that summation, compute DANLtotal = 10 × log
(10^(DANLhigh/10) + 10^(AdditionalDANL / 10)). In FFT sweeps, the same behavior occurs, except that FFT IF
Gain can be set to autorange, where it varies with the input signal level, in addition to forced High and Low settings.
45
Page 46
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Displayed Average Noise Level
(DANL) without Noise Floor
Extension (RF/µW)
a
Option 508, 513, or 526LNP
off
3 to 10 Hz
10 to 100 Hz
100 Hz to 1 kHz
1 to 9 kHz
9 to 100 kHz
100 kHz to 1 MHz
1 to 10 MHz
b
10 MHz to 1.2 GHz
1.2 to 2.1 GHz
x–100 dBm (nominal)
x–125 dBm (nominal)
x–130 dBm (nominal)
x–137 dBm (nominal)
x−141 dBm−141 dBm−146 dBm
x−150 dBm−150 dBm−155 dBm
x−155 dBm−152 dBm−157 dBm
x−155 dBm−153 dBm−156 dBm
x−153 dBm−152 dBm−155 dBm
Input terminated
Sample or Average detector
Averaging type = Log
Refer to the footnote for
Band Overlaps on
page 17.
0 dB input attenuation
IF Gain = High
1 Hz Resolution Bandwidth
LNP on20 to 30°CFull range Typical
2.1 to 3 GHz
3.0 to 3.6 GHz
3.5 to 4.2 GHz
3.6 to 4.2 GHz
3.6 to 3.7 GHz
x−152 dBm−151 dBm−153 dBm
x−151 dBm−149 dBm−152 dBm
x−149 dBm−147 dBm−152 dBm
x−154 dBm−152 dBm−155 dBm
x
See note
4.2 to 8.4 GHzx−150 dBm−148 dBm−152 dBm
4.2 to 8.4 GHz
8.3 to 13.6 GHz
8.3 to 13.6 GHz
13.5 to 16.9 GHz
13.5 to 16.9 GHz
16.9 to 20 GHz
16.9 to 20 GHz
x−149 dBm−147 dBm−151 dBm
x−145 dBm−143 dBm−147 dBm
x−143 dBm−140 dBm−146 dBm
x−155 dBm−153 dBm−156 dBm
x−155 dBm−153 dBm−156 dBm
x−152 dBm−150 dBm−155 dBm
x−151 dBm−149 dBm−154 dBm
c
46
Page 47
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
20.0 to 26.5 GHz
20.0 to 26.5 GHz
Additional DANL, IF Gain = Low
x−136 dBm−133 dBm−139 dBm
x−148 dBm−146 dBm−151 dBm
d
xx–164.5 dBm (nominal)
a. DANL for zero span and swept is measured in a 1 kHz RBW and normalized to the narrowest available RBW,
because the noise figure does not depend on RBW and 1 kHz measurements are faster.
b. DANL below 10 MHz is affected by phase noise around the LO feedthrough signal. Specifications apply with the
best setting of the Phase Noise Optimization control, which is to choose the “Best Close-in φ Noise" for frequencies below about 150 kHz, and “Best Wide Offset φ Noise" for frequencies above about 150 kHz.
c. Band 0 is extendable (set “Extend Low Band” to On) to 3.7 GHz instead of 3.6 GHz in instruments with frequency
option 508, 513 or 526 and with firmware of version A.16.17 or later. Subject to these conditions, statistical
observations show that performance nominally fits within the same range within the 3.6 to 3.7 GHz frequencies
as within the next lower specified frequency range, but is not warranted.
d. Setting the IF Gain to Low is often desirable in order to allow higher power into the mixer without overload, bet-
ter compression and better third-order intermodulation. When the Swept IF Gain is set to Low, either by auto
coupling or manual coupling, there is noise added above that specified in this table for the IF Gain = High case.
That excess noise appears as an additional noise at the input mixer. This level has sub-decibel dependence on
center frequency. To find the total displayed average noise at the mixer for Swept IF Gain = Low, sum the powers
of the DANL for IF Gain = High with this additional DANL. To do that summation, compute DANLtotal = 10 × log
(10^(DANLhigh/10) + 10^(AdditionalDANL / 10)). In FFT sweeps, the same behavior occurs, except that FFT IF
Gain can be set to autorange, where it varies with the input signal level, in addition to forced High and Low settings.
47
Page 48
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Displayed Average Noise
95th Percentile (≈2σ)b
Level with Noise Floor
Extension Improvement
(mmW)
Option 544 or 550Preamp Off
Band 0, f > 20 MHz
a
Preamp On
d
10 dB9 dBn/a
c
LNP On
Band 18 dB9 dB9 dB
Band 28 dB8 dB9 dB
Band 39 dB8 dB10 dB
Band 410 dB8 dB11 dB
Band 511 dB8 dB11 dB
Band 611 dB7 dB11 dB
Improvement for CW Signals
Improvement, Pulsed-RF Signals
e
f
3.5 dB (nominal)
10.8 dB (nominal)
Improvement, Noise-Like Signals9.1 dB (nominal)
a. This statement on the improvement in DANL is based on the statistical observations of the error in the effective
noise floor after NFE is applied. That effective noise floor can be a negative or a positive power at any frequency.
These 95th percentile values are based on the absolute value of that effective remainder noise power.
b. Unlike other 95th percentiles, these table values do not include delta environment effects. NFE is aligned in the
factory at room temperature. For best performance, in an environment that is different from room temperature,
such as an equipment rack with other instruments, we recommend running the "Characterize Noise Floor"
operation after the first time the analyzer has been installed in the environment, and given an hour to stabilize.
c. DANL of the preamp is specified with a 50Ω source impedance. Like all amplifiers, the noise varies with the
source impedance. When NFE compensates for the noise with an ideal source impedance, the variation in the
remaining noise level with the actual source impedance is greatly multiplied in a decibel sense.
d. NFE does not apply to the low frequency sensitivity. At frequencies below about 1 MHz, the sensitivity is domi-
nated by phase noise surrounding the LO feedthrough. The NFE is not designed to improve that performance. At
frequencies between 1 and 20 MHz the NFE effectiveness increases from nearly none to near its maximum.
e. Improvement in the uncertainty of measurement due to amplitude errors and variance of the results is modestly
improved by using NFE. The nominal improvement shown was evaluated for a 2 dB error with 250 traces averaged. For extreme numbers of averages, the result will be as shown in the "Improvement for Noise-like Signals"
and DANL sections of this table.
f. Pulsed-RF signals are usually measured with peak detection. Often, they are also measured with many “max hold”
traces. When the measurement time in each display point is long compared to the reciprocal of the RBW, or the
number of traces max held is large, considerable variance reduction occurs in each measurement point. When
the variance reduction is large, NFE can be quite effective; when it is small, NFE has low effectiveness. For example, in Band 0 with 100 pulses per trace element, in order to keep the error within ±3 dB error 95% of the time,
the signal can be 10.8 dB lower with NFE than without NFE.
48
Page 49
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Displayed Average Noise
95th Percentile (≈2σ)b
Level with Noise Floor
Extension Improvement
(RF/µW)
Option 508, 513, or 526Preamp Off
Band 0, f > 20 MHz
a
c
Preamp On
d
9 dB10 dB
LNP On
n/a
Band 110 dB9 dB10 dB
Band 210 dB10 dB10 dB
Band 39 dB9 dB10 dB
Band 49 dB8 dB9 dB
Improvement for CW Signals
Improvement, Pulsed-RF Signals
e
f
3.5 dB (nominal)
10.8 dB (nominal)
Improvement, Noise-Like Signals9.1 dB (nominal)
a. This statement on the improvement in DANL is based on the statistical observations of the error in the effective
noise floor after NFE is applied. That effective noise floor can be a negative or a positive power at any frequency.
These 95th percentile values are based on the absolute value of that effective remainder noise power.
b. Unlike other 95th percentiles, these table values do not include delta environment effects. NFE is aligned in the
factory at room temperature. For best performance, in an environment that is different from room temperature,
such as an equipment rack with other instruments, we recommend running the "Characterize Noise Floor"
operation after the first time the analyzer has been installed in the environment, and given an hour to stabilize.
c. DANL of the preamp is specified with a 50Ω source impedance. Like all amplifiers, the noise varies with the
source impedance. When NFE compensates for the noise with an ideal source impedance, the variation in the
remaining noise level with the actual source impedance is greatly multiplied in a decibel sense.
d. NFE does not apply to the low frequency sensitivity. At frequencies below about 1 MHz, the sensitivity is domi-
nated by phase noise surrounding the LO feedthrough. The NFE is not designed to improve that performance. At
frequencies between 1 and 20 MHz the NFE effectiveness increases from nearly none to near its maximum.
e. Improvement in the uncertainty of measurement due to amplitude errors and variance of the results is modestly
improved by using NFE. The nominal improvement shown was evaluated for a 2 dB error with 250 traces averaged. For extreme numbers of averages, the result will be as shown in the "Improvement for Noise-like Signals"
and DANL sections of this table.
f. Pulsed-RF signals are usually measured with peak detection. Often, they are also measured with many “max hold”
traces. When the measurement time in each display point is long compared to the reciprocal of the RBW, or the
number of traces max held is large, considerable variance reduction occurs in each measurement point. When
the variance reduction is large, NFE can be quite effective; when it is small, NFE has low effectiveness. For example, in Band 0 with 100 pulses per trace element, in order to keep the error within ±3 dB error 95% of the time,
the signal can be 10.8 dB lower with NFE than without NFE.
49
Page 50
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Displayed Average Noise
95th Percentile (≈2σ)
b
Level with Noise Floor
Extension (mmW)
Option 544 or 550Preamp Off
Band 0, f >20 MHz
a
Preamp On
e
−163 dBm−174 dBmn/a
cd
LNP On
Band 1 −157 dBm−173 dBm−163 dBm
Band 2 −159 dBm−174 dBm−164 dBm
Band 3−160 dBm−174 dBm−164 dBm
Band 4−155 dBm−171 dBm−163 dBm
Band 5−156 dBm−169 dBm−162 dBm
Band 6−148 dBm−161 dBm−156 dBm
a. DANL with NFE is unlike DANL without NFE. It is based on the statistical observations of the error in the effec-
tive noise floor after NFE is applied. That effective noise floor can be a negative or a positive power at any frequency. These 95th percentile values are based on the absolute value of that effective remainder noise power.
b. Unlike other 95th percentiles, these table values do not include delta environment effects. NFE is aligned in the
factory at room temperature. For best performance, in an environment that is different from room temperature,
such as an equipment rack with other instruments, we recommend running the "Characterize Noise Floor"
operation after the first time the analyzer has been installed in the environment, and given an hour to stabilize.
c. DANL of the preamp is specified with a 50Ω source impedance. Like all amplifiers, the noise varies with the
source impedance. When NFE compensates for the noise with an ideal source impedance, the variation in the
remaining noise level with the actual source impedance is greatly multiplied in a decibel sense.
d. NFE performance can give results below theoretical levels of noise in a termination resistor at room tempera-
ture, about –174 dBm/Hz. this is intentional and usually desirable. NFE is not designed to report the noise at
the input of the analyzer; it reports how much more noise is at the input of the analyzer than was present in its
alignment. And its alignment includes the noise of a termination at room temperature. So it can often see the
added noise below the theoretical noise. Furthermore, DANL is defined with log averaging in a 1 Hz RBW, which
is about 2.3 dB lower than the noise density (power averaged) in a 1 Hz noise bandwidth.
e. NFE does not apply to the low frequency sensitivity. At frequencies below about 1 MHz, the sensitivity is domi-
nated by phase noise surrounding the LO feedthrough. The NFE is not designed to improve that performance.
At frequencies between 1 and 20 MHz the NFE effectiveness increases from nearly none to near its maximum.
50
Page 51
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Displayed Average Noise
95th Percentile (≈2σ)b
Level with Noise Floor
Extension (RF/µW)
Option 508, 513, or 526Preamp OffPreamp
Band 0, f >20 MHz
a
LNP On
cd
On
e
−163 dBm−174 dBmn/a
Band 1 −162 dBm−174 dBm−166 dBm
Band 2 −162 dBm−174 dBm−167 dBm
Band 3−159 dBm−172 dBm−165 dBm
Band 4−148 dBm−166 dBm−162 dBm
a. DANL with NFE is unlike DANL without NFE. It is based on the statistical observations of the error in the effective
noise floor after NFE is applied. That effective noise floor can be a negative or a positive power at any frequency.
These 95th percentile values are based on the absolute value of that effective remainder noise power.
b. Unlike other 95th percentiles, these table values do not include delta environment effects. NFE is aligned in the
factory at room temperature. For best performance, in an environment that is different from room temperature,
such as an equipment rack with other instruments, we recommend running the "Characterize Noise Floor"
operation after the first time the analyzer has been installed in the environment, and given an hour to stabilize.
c. DANL of the preamp is specified with a 50Ω source impedance. Like all amplifiers, the noise varies with the
source impedance. When NFE compensates for the noise with an ideal source impedance, the variation in the
remaining noise level with the actual source impedance is greatly multiplied in a decibel sense.
d. NFE performance can give results below theoretical levels of noise in a termination resistor at room temperature,
about –174 dBm/Hz. this is intentional and usually desirable. NFE is not designed to report the noise at the input
of the analyzer; it reports how much more noise is at the input of the analyzer than was present in its alignment.
And its alignment includes the noise of a termination at room temperature. So it can often see the added noise
below the theoretical noise. Furthermore, DANL is defined with log averaging in a 1 Hz RBW, which is about 2.3
dB lower than the noise density (power averaged) in a 1 Hz noise bandwidth.
e. NFE does not apply to the low frequency sensitivity. At frequencies below about 1 MHz, the sensitivity is domi-
nated by phase noise surrounding the LO feedthrough. The NFE is not designed to improve that performance. At
frequencies between 1 and 20 MHz the NFE effectiveness increases from nearly none to near its maximum.
51
Page 52
UXA Signal Analyzer
Dynamic Range
Spurious Responses
DescriptionSpecificationsSupplemental Information
Spurious Responses
Preamp Off
a
(see Band Overlaps on page 17)
Residual Responses
b
200 kHz to 8.4 GHz (swept)
Zero span or FFT or other frequencies
13.5 to 17.1 GHzf+645 MHz−10 dBm−74 dBc−80 dBc−84 dBc−106 dBc
17.0 to 22 GHzf+645 MHz−10 dBm−70 dBc−80 dBc−78 dBc−101 dBc
22 to 26.5 GHzf+645 MHz−10 dBm−66 dBc−70 dBc−75 dBc−102 dBc
26.5 to 34.5 GHzf+645 MHz−30 dBm−70 dBc−98 dBc
34.4 to 42 GHzf+645 MHz−30 dBm−60 dBc−84 dBc
42 to 50 GHzf+645 MHz−30 dBm−75 dBc
(nominal)
Other Spurious Responses
Mixer Level
c
Response
Carrier Frequency ≤26.5 GHz
First RF Order
(f ≥ 10 MHz from carrier)
e
−10 dBm−80 dBc + 20 ×
f
log(N
)
Includes IF feedthrough,
LO harmonic mixing
responses
Higher RF Order
g
(f ≥ 10 MHz from carrier)
−40 dBm
−80 dBc + 20 ×
log(Nf)
Includes higher order mixer
responses
52
Page 53
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Carrier Frequency >26.5 GHz
First RF Order
e
−30 dBm−90 dBc (nominal)
(f ≥ 10 MHz from carrier)
Higher RF Order
g
−30 dBm−90 dBc (nominal)
(f ≥ 10 MHz from carrier)
LO-Related Spurious Responses
(Offset from carrier 200 Hz to 10 MHz)
Line-Related Spurious Responses
−10 dBm
−68 dBc
f
log(N
)
+ 20 ×
−72 dBc + 20 × log(N
(typical)
−73 dBc
h
+ 20 x log(Nf)
hd
(nominal)
a. The spurious response specifications only apply with the preamp turned off. When the preamp is turned on, per-
formance is nominally the same as long as the mixer level is interpreted to be: Mixer Level = Input Level − Input
Attenuation + Preamp Gain.
b. Input terminated, 0 dB input attenuation.
c. Mixer Level = Input Level − Input Attenuation.
d. The following additional spurious responses specifications are supported from 8 to 12 GHz at 20 to 30º C. Image
responses are warranted to be better than –81 dBc, with 95th percentile performance of –87 dBc. LO-related
spurious responses are warranted to be better than –83 dBc at 1 to 10 MHz offsets from the carrier, with phase
noise optimization set to Best Wide-Offset.
e. With first RF order spurious products, the indicated frequency will change at the same rate as the input, with
higher order, the indicated frequency will change at a rate faster than the input.
f. N is the LO multiplication factor.
g. RBW=100 Hz. With higher RF order spurious responses, the observed frequency will change at a rate faster than
the input frequency.
h. Nominally −40 dBc under large magnetic (0.38 Gauss rms) or vibrational (0.21 g rms) environmental stimuli.
f
)
53
Page 54
UXA Signal Analyzer
Dynamic Range
Second Harmonic Distortion
DescriptionSpecificationsSupplemental Information
Second Harmonic Distortion (mmW)
bc
Option 544 or 550Mixer
Level
Source FrequencyLNP off LNP on
Distortion
a
SHI
Distortion
(nominal)
SHI
(nominal)
10 MHz to 1.8 GHz
d
x−15 dBm−60 dBc+45 dBm
1.75 to 2.5 GHzx−15 dBm−95 dBc+80 dBm
1.75 to 3 GHzx−15 dBm−72 dBc+57 dBm
3 to 6.5 GHzx−15 dBm−77 dBc+62 dBm
2.5 to 5 GHzx−15 dBm−99 dBc+84 dBm
6.5 to 10 GHzx−15 dBm−70 dBc+55 dBm
10 to 13.25 GHzx−15 dBm−62 dBc+47 dBm
5 to 13.5 GHzx–15 dBm−105 dBc+90 dBm
13.25 to 25 GHzx−15 dBm−65 dBc+50 dBm
13.25 to 25 GHzx−15 dBm−105 dBc+90 dBm
a. Mixer level = Input Level − Input Attenuation
b. SHI = second harmonic intercept. The SHI is given by the mixer power in dBm minus the second harmonic
distortion level relative to the mixer tone in dBc.
c. Performance >3.6 GHz improves greatly with standard Option LNP enabled.
d. These frequencies are half of the band edge frequencies. See Band Overlaps on page 17.
54
Page 55
UXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Second Harmonic Distortion
(RF/µW)
bc
Option 508, 513, or 526Mixer
Level
Distortion
a
SHI
Source FrequencyLNP
off
10 MHz to 1.8 GHz
d
1.75
to 3 GHz
d
LNP
on
x–15 dBm–60 dBc+45 dBm
x–15 dBm–77 dBc+62 dBm
1.75 to 2.5 GHzx–15 dBm–95 dBc+80 dBm
3 to 6.5 GHzx−15 dBm–77 dBc+62 dBm
2.5 to 4 GHzx–15 dBm–101 dBc+86 dBm
6.5 to 10 GHzx−15 dBm–70 dBc+55 dBm
10 to 13.25 GHzx−15 dBm–62 dBc+47 dBm
4 to 13.25 GHzx–15 dBm–105 dBc+90 dBm
a. Mixer level = Input Level − Input Attenuation
b. SHI = second harmonic intercept. The SHI is given by the mixer power in dBm minus the second harmonic dis-
tortion level relative to the mixer tone in dBc.
c. Performance >3.6 GHz improves greatly with standard Option LNP enabled.
d. These frequencies are half of the band edge frequencies. See Band Overlaps on page 17.
55
Page 56
UXA Signal Analyzer
Dynamic Range
Third Order Intermodulation
DescriptionSpecificationsSupplemental Information
Third Order
Intermodulation
(Tone separation > 5 times IF
Prefilter Bandwidth
Sweep rate reduced
a
b
Refer to the footnote for
Band Overlaps on page 17.
Refer to footnote
e
for the "Extrapolated
Distortion".
Verification conditionsc,
d
LNP off
20 to 30°C
)
Intercept
f
Intercept (typical)
10 to 300 MHz+13.5 dBm+16 dBm
300 to 600 MHz+18 dBm+21 dBm
600 MHz to 1.5 GHz+20 dBm+22 dBm
1.5 to 3.6 GHz+21 dBm+23 dBm
RF/μWmmW RF/μWmmW
3.5 to 8.4 GHz+19 dBm+16 dBm+23 dBm+23 dBm
3.6 to 3.7 GHz
See note
g
8.3 to 13.6 GHz +19 dBm+16 dBm+23 dBm+23 dBm
13.5 to 17.1 GHz +18 dBm+13 dBm+23 dBm+17 dBm
17.0 to 26.5 GHz +19 dBm+13 dBm+24 dBm+20 dBm
26.4 to 34.5 GHz+13 dBm+18 dBm
34.4 to 50 GHz+7 dBm+12 dBm
Full temperature range
10 to 300 MHz+12.5 dBm
300 to 600 MHz+17 dBm
600 MHz to 1.5 GHz+18 dBm
1.5 to 3.6 GHz+19 dBm
3.5 to 13.6 GHz+17 dBm+13 dBm
13.5 to 26.5 GHz+17 dBm+10 dBm
26.4 to 34.5 GHz+11 dBm
34.4 to 50 GHz+3 dBm
56
Page 57
UXA Signal Analyzer
Dynamic Range
a. See the IF Prefilter Bandwidth table in the Gain Compression specifications on page 42. When the tone
separation condition is met, the effect on TOI of the setting of IF Gain is negligible. TOI is verified with IF
Gain set to its best case condition, which is IF Gain = Low.
b. Autocoupled sweep rates using Option FS1 or FS2 are often too fast for excellent TOI performance. A
sweep rate of 1.0 × RBW
IF Prefilter setting. Footnote
2
is often suitable for best TOI performance, because of how it affects the
a
links to the details.
c. TOI is verified with two tones, each at −16 dBm at the mixer, spaced by 100 kHz.
d. When LNP is on, the low noise path is enabled, which causes third-order intercept (TOI) to decrease to the
same extent as that to which the DANL decreases. Therefore, LNP on does not substantially change the TOI
to-noise dynamic range.
e. Traditionally, the distortion components from two tones, each at −30 dBm, were given as specifications.
When spectrum analyzers were not as good as they are now, these distortion products were easily measured. As spectrum analyzers improved, the measurement began to be made at higher levels and extrapolated to the industry-standard −30 dBm test level. This extrapolation was justified by excellent conformance
with the third-order model, wherein distortion in dBc was given by twice the difference between the test
tone level and the intercept, both given in dBm units. In UXA, we no longer make that extrapolation in this
Specifications Guide.
One reason we don’t extrapolate is that the model does not work as well as it had with higher levels of distortion in older and less capable analyzers, so that the computation is misleading; distortions at low test levels will be modestly higher than predicted from the formula. The second reason is that the distortion
components are so small as to be unmeasurable, and thus highly irrelevant, in many cases.
Please note the slope of the third-order intermodulation lines in the graphs that follow. The slope differs
somewhat from that of the ideal third-order model, which would have a slope of 2.
f. Intercept = TOI = third order intercept. The TOI is given by the mixer tone level (in dBm) minus (distortion/2)
where distortion is the relative level of the distortion tones in dBc.
g. Band 0 is extendable (set “Extend Low Band” to On) to 3.7 GHz instead of 3.6 GHz in instruments with fre-
quency option 508, 513 or 526 and with firmware of version A.16.17 or later. Subject to these conditions,
statistical observations show that performance nominally fits within the same range within the 3.6 to 3.7
GHz frequencies as within the next lower specified frequency range, but is not warranted.
57
Page 58
UXA Signal Analyzer
Dynamic Range
Nominal Dynamic Range vs. Offset Frequency vs. RBW [Plot]
a. Noise sidebands around a signal are dominantly phase noise sidebands. With the extremely low phase noise of
the UXA, AM sidebands are non-negligible contributors. These specifications apply to the sum of the AM and
PM sidebands.
b. The nominal performance of the phase noise at center frequencies different than the one at which the
specifications apply (1 GHz) depends on the center frequency, band and the offset. For low offset frequencies,
offsets well under 100 Hz, the phase noise changes by 20 × log[(f+0.3225)/1.3225]. For mid-offset frequencies
such as 50 kHz, phase noise changes as 20 × log[(f+5.1225)/6.1225]. In both expressions, f is the larger of 0.5
and the carrier frequency in GHz units. For wide offset frequencies, offsets above about 500 kHz, phase noise
increases as 20 × log(N). N is the LO Multiple as shown on page 9.
c. Noise sidebands for lower offset frequencies, for example, 10 kHz, apply with phase noise optimization (PNO)
set to Balance Noise and Spurs. In some frequency settings of the analyzer, a spurious response 60 to 180 MHz
offset from the carrier may be present unless the phase locked loop behavior is changed in a way that increases
the phase noise. This tradeoff is controlled such that the spurs are better than –70 dBc, at the expense of up to
7 dB increase in phase noise within ±1 octave of 1 MHz offset for those settings where this spurious is likely to
be visible. To eliminate this phase noise degradation in exchange for the aforementioned spurs, Best Close-in
Noise should be used. When the setting is changed to Best Spurs, the maximum spurious response is held to
–90 dBc, but the phase noise at all center frequencies is degraded by up to approximately 12 dB from the best
possible setting, mostly within ±1 octave of an offset of 400 kHz from the carrier. Noise sidebands for higher
offset frequencies, for example, 1 MHz, apply with the phase noise optimization set to Best Wide-Offset Noise.
59
Page 60
UXA Signal Analyzer
Dynamic Range
d. Specifications are given with the internal frequency reference. The phase noise at offsets below 100 Hz is
impacted or dominated by noise from the reference. Thus, performance with external references will not follow
the curves and specifications. When using an external reference with superior phase noise, we recommend set-
ting the external reference phase-locked-loop bandwidth to wide (60 Hz), to take advantage of that superior
performance. When using an external reference with inferior phase noise performance, we recommend setting
that bandwidth to narrow (15 Hz). In these relationships, inferior and superior phase noise are with respect to
–134 dBc/Hz at 30 Hz offset from a 10 MHz reference. Because most reference sources have phase noise
behavior that falls off at a rate of 30 dB/decade, this is usually equivalent to –120 dBc/Hz at 10 Hz offset.
e. Keysight measures 100% of the signal analyzers for phase noise at 10 Hz offset from a 1 GHz carrier in the fac-
tory production process. This measurement requires a signal of exceptionally low phase noise that is character-
ized with specialized processes. It is impractical for field and customer use. Because field verification is
impractical, Keysight only gives a typical result. More than 80% of prototype instruments met this "typical"
specification; the factory test line limit is set commensurate with an on-going 80% yield to this typical. Like all
typical specifications, there is no guardbanding for measurement uncertainty. The factory test line limit is con-
sistent with a warranted specification of –89 dBc/Hz.
f. Analyzer-contributed phase noise at the low levels of this offset requires advanced verification techniques
because broadband noise would otherwise cause excessive measurement error. Keysight uses a high level low
phase noise CW test signal and sets the input attenuator so that the mixer level will be well above the normal
top-of-screen level (-10 dBm) but still well below the 1 dB compression level. This improves dynamic range
(carrier to broadband noise ratio) at the expense of amplitude uncertainty due to compression of the phase
noise sidebands of the analyzer. (If the mixer level were increased to the "1 dB Gain Compression Point," the
compression of a single sideband is specified to be 1 dB or lower. At lower levels, the compression falls off rap-
idly. The compression of phase noise sidebands is substantially less than the compression of a single-sideband
test signal, further reducing the uncertainty of this technique.) Keysight also measures the broadband noise of
the analyzer without the CW signal and subtracts its power from the measured phase noise power. The same
techniques of overdrive and noise subtraction can be used in measuring a DUT, of course.
60
Page 61
UXA Signal Analyzer
Dynamic Range
Nominal Phase Noise at Different Carrier Frequencies, Phase Noise Optimized vs Offset Frequency [Plot]
61
Page 62
UXA Signal Analyzer
Dynamic Range
Nominal Phase Noise at Different Phase Noise/Spurs Optimization [Plot]
62
Page 63
UXA Signal Analyzer
Power Suite Measurements
Power Suite Measurements
The specifications for this section apply only to instruments with Frequency Option 508, 513, or
526.
DescriptionSpecificationsSupplemental Information
Channel Power
Amplitude Accuracy
Case: Radio Std = 3GPP W-CDMA, or IS-95
Absolute Power Accuracy
(20 to 30°C, Attenuation = 10 dB)
a. See “Absolute Amplitude Accuracy” on page 34.
b. See “Frequency and Time” on page 17.
c. Expressed in dB.
DescriptionSpecificationsSupplemental Information
Occupied Bandwidth
Frequency Accuracy±(Span/1000) (nominal)
±0.61 dB
Absolute Amplitude Accuracy
Power Bandwidth Accuracy
±0.19 dB (95th percentile)
a
+
bc
63
Page 64
UXA Signal Analyzer
Power Suite Measurements
DescriptionSpecificationsSupplemental Information
Adjacent Channel Power (ACP)
Case: Radio Std = None
Accuracy of ACP Ratio (dBc)
Accuracy of ACP Absolute Power
(dBm or dBm/Hz)
Accuracy of Carrier Power (dBm), or
Carrier Power PSD (dBm/Hz)
Passband Width
e
Case: Radio Std = 3GPP W-CDMA
Display Scale Fidelity
Absolute Amplitude Accuracy
Power Bandwidth Accuracy
Absolute Amplitude Accuracy
Power Bandwidth Accuracy
MS (UE)5 MHz±0.08 dB At ACPR range of −30 to −36 dBc with optimum
mixer level
h
MS (UE)10 MHz±0.09 dBAt ACPR range of −40 to −46 dBc with optimum
mixer level
i
BTS5 MHz±0.22 dBAt ACPR range of −42 to −48 dBc with optimum
mixer level
j
BTS10 MHz±0.18 dBAt ACPR range of −47 to −53 dBc with optimum
i
k
BTS5 MHz±0.10 dB
mixer level
At −48 dBc non-coherent ACPR
Dynamic RangeRRC weighted, 3.84 MHz noise
bandwidth
m
Noise
Correction
l
Offset
Freq
Method
ACLR (typical)
Optimum MLn
(Nominal)
Off5 MHzFiltered IBW−81 dB−8 dBm
Off5 MHzFast−81 dB−8 dBm
Off10 MHzFiltered IBW−87 dB−4 dBm
On5 MHzFiltered IBW−82.5 dB−8 dBm
On10 MHzFiltered IBW−89 dB−4 dBm
64
Page 65
UXA Signal Analyzer
Power Suite Measurements
DescriptionSpecificationsSupplemental Information
RRC Weighting Accuracy
White noise in Adjacent Channel
TOI-induced spectrum
rms CW error
o
0.00 dB nominal
0.001 dB nominal
0.012 dB nominal
a. The effect of scale fidelity on the ratio of two powers is called the relative scale fidelity. The scale fidelity speci-
fied in the Amplitude section is an absolute scale fidelity with –35 dBm at the input mixer as the reference point.
The relative scale fidelity is nominally only 0.01 dB larger than the absolute scale fidelity.
b. See Amplitude Accuracy and Range section.
c. See Frequency and Time section.
d. Expressed in decibels.
e. An ACP measurement measures the power in adjacent channels. The shape of the response versus frequency of
those adjacent channels is occasionally critical. One parameter of the shape is its 3 dB bandwidth. When the
bandwidth (called the Ref BW) of the adjacent channel is set, it is the 3 dB bandwidth that is set. The passband
response is given by the convolution of two functions: a rectangle of width equal to Ref BW and the power
response versus frequency of the RBW filter used. Measurements and specifications of analog radio ACPs are
often based on defined bandwidths of measuring receivers, and these are defined by their −6 dB widths, not
their −3 dB widths. To achieve a passband whose −6 dB width is x, set the Ref BW to be x − 0.572 × RBW.
f. Most versions of adjacent channel power measurements use negative numbers, in units of dBc, to refer to the
power in an adjacent channel relative to the power in a main channel, in accordance with ITU standards. The
standards for W-CDMA analysis include ACLR, a positive number represented in dB units. In order to be consis-
tent with other kinds of ACP measurements, this measurement and its specifications will use negative dBc
results, and refer to them as ACPR, instead of positive dB results referred to as ACLR. The ACLR can be deter-
mined from the ACPR reported by merely reversing the sign.
g. The accuracy of the Adjacent Channel Power Ratio will depend on the mixer drive level and whether the distor-
tion products from the analyzer are coherent with those in the UUT. These specifications apply even in the worst
case condition of coherent analyzer and UUT distortion products. For ACPR levels other than those in this spec-
ifications table, the optimum mixer drive level for accuracy is approximately −37 dBm − (ACPR/3), where the
ACPR is given in (negative) decibels.
h. To meet this specified accuracy when measuring mobile station (MS) or user equipment (UE) within 3 dB of the
required −33 dBc ACPR, the mixer level (ML) must be optimized for accuracy. This optimum mixer level is
−22 dBm, so the input attenuation must be set as close as possible to the average input power − (−22 dBm).
For example, if the average input power is −6 dBm, set the attenuation to 16 dB. This specification applies for
the normal 3.5 dB peak-to-average ratio of a single code. Note that, if the mixer level is set to optimize dynamic
range instead of accuracy, accuracy errors are nominally doubled.
i. ACPR accuracy at 10 MHz offset is warranted when the input attenuator is set to give an average mixer level of
−14 dBm.
j. In order to meet this specified accuracy, the mixer level must be optimized for accuracy when measuring node B
Base Transmission Station (BTS) within 3 dB of the required −45 dBc ACPR. This optimum mixer level is −18
dBm, so the input attenuation must be set as close as possible to the average input power − (−18 dBm). For
example, if the average input power is −6 dBm, set the attenuation to 12 dB. This specification applies for the
normal 10 dB peak-to-average ratio (at 0.01% probability) for Test Model 1. Note that, if the mixer level is set to
optimize dynamic range instead of accuracy, accuracy errors are nominally doubled.
k. Accuracy can be excellent even at low ACPR levels assuming that the user sets the mixer level to optimize the
dynamic range, and assuming that the analyzer and UUT distortions are incoherent. When the errors from the
UUT and the analyzer are incoherent, optimizing dynamic range is equivalent to minimizing the contribution of
analyzer noise and distortion to accuracy, though the higher mixer level increases the display scale fidelity
errors. This incoherent addition case is commonly used in the industry and can be useful for comparison of
analysis equipment, but this incoherent addition model is rarely justified. This derived accuracy specification is
based on a mixer level of −14 dBm.
65
Page 66
UXA Signal Analyzer
Power Suite Measurements
l. The dynamic range shown with Noise Correction = Off applies with Noise Floor Extension On. (Noise Correction
is the process within the measurement of making a calibration of the noise floor at the exact analyzer settings
used for the measurement. Noise Floor Extension is the factory calibration of the noise floor.)
m. Keysight measures 100% of the signal analyzers for dynamic range in the factory production process. This mea-
surement requires a near-ideal signal, which is impractical for field and customer use. Because field verification
is impractical, Keysight only gives a typical result. More than 80% of prototype instruments met this “typical”
specification; the factory test line limit is set commensurate with an on-going 80% yield to this typical.
The ACPR dynamic range is verified only at 2 GHz, where Keysight has the near-perfect signal available. The
dynamic range is specified for the optimum mixer drive level, which is different in different instruments and dif-
ferent conditions. The test signal is a 1 DPCH signal.
The ACPR dynamic range is the observed range. This typical specification includes no measurement uncertainty.
n. ML is Mixer Level, which is defined to be the input signal level minus attenuation.
o. 3GPP requires the use of a root-raised-cosine filter in evaluating the ACLR of a device. The accuracy of the
passband shape of the filter is not specified in standards, nor is any method of evaluating that accuracy. This
footnote discusses the performance of the filter in this instrument. The effect of the RRC filter and the effect of
the RBW used in the measurement interact. The analyzer compensates the shape of the RRC filter to accommo-
date the RBW filter. The effectiveness of this compensation is summarized in three ways:
− White noise in Adj Ch: The compensated RRC filter nominally has no errors if the adjacent channel has a
spectrum that is flat across its width.
− TOI−induced spectrum: If the spectrum is due to third−order intermodulation, it has a distinctive shape. The
computed errors of the compensated filter are −0.001 dB for the 100 kHz RBW used for UE testing with the
IBW method. It is 0.000 dB for the 27 kHz RBW filter used for BTS testing with the Filtered IBW method. The
worst error for RBWs between 27 and 390 kHz is 0.05 dB for a 330 kHz RBW filter.
− rms CW error: This error is a measure of the error in measuring a CW−like spurious component. It is evaluated
by computing the root of the mean of the square of the power error across all frequencies within the adjacent
channel. The computed rms error of the compensated filter is 0.012 dB for the 100 kHz RBW used for UE test-
ing with the IBW method. It is 0.000 dB for the 27 kHz RBW filter used for BTS testing. The worst error for
RBWs between 27 kHz and 470 kHz is 0.057 dB for a 430 kHz RBW filter.
a. Coher = no means that the specified accuracy only applies when the distortions of the device under test are not
coherent with the third-order distortions of the analyzer. Incoherence is often the case with advanced
multi-carrier amplifiers built with compensations and predistortions that mostly eliminate coherent third-order
effects in the amplifier.
b. Optimum mixer level (MLOpt). The mixer level is given by the average power of the sum of the four carriers
minus the input attenuation.
66
Page 67
UXA Signal Analyzer
Power Suite Measurements
DescriptionSpecificationsSupplemental Information
Power Statistics CCDF
Histogram Resolution
a
0.01 dB
a. The Complementary Cumulative Distribution Function (CCDF) is a reformatting of a histogram of the power
envelope. The width of the amplitude bins used by the histogram is the histogram resolution. The resolution of
the CCDF will be the same as the width of those bins.
DescriptionSpecificationsSupplemental Information
Burst Power
MethodsPower above threshold
Power within burst width
ResultsOutput power, average
Output power, single burst
Maximum power
Minimum power within burst
Burst width
DescriptionSpecificationsSupplemental Information
TOI (Third Order
Intermodulation)
Measures TOI of a signal with two
dominant tones
ResultsRelative IM tone powers (dBc)
Absolute tone powers (dBm)
Intercept (dBm)
DescriptionSpecificationsSupplemental Information
Harmonic Distortion
Maximum harmonic number10th
ResultsFundamental Power (dBm)
Relative harmonics power (dBc)
Total harmonic distortion (%, dBc)
67
Page 68
UXA Signal Analyzer
Power Suite Measurements
Description SpecificationsSupplemental Information
Spurious EmissionsTable-driven spurious signals;
search across regions
Case: Radio Std = 3GPP W-CDMA
Dynamic Range
a
, relative (RBW=1 MHz)
88.4 dB 90.7 dB (typical)
(1 to 3.0 GHz)
Sensitivity
b
, absolute (RBW=1 MHz)
−88.5 dBm−90.5 dBm (typical)
(1 to 3.0 GHz)
AccuracyAttenuation = 10 dB
20 Hz to 3.6 GHz±0.19 dB (95th percentile)
3.5 to 8.4 GHz±1.13 dB (95th percentile)
8.3 to 13.6 GHz±1.50 dB (95th percentile)
a. The dynamic range is specified at 12.5 MHz offset from center frequency with mixer level of 1 dB compression
point, which will degrade accuracy 1 dB.
b. The sensitivity is specified at far offset from carrier, where phase noise does not contribute. You can derive the
dynamic range at far offset from 1 dB compression mixer level and sensitivity.
a. The dynamic range specification is the ratio of the channel power to the power in the offset specified. The
dynamic range depends on the measurement settings, such as peak power or integrated power. Dynamic range
specifications are based on default measurement settings, with detector set to average, and depend on the
mixer level. Default measurement settings include 30 kHz RBW.
b. This dynamic range specification applies for the optimum mixer level, which is about −18 dBm. Mixer level is
defined to be the average input power minus the input attenuation.
c. The sensitivity is specified with 0 dB input attenuation. It represents the noise limitations of the analyzer. It is
tested without an input signal. The sensitivity at this offset is specified in the default 30 kHz RBW, at a center
frequency of 2 GHz.
d. The relative accuracy is a measure of the ratio of the power at the offset to the main channel power. It applies for
spectrum emission levels in the offsets that are well above the dynamic range limitation.
e. The absolute accuracy of SEM measurement is the same as the absolute accuracy of the spectrum analyzer. See
“Absolute Amplitude Accuracy” on page 34 for more information. The numbers shown are for 0 to 3.6
GHz, with attenuation set to 10 dB.
69
Page 70
UXA Signal Analyzer
Options
Options
The following options and applications affect instrument specifications.
Standard Option CR3:Connector Rear, second IF Out
Standard Option EXM:External mixing
Standard Option LNP:Low Noise Path
Standard Option MPB:Preselector bypass
Standard Option NFE:Noise floor extension, instrument alignment
Option 508:Frequency range, 2 Hz to 8.4 GHz
Option 513:Frequency range, 2 Hz to 13.6 GHz
Option 526:Frequency range, 2 Hz to 26.5 GHz
Option 544:Frequency range, 2 Hz to 44 GHz
Option 550:Frequency range, 2 Hz to 50 GHz
Option ALV:Auxiliary Log Video output
Option B25:Analysis bandwidth, 25 MHz
Option B40:Analysis bandwidth, 40 MHz
Option B2X:Analysis bandwidth, 255 MHz
Option B5X:Analysis bandwidth, 510 MHz
Option C35:APC 3.5 mm connector (for Freq Option 526 only)
Option CRP:Connector Rear, arbitrary IF Out
Option EA3:Electronic attenuator, 3.6 GHz
Option EMC:Precompliance EMC Features
Option P08:Preamplifier, 8.4 GHz
Option P13:Preamplifier, 13.6 GHz
Option P26:Preamplifier, 26.5 GHz
Option P44:Preamplifier, 44 GHz
Option P50:Preamplifier, 50 GHz
Option RT1:Real-time analysis up to the maximum analysis bandwidth, basic detection
Option RT2:Real-time analysis up to the maximum analysis bandwidth, optimum detection
N9083EM0E:Multi-Standard Radio measurement application
N9084EM0E:Short Range Communications measurement application
N9085EM0E:5G NR measurement application
71
Page 72
UXA Signal Analyzer
General
General
DescriptionSpecificationsSupplemental Information
Calibration Cycle1 year
DescriptionSpecificationsSupplemental Information
Environmental
Indoor use
Temperature Range
Operating
Altitude ≤ 2,300 m0 to 55°C
Altitude = 4,600 m0 to 47°C
Derating
Storage−40 to +70°C
Altitude4,600 m (approx 15,000 feet)
Humidity
Relative humidityType tested at 95%, +40°C
DescriptionSpecificationsSupplemental Information
Environmental and Military
Specifications
a
(non-condensing)
a. The maximum operating temperature derates linearly from altitude of 4,600 m to 2,300 m.
Samples of this product have been type tested in
accordance with the Keysight Environmental Test
Manual and verified to be robust against the
environmental stresses of Storage, Transportation
and End-use; those stresses include but are not
limited to temperature, humidity, shock, vibration,
altitude and power line conditions. Test Methods are
aligned with IEC 60068-2 and levels are similar to
MIL-PRF-28800F Class 3.
72
Page 73
UXA Signal Analyzer
General
DescriptionSpecificationSupplemental Information
Acoustic NoiseValues given are per ISO 7779 standard in the "Operator
Sitting" position
Ambient Temperature
< 35°CNominally under 55 dBA Sound Pressure. 55 dBA is generally
considered suitable for use in quiet office environments.
≥ 35°CNominally under 65 dBA Sound Pressure. 65 dBA is generally
considered suitable for use in noisy office environments. (The
fan speed, and thus the noise level, increases with increasing
ambient temperature.)
73
Page 74
UXA Signal Analyzer
General
DescriptionSpecificationSupplemental Information
Power Requirements
Low Range
Voltage100 /120 V± 10% operating range
Frequency50/60/400 Hz
High Range
Voltage220/240 V± 10% operating range
Frequency 50/60 Hz
Power Consumption, On850 W (Maximum)470 W (typical)
Power Consumption, Standby25 WStandby power is supplied to both the
CPU and the frequency reference
oscillator.
The UXA has autoranging line voltage input. Before switching on the instrument,
be sure the supply voltage is within the specified range and voltage fluctuations
do not exceed 10 percent of the nominal supply voltage.
DescriptionSupplemental Information
Measurement Speed
Local measurement and display update rate
Remote measurement and LAN transfer rate
a
bc
bc
Nominal
10 ms
10.7 ms
Marker Peak Search4.4 ms
Center Frequency Tune and Transfer (Band 0)20 ms
Center Frequency Tune and Transfer (Bands 1-4)48 ms
Measurement/Mode Switching100 ms
a. Sweep Points = 101.
b. Factory preset, fixed center frequency, RBW = 1 MHz, 10 MHz < span ≤ 600 MHz, stop frequency ≤ 3.6 GHz,
Auto Align Off.
c. Phase Noise Optimization set to Fast Tuning, Display Off, 32 bit REAL, markers Off, single sweep, measured with
HP Z420(memory 120 Gb, Windows 7. Intel Xcon CPU E5-1620 3.6 GHz), Keysight I/O Libraries Suite Version
16.317914, one meter GPIB cable, Keysight GPIB Card.
74
Page 75
UXA Signal Analyzer
General
DescriptionSpecificationsSupplemental Information
Display
Resolution1280 × 800Capacitive multi-touch screen
Size357 mm (14.1 in) diagonal (nominal)
DescriptionSpecificationsSupplemental Information
Data Storage
Removable solid state drive (SSD)≥80 GB total volume; ≥9 GB for user data,
available on separate partition.
Secured digital (SD) memory deviceFor calibration data backup.
DescriptionSpecificationsSupplemental Information
WeightWeight without options
Net30.9 kg (68 lbs) (nominal)
Shipping 39.5 kg (87 lbs) (nominal)
Cabinet DimensionsCabinet dimensions exclude front and rear
This product is designed for use in Installation Category II and Pollution Degree 2 per IEC 61010 3rd
ed, and 664 respectively.
This product has been designed and tested in accordance with accepted industry standards, and
has been supplied in a safe condition. The instruction documentation contains information and
warnings which must be followed by the user to ensure safe operation and to maintain the product
in a safe condition.
This product is intended for indoor use.
The CE mark is a registered trademark of the European Community (if accompanied
by a year, it is the year when the design was proven). This product complies with all
relevant directives.
ccr.keysight@keysight.com
ICES/NMB-001“This ISM device complies with Canadian ICES-001.”
ISM 1-A (GRP.1 CLASS A)This is a symbol of an Industrial Scientific and Medical Group 1 Class A product.
The CSA mark is a registered trademark of the CSA International.
The Keysight email address is required by EU directives applicable to our product.
“Cet appareil ISM est conforme a la norme NMB du Canada.”
(CISPR 11, Clause 4)
The RCM mark is a registered trademark of the Australian Communications and
Media Authority.
This symbol indicates separate collection for electrical and electronic equipment
mandated under EU law as of August 13, 2005. All electric and electronic equipment
are required to be separated from normal waste for disposal (Reference WEEE
Directive 2002/96/EC).
China RoHS regulations include requirements related to packaging, and require
compliance to China standard GB18455-2001.
This symbol indicates compliance with the China RoHS regulations for
paper/fiberboard packaging.
South Korean Certification (KC) mark; includes the marking’s identifier code which
follows this format:
MSIP-REM-YYY-ZZZZZZZZZZZZZZ.
80
Page 81
UXA Signal Analyzer
Regulatory Information
EMC: Complies with the essential requirements of the European EMC Directive as well as current
editions of the following standards (dates and editions are cited in the Declaration of Conformity):
— IEC/EN 61326-1
— CISPR 11, Group 1, Class A
— AS/NZS CISPR 11
— ICES/NMB-001
This ISM device complies with Canadian ICES-001.
Cet appareil ISM est conforme a la norme NMB-001 du Canada.
This is a sensitive measurement apparatus by design and may have some performance
loss (up to 25 dBm above the Spurious Responses, Residual specification of -100 dBm)
when exposed to ambient continuous electromagnetic phenomenon in the range of
80 MHz -2.7 GHz when tested per IEC 61000-4-3.
South Korean Class A EMC declaration:
This equipment has been conformity assessed for use in business environments. In a residential
environment this equipment may cause radio interference.
This EMC statement applies to the equipment only for use in business environment.
SAFETY: Complies with the essential requirements of the European Low Voltage Directive as well as
current editions of the following standards (dates and editions are cited in the Declaration of
To fin d a cu rrent Declaration of Conformity for a specific Keysight product, go to:
http://www.keysight.com/go/conformity
82
Page 83
Keysight X-Series Signal Analyzer
N9040B
Specification Guide
2I/Q Analyzer, Standard
This chapter contains specifications for the I/Q Analyzer measurement application (Basic Mode).
83
Page 84
I/Q Analyzer, Standard
Specifications Affected by I/Q Analyzer
Specifications Affected by I/Q Analyzer
Specification NameInformation
Number of Frequency Display Trace Points
(buckets)
Resolution BandwidthSee “Frequency” on page 85 in this chapter.
Video BandwidthNot available.
Clipping-to-Noise Dynamic RangeSee “Clipping-to-Noise Dynamic Range” on page 86 in this
Resolution Bandwidth Switching UncertaintyNot specified because it is negligible.
Available DetectorsDoes not apply.
Spurious ResponsesThe “Spurious Responses” on page 52 of core specifications still
IF Amplitude FlatnessSee “IF Frequency Response” on page 32 of the core
IF Phase LinearitySee “IF Phase Linearity” on page 33 of the core specifications for
Does not apply.
chapter.
apply. Additional bandwidth-option-dependent spurious responses are
given in the Analysis Bandwidth chapter for any optional bandwidths in
use.
specifications for the 10 MHz bandwidth. Specifications for wider
bandwidths are given in the Analysis Bandwidth chapter for any
optional bandwidths in use.
the 10 MHz bandwidth. Specifications for wider bandwidths are given
in the Analysis Bandwidth chapter for any optional bandwidths in use.
Data AcquisitionSee “Data Acquisition” on page 87 in this chapter for the 10 MHz
bandwidth. Specifications for wider bandwidths are given in the
Analysis Bandwidth chapter for any optional bandwidths in use.
84
Page 85
I/Q Analyzer, Standard
Frequency
Frequency
DescriptionSpecificationsSupplemental Information
Frequency Span
Option B25 (Standard)10 Hz to 25 MHz
Option B4010 Hz to 40 MHz
Option B2X10 Hz to 255 MHz
Option B5X10 Hz to 510 MHz
Option H1G
Resolution Bandwidth
(Spectrum Measurement)
Range
Overall
Span = 1 MHz
Span = 10 kHz
Span = 100 Hz
Window ShapesFlat Top, Uniform, Hanning, Hamming,
Analysis Bandwidth (Span)
(Waveform Measurement)
Option B25 (Standard)10 Hz to 25 MHz
Option B4010 Hz to 40 MHz
Option B2X10 Hz to 255 MHz
Option B5X10 Hz to 510 MHz
Option H1G
a
a
40 MHz to 1 GHz
100 mHz to 3 MHz
50 Hz to 1 MHz
1 Hz to 10 kHz
100 mHz to 100 Hz
Gaussian, Blackman, Blackman-Harris, Kaiser
Bessel (K-B 70 dB, K-B 90 dB & K-B 110 dB)
40 MHz to 1 GHz
a. In the 1 GHz bandwidth path, the span and bandwidth will be 40 MHz minimum. Below 40 MHz, a
narrower IF path is used.
85
Page 86
I/Q Analyzer, Standard
Clipping-to-Noise Dynamic Range
Clipping-to-Noise Dynamic Range
DescriptionSpecificationsSupplemental Information
Clipping-to-Noise Dynamic Range
a
Excluding residuals and spurious
responses
Clipping Level at MixerCenter frequency ≥ 20 MHz
IF Gain = Low−10 dBm−8 dBm (nominal)
IF Gain = High−20 dBm−17.5 dBm (nominal)
Noise Density at Mixer
at center frequency
c
(DANL
+ IFGainEffectd) + 2.25 dB
b
e
Example
f
a. This specification is defined to be the ratio of the clipping level (also known as “ADC Over Range”) to the noise
density. In decibel units, it can be defined as clipping_level [dBm] − noise_density [dBm/Hz]; the result has
units of dBFS/Hz (fs is “full scale”).
b. The noise density depends on the input frequency. It is lowest for a broad range of input frequencies near the
center frequency, and these specifications apply there. The noise density can increase toward the edges of the
span. The effect is nominally well under 1 dB.
c. The primary determining element in the noise density is the “Displayed Average Noise Level” on
page 44.
d. DANL is specified with the IF Gain set to High, which is the best case for DANL but not for Clipping-to-noise
dynamic range. The core specifications “Displayed Average Noise Level” on page 44, gives a line entry
on the excess noise added by using IF Gain = Low, and a footnote explaining how to combine the IF Gain noise
with the DANL.
e. DANL is specified for log averaging, not power averaging, and thus is 2.51 dB lower than the true noise density.
It is also specified in the narrowest RBW, 1 Hz, which has a noise bandwidth slightly wider than 1 Hz. These
two effects together add up to 2.25 B.
f. As an example computation, consider this: For the case where DANL = −151 dBm in 1 Hz, IF Gain is set to low,
and the “Additional DANL” is −160 dBm, the total noise density computes to −148.2 dBm/Hz and the Clipping-to-noise ratio for a −10 dBm clipping level is −138.2 dBFS/Hz.
86
Page 87
I/Q Analyzer, Standard
Data Acquisition
Data Acquisition
DescriptionSpecificationsSupplemental Information
Time Record Length
IQ Analyzer8,000,000 IQ sample pairs
Advanced ToolsData Packing
Waveform measurement
89600 VSA software or Fast Capture
a
b
32-bit 64-bit
Length (IQ sample pairs)
29
IFBW ≤255.176 MHz
IFBW >255.176 MHz
536 MSa (2
1073 MSa (2
Sa)268 MSa (228 Sa)
29
Sa)2536 MSa (228 Sa)
2 GB total memory
2 GB total memory
Maximum IQ Capture TimeData PackingData Packing
(89600VSA and Fast Capture)32-bit 64-bit 32-bit 64-bit
29
(2
10 MHz IFBW42.94 s21.47 s
25 MHz IFBW17.17 s8.58 s
40 MHz IFBW10.73 s5.36 s
240 MHz IFBW1.78 s0.89 s
255 MHz IFBW1.78 s0.89 s
256 MHz IFBW3.35 s1.67 s
)/10 MHz ×1.25)(228)/10 MHz ×1.25)
29
)/25 MHz ×1.25)(228)/25 MHz ×1.25)
(2
29
)/40 MHz ×1.25)(228)/40 MHz ×1.25)
(2
29
)/240 MHz ×1.25)(228)/240 MHz ×1.25)
(2
29
)/300 MSA/s)(228)/300 MSa/s)
(2
30
)/256 MHz ×1.25)(229)/256 MHz ×1.25)
(2
30
)/480 MHz ×1.25)(229)/480 MHz ×1.25)
480 MHz IFBW1.78 s0.89 s
510 MHz IFBW
1.78 s0.89 s
(2
30
)/600 MSa/s)(229)/600 MSa/s)
(2
Maximum IQ Capture TimeData Packing
(89600 VSA and Fast Capture)32-bit 64-bit Calculated by: Length
of IQ sample
10 MHz IFBW42.94 s21.47 s
pairs/Sample Rate
(IQ Pairs)
c
Sample Rate (IQ Pairs)1.25 × IFBW
ADC Resolution16 bits
a. This can also be accessed with the remote programming command of "read:wav0?".
b. This can only be accessed with the remote programming command of "init:fcap" in the IQ Analyzer (Basic) waveform
measurement.
c. For example, using 32-bit data packing at 10 MHz IF bandwidth (IFBW) the Maximum Capture Time is calculated using
29
the formula: "Max Capture Time = (2
)/(10 MHz × 1.25)".
87
Page 88
I/Q Analyzer, Standard
Data Acquisition
88
Page 89
Keysight X-Series Signal Analyzer
N9040B
Specification Guide
3Standard Option CR3 - Connector Rear, 2nd IF Output
This chapter contains specifications for Option CR3, Connector Rear, 2nd IF Output.
89
Page 90
Standard Option CR3 - Connector Rear, 2nd IF Output
Specifications Affected by Connector Rear, 2nd IF Output
Specifications Affected by Connector Rear, 2nd IF Output
No other analyzer specifications are affected by the presence or use of this option. New
specifications are given in the following pages.
90
Page 91
Standard Option CR3 - Connector Rear, 2nd IF Output
Other Connector Rear, 2nd IF Output Specifications
Other Connector Rear, 2nd IF Output Specifications
Aux IF Out Port
DescriptionSpecificationsSupplemental Information
ConnectorSMA femaleShared with other options
Impedance50Ω (nominal)
Second IF Out
DescriptionSpecificationsSupplemental Information
Second IF Out
Output Center Frequency
SA Mode322.5 MHz
I/Q Analyzer Mode
IF Path ≤ 25 MHz322.5 MHz
IF Path 40 MHz250 MHz
IF Path 255 MHz750 MHz
IF Path 510 MHz877.1484375 MHz
IF Path 1 GHz750 MHz
Conversion Gain at 2nd IF output
center frequency
Bandwidth (−6 dB)
Low band
IF Path ≤ 40 MHz
IF Path 255 MHz255 MHz (nominal)
IF Path 510 MHz510 MHz (nominal)
High band
With preselector
1 dB (nominal)
Up to 140 MHz (nominal)
Depends on RF center frequency
a
b
c
Range
Preselector bypassed
External Mixing 100 - 1200 MHz ±6 dB (nominal)
Residual Output Signals−94 dBm or lower (nominal)
100 - 800 MHz ±3 dB (nominal)
91
Page 92
Standard Option CR3 - Connector Rear, 2nd IF Output
Other Connector Rear, 2nd IF Output Specifications
a. "Conversion Gain" is defined from RF input to IF out with 0 dB mechanical attenuation and the electronic atten-
uator off. The nominal performance applies in zero span.
b. The passband width at –3 dB nominally extends from IF frequencies of 230 to 370 MHz. When the IF in use is
centered at a frequency different from 300 MHz, the passband will be asymmetric.
c. The YIG-tuned preselector bandwidth nominally varies from 55 MHz for a center frequencies of 3.6 GHz
through 57 MHz at 15 GHz to 75 MHz at 26.5 GHz. The preselector effect will dominate the passband width.
See “Preselector Bandwidth” on page 28.
92
Page 93
Keysight X-Series Signal Analyzer
N9040B
Specification Guide
4Standard Option EXM - External Mixing
This chapter contains specifications for the Option EXM External Mixing.
93
Page 94
Standard Option EXM - External Mixing
Specifications Affected by External mixing
Specifications Affected by External mixing
Specification NameInformation
RF-Related Specifications, such as TOI,
DANL, SHI, Amplitude Accuracy, and so
forth.
IF-Related Specifications, such as RBW
range, RBW accuracy, RBW switching
uncertainty, and so forth.
New specifications:
IF Input
Mixer Bias
LO Output
Specifications do not apply; some related specifications are contained in IF
Input in this chapter
Specifications unchanged, except IF Frequency Response — see
specifications in this chapter.
See specifications in this chapter.
94
Page 95
Standard Option EXM - External Mixing
Other External Mixing Specifications
Other External Mixing Specifications
DescriptionSpecificationsSupplemental Information
Connection Port EXT MIXER
ConnectorSMA, female
Impedance50Ω (nominal) at IF and LO frequencies
FunctionsTriplexed for Mixer Bias, IF
Input and LO output
DescriptionSpecificationsSupplemental Information
Mixer Bias
Bias Current
Range±10 mA
Resolution10 μA
Output impedance477Ω (nominal)
Voltage clamp±3.7 V (nominal)
a. The mixer bias circuit has a Norton equivalent, characterized by its short circuit current and its impedance. It is
b. The actual port current is often less than the short circuit current, due to the diode voltage drop of many mixers.
a
also clamped to a voltage range less than the Thevenin voltage capability.
Short circuit current
b
95
Page 96
Standard Option EXM - External Mixing
Other External Mixing Specifications
DescriptionSpecificationsSupplemental Information
IF Input
Maximum Safe Level+7 dBm
Center Frequency
IF BW ≤25 MHz322.5 MHz
40 MHz IF path250 MHz
255 MHz IF path750 MHz
510 MHz IF path877.1484375 MHz
1000 MHz IF path750 MHz
BandwidthSupports all optional IFs
ADC Clipping Level
25, 255, or 510 MHz IF paths−15 dBm (nominal)
40 MHz IF path−20 dBm (nominal)
1000 MHz IF path−15dBm (nominal)
1 dB Gain Compression−2 dBm (nominal)
Gain Accuracy
a
20 to 30°CFull Range
IF BW ≤25 MHz±1.2 dB±2.5 dBSwept and narrowband
Wider IF BW±1.2 dB (nominal)
IF Frequency ResponseRMS (nominal)
CFWidth
322.5 MHz (10 MHz IF path)±5 MHz0.05 dB
322.5 MHz (25 MHz IF path)±12.5 MHz0.07 dB
250 MHz (40 MHz IF path)±20 MHz0.10 dB
750 MHz (255 MHz IF path)±127.5 MHz0.12 dB
877.1484375 MHz
±255 MHz0.15 dB
(510 MHz IF path)
750 MHz (1 GHz IF path)±500 MHz0.18 dB
Noise Figure
9 dB (nominal)
(322.5 MHz, swept operation high IF gain)
VSWRSee plot below.
a. The amplitude accuracy of a measurement includes this term and the accuracy with which the settings of correc-
tions model the loss of the external mixer.
96
Page 97
Standard Option EXM - External Mixing
Other External Mixing Specifications
External Mixer IF Input VSWR [Plot]
DescriptionSpecificationsSupplemental Information
LO Output
Frequency Range3.75 to 14.1 GHz
Output Power
3.75 to 8.72 GHz
7.8 to 14.1 GHz
a
b
c
Second Harmonic
Fundamental Feedthrough and
Undesired Harmonics
c
VSWR
20 to 30°CFull Range
+15.0 to 18.0 dBm+13.5 to 19 dBm
+14.0 to 18.5 dBmNot specified
−20 dB (nominal)
−30 dB (nominal)
1.8:1 (nominal)
d
a. The LO output port power is compatible with Keysight M1970 and 11970 Series mixers except for the 11970K.
The power is specified at the connector. Cable loss will affect the power available at the mixer. With nonKeysight/Agilent mixer units, supplied loss calibration data may be valid only at a specified LO power that may
differ from the power available at the mixer. In such cases, additional uncertainties apply.
b. LO Doubler = Off settings.
c. LO Doubler = On setting. Fundamental frequency = 3.9 to 7.05 GHz.
d. The reflection coefficient has a Rayleigh probability distribution from 3.75 GHz to 14.1 GHz with a median
VSWR of 1.22:1.
b
97
Page 98
Standard Option EXM - External Mixing
Other External Mixing Specifications
This chapter contains specifications for the Option LNP, Low Noise Path.
99
Page 100
Standard Option LNP - Low Noise Path Specifications
Specifications Affected by Low Noise Path
Specifications Affected by Low Noise Path
The low noise path is in use when all the following are true:
— The setting of the Microwave Path is "Low Noise Path Enabled"
— The start frequency is at least 3.5 GHz and the stop frequency is above 3.6 GHz
— The preamp is either not licensed, or set to Off, or set to “Low Band”
Specification NameInformation
Displayed Average Noise Level (DANL)See DANL specifications on page 44 of the core specifications.
Compression
VSWRThe magnitude will be very similar between LNP and non-LNP
Frequency ResponseSee specifications in this chapter. The specifications are very similar
Second Harmonic DistortionSee “Second Harmonic Distortion” on page 54 of the core
Third-Order Intermodulation
Other Input Related SpuriousSee “Spurious Responses” on page 52 of the core
a. The low noise path, when in use, does not substantially change the compression-to-noise dynamic range or the
TOI-to-noise dynamic range because it mostly just reduces losses in the signal path in front of all significant
noise, TOI and compression-affecting circuits. In other words, the compression threshold and the third-order
intercept both decrease, and to the same extent as that to which the DANL decreases.
Little change in dynamic range
operation, but the details, such as the frequencies of the peaks and
valleys, will shift.
to the normal path. But the details of the response can be quite
different, with the frequencies of the peaks and valleys shifting
between LNP and non-LNP operation. That means that any relative
measurements between, for example, a large signal measured without
LNP, and a small signal measured with LNP, could be subject to
relative frequency response errors as large as the sum of the
individual errors.
specifications.
Little change in dynamic range
specifications. This performance with LNP is not warranted, but is
nominally the same as non-LNP performance.
a
a
100
Loading...
+ hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.