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
N9030-90089
Edition
Edition 2, March 2021
Supersedes: January 2021
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.
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/pxa
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.
17
PXA 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
18
PXA Signal Analyzer
Definitions and Requirements
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.
19
PXA Signal Analyzer
Frequency and Time
Frequency and Time
DescriptionSpecificationsSupplemental Information
Frequency Range
Maximum Frequency
Option 5033.6 GHz
Option 5088.4 GHz
Option 51313.6 GHz
Option 52626.5 GHz
Option 54444 GHz
Option 55050 GHz
Preamp Option P033.6 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
Preamp
Off10 MHz2 Hz
On10 MHz9 kHz
AC Coupled
a
DC Coupled
20
PXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Band
Harmonic
LO Multiple (N
b
)Band Overlaps
c
Mixing Mode
0 (2 Hz to 3.6 GHz)
d
1−1Options 503, 508, 513, 526, 544,
550
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
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. AC Coupled only applicable to frequency Options 503, 508, 513, and 526.
b. 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.
c. 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.
d. 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.
21
PXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Precision Frequency Reference (Option EP1)
Accuracy±[(time since last adjustment ×
aging rate) + temperature
stability + calibration accuracy
a]b
Temperature Stability
−8
−8
Nominally linear
−10
±5 × 10
20 to 30°C
Full temperature range
Aging Rate
±1.5 × 10
±5 × 10
Total Aging
−7
1 Year
2 Years
Settability
Warm-up and Retrace
d
300 s after turn on
900 s after turn on
Achievable Initial Calibration Accuracy
e
±1 × 10
±1.5 × 10
±2 × 10
±4 × 10
−9
−8
−7
Nominal
±1 × 10
±1 × 10
−7
of final frequency
−8
of final frequency
Standby power to reference oscillatorNot supplied
Residual FM
(Center Frequency = 1 GHz
≤0.25 Hz × N
(nominal)
10 Hz RBW, 10 Hz VBW)
c
/day (nominal)
f
p-p in 20 ms
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. Narrow temperature range performance is nominally linear with temperature. For example, for
25±3º C, the stability would be only three-fifths as large as the warranted 25±5º C, thus ±0.9 × 10
−8
.
d. Standby mode does not apply power to the oscillator. Therefore warm-up applies every time the power is
turned on. The warm-up reference is one hour after turning the power on. Retracing also occurs every time
warm-up occurs. The effect of retracing is included within the “Achievable Initial Calibration Accuracy” term of
the Accuracy equation.
22
PXA Signal Analyzer
Frequency and Time
e. 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
f. N is the LO multiplication factor.
DescriptionSpecificationsSupplemental Information
Precision Frequency Reference (Option EP0)
Accuracy±[(time since last adjustment ×
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.
23
PXA Signal Analyzer
Frequency and Time
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.
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.
c
24
PXA Signal Analyzer
Frequency and Time
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 >1GHz.
DescriptionSpecificationsSupplemental Information
Frequency Span
Range
Option 5030 Hz, 10 Hz to 3.6 GHz
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
25
a
)
a
)
PXA Signal Analyzer
Frequency and Time
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.
26
PXA 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
−40 to −10 dBm plus attenuation (nominal)
c
b
Absolute±2 dB + Absolute Amplitude Accuracy (nominal)
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.
28
PXA 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
29
PXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Resolution Bandwidth (RBW)
Range (−3.01 dB bandwidth)
Standard
With Option B85 and Option RBE
With Option B1X and Option RBE
a
a
With Option B2X or B5X and Option RBE
Power bandwidth accuracy
b
RBW RangeCF Range
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.
10, 15, 20, 25, 30, 40, 50, 60, and
70 MHz, in Spectrum Analyzer mode and
zero span.
10, 15, 20, 25, 30, 40, 50, 60, 70, 80,
100, and 133 MHz, in Spectrum Analyzer
m o d e a n d z e ro s p a n .
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.
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 −0.2dB (nominal)
4 to 8 MHz< 3.6 GHz0 to −0.4dB (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)
30
PXA Signal Analyzer
Frequency and Time
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.
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 tol-
erance 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 by nominally 6%. This
widening declines to 0.6% nominal when the Swp Time Rules key is set to Accuracy instead of Normal. 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
Standard25 MHz
With Option B4040 MHz
With Option B8585 MHz
With Option B1X160 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.
31
PXA Signal Analyzer
Frequency and Time
DescriptionSpecificationsSupplemental Information
Preselector BandwidthRelevant to many options, such as B1X Wide IF Bandwidth, in
Bands 1 and higher. Nominal.
Mean Bandwidth at CF
a
Freq option ≤ 526Freq option > 526
5 GHz58 MHz46 MHz
10 GHz57 MHz52 MHz
15 GHz59 MHz53 MHz
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
32
PXA Signal Analyzer
Amplitude Accuracy and Range
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
Measurement Range
Preamp OffDisplayed Average Noise Level to +30 dBm
Preamp On
RF (Option 503)Displayed Average Noise Level to +30 dBm
μW (Options 508, 513, 526)Displayed Average Noise Level to +24 dBm
mmW (Options 544, 550)Displayed Average Noise Level to +20 dBm
Input Attenuation Range0 to 70 dB, in 2 dB steps
DescriptionSpecificationsSupplemental Information
Maximum Safe Input LevelApplies with or without preamp
(Options P03, 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
+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
33
PXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
Marker Readout
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)
34
PXA Signal Analyzer
Amplitude Accuracy and Range
Frequency Response
DescriptionSpecificationsSupplemental Information
Frequency ResponseRefer to the footnote for
(Maximum error relative to reference
condition (50 MHz)
Mechanical attenuator only
Swept operation
c
b
Attenuation 10 dB)
Option 544, or 550 (mmW)
Option 503, 508, 513, or 526 (RF/μW)
20 to 30°CFull range95th Percentile (≈2σ)
Band Overlaps on
page 21
Freq Option 526 only: Modes
above 18 GHz
.
a
3 Hz to 10 MHzx
d
10 to 20 MHz
10 to 20 MHz
20 to 50 MHz
d
20 to 50 MHz
x±0.46 dB±0.54 dB
x±0.35 dB±0.44 dB
x±0.46 dB±0.54 dB
x±0.35 dB±0.44 dB±0.16 dB
x±0.35 dB±0.44 dB±0.19 dB
50 MHz to 3.6 GHzx±0.35 dB±0.44 dB±0.16 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
fg
fg
fg
fg
fg
fg
x±0.35 dB±0.47 dB±0.15 dB
See note
x±1.5 dB±2.5 dB±0.54 dB
x±1.7 dB±3.5 dB±0.70 dB
x±1.5 dB±2.5 dB±0.54 dB
x±1.5 dB±2.5 dB±0.57 dB
x±2.0 dB±2.7 dB±0.56 dB
x±2.0 dB±2.5 dB±0.54 dB
e
13.5 to 16 GHz (with Option EP0)
16 to 17 GHz (with Option EP0)
fg
fg
13.5 to 17.1 GHz (without Option EP0)
x
x
fg
x±2.0 dB±2.7 dB
±2.0 dB±2.7 dB±0.55 dB
±2.0 dB±3.0 dB±0.81 dB
35
±0.62 dB
PXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
13.5 to 17.1 GHz
fg
x±2.0 dB±2.7 dB±0.64 dB
Frequency Response (cont.)
17.0 to 22 GHz
17.0 to 22 GHz
22.0 to 26.5 GHz
22.0 to 26.5 GHz
26.4 to 34.5 GHz
34.4 to 50 GHz
fg
fg
fg
fg
fg
fg
x±2.0 dB±2.7 dB±0.65 dB
x±2.0 dB±2.8 dB±0.72 dB
x±2.5 dB±3.7 dB±0.82 dB
x±2.5 dB±3.5 dB±0.71 dB
x±2.5 dB±3.5 dB±1.00 dB
x±3.2 dB±4.9 dB±1.37 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. 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-
tion, 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 MHz is negligible, but not warranted.
e. 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. 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.
f. Specifications for frequencies >3.5 GHz apply for sweep rates ≤100 MHz/ms.
g. Preselector centering applied.
36
PXA Signal Analyzer
Amplitude Accuracy and Range
Nominal Frequency Response Band 0 without Option EP0 [Plot]
Nominal Frequency Response Band 0 with Option EP0 [Plot]
37
PXA 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)
(MHz)
Preselector
Max Errord
Midwidth Error
(95th Percentile)
Slope (dB/MHz)
(95th Percentile)
RMSe
(nominal)
c
Span
< 3.6≤10±0.20 dB±0.12 dB±0.100.02 dB
≥ 3.6, ≤ 26.5≤10 On0.23 dB
f
≥ 3.6, ≤ 26.5≤10
Off
±0.25 dB±0.12 dB±0.100.02 dB
≥ 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. Option MPB is installed and enabled.
38
PXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
IF Phase LinearityDeviation from mean phase linearity
Freq Option 526 only: Modes above 18 GHz
Without Option EP0With Option EP0
a
Center Freq
(GHz)
Span
(MHz)
PreselectorPeak-to-
peak
(nominal)
RMS
(nominal)
b
Peak-to
-peak
(nominal)
RMS
(nominal)
≥ 0.02, < 3.6≤10n/a0.06°0.012°0.17°0.037°
≥ 3.6≤10
Off
0.10°0.022°0.31°0.067°
c
≥ 3.6≤10On0.11°0.024°0.83°0.170°
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. Option MPB is installed and enabled.
b
39
PXA 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
95th Percentile Absolute
Amplitude Accuracy
a
±0.24 dB
±0.13 dB
(95th percentile)
±0.28 dB
a
±(0.24 dB + frequency response)
±(0.28 dB + frequency response)
±0.19 dB
b
(Wide range of signal levels,
RBWs, RLs, etc.,
0.01 to 3.6 GHz,
Atten = 10 dB)
Amplitude Reference Accuracy±0.05 dB (nominal)
Preamp On
c
±(0.36 dB + frequency response)
(P03, 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.
40
PXA 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 21
(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 50.0 GHz±1.0 dB (nominal)
41
PXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
RF Input VSWR
(at tuned frequency, DC coupled)
mmW (Option544, or 550)
RF/μW (Option 503, 508, 513, or 526)
10 dB atten, 50 MHz (ref condition) x1.07:1 (nominal)
10 dB atten, 50 MHz (ref condition)
0 dB atten, 0.01 to 3.6 GHzx
x1.09:1 (nominal)
x< 2.2:1 (nominal)
95th Percentile
Band 0 (0.01 to 3.6 GHz, 10 dB atten)x1.139
Band 0 (0.01 to 3.6 GHz, 10 dB atten)
x1.134
Band 1 (3.5 to 8.4 GHz, 10 dB atten)x1.290
Band 1 (3.5 to 8.4 GHz, 10 dB atten)
x1.152
Band 2 (8.3 to 13.6 GHz, 10 dB atten)x1.388
Band 2 (8.3 to 13.6 GHz, 10 dB atten)
x1.178
Band 3 (13.5 to 17.1 GHz, 10 dB atten)x1.41
Band 3 (13.5 to 17.1 GHz, 10 dB atten)
x1.212
Band 4 (17.0 to 26.5 GHz, 10 dB atten)x1.48
Band 4 (17.0 to 26.5 GHz, 10 dB atten)
Band 5 (26.4 to 34.5 GHz, 10 dB atten)
Band 6 (34.4 to 50 GHz, 10 dB atten)
x1.331
x1.373
x1.389
a
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
42
PXA Signal Analyzer
Amplitude Accuracy and Range
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.
Nominal VSWR Low Band [Plot]
43
PXA Signal Analyzer
Amplitude Accuracy and Range
Nominal VSWR, above 3.5 GHz [Plot]
44
PXA Signal Analyzer
Amplitude Accuracy and Range
DescriptionSpecificationsSupplemental Information
Resolution Bandwidth Switching UncertaintyRelative to reference BW of 30 kHz,
1.0 Hz to 1.5 MHz RBW±0.03 dB
verified in low band
a
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
Switching between Linear and Log
Log Scale Switching
0 dB
0 dB
a
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.
45
PXA 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
Typical
−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.
46
PXA 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)] −
x
exp[(P2 − Pref)/(8.69 dB)]} (exp is the natural exponent function, e
). 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 PXA. 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
47
PXA Signal Analyzer
Dynamic Range
Dynamic Range
Gain Compression
DescriptionSpecificationsSupplemental Information
1 dB Gain Compression Point (Two-tone)
abc
Maximum power at mixer
d
20 to 30°CFull range
20 to 40 MHz−3 dBm−4 dBm0 dBm (typical)
40 to 200 MHz+1 dBm−1 dBm+3 dBm (typical)
200 MHz to 3.6 GHz+3 dBm+2 dBm+5 dBm (typical)
3.6 to 16 GHz+1 dBm0 dBm+4 dBm (typical)
16 to 26.5 GHz−1 dBm−3 dBm+2 dBm (typical)
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
Low frequency exceptions
+12 dBm (nominal)
Gain set to Low
IF Prefilter Bandwidth
e
Zero Span or
f
Swept
, RBW =
Sweep Type = FFT,
FFT Width =
–3 dB Bandwidth
(nominal)
≤ 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
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.
48
PXA 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, 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 FS1 or 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.
49
PXA Signal Analyzer
Dynamic Range
Displayed Average Noise Level
DescriptionSpecificationsSupplemental
Information
DANL without Noise Floor Extension and without
Option EP0
a
Input terminated
Sample or Average detector
Averaging type = Log
0 dB input attenuation
Refer to the footnote
for
Band Overlaps on
page 21.
IF Gain = High
1 Hz Resolution Bandwidth
mmW with no signal path options
mmW with one or more signal path options
b
c
RF/μW without Option EP0
20 to 30°CFull rangeTypical
3 to 10 Hz
10 to 100 Hz
100 Hz to 1 kHz
1 to 9 kHzx
9 to 100 kHzx
100 kHz to 1 MHzx
1 to 10 MHz
d
10 MHz to 1.2 GHzx
xxx–100 dBm (nominal)
xxx–125 dBm (nominal)
xxx–130 dBm (nominal)
xx–145 dBm (nominal)
xx−146 dBm−146 dBm−151 dBm
xx−150 dBm−150 dBm−156 dBm
xxx−155 dBm−152 dBm−158 dBm
xx−154 dBm−152 dBm−155 dBm
1.2 to 2.1 GHzx
2.1 to 3 GHzx
3.0 to 3.6 GHzx
3.5 to 4.2 GHz
3.5 to 4.2 GHz
3.5 to 4.2 GHz
4.2 to 8.4 GHz
4.2 to 6.6 GHz
4.2 to 6.6 GHz
6.6 to 8.4 GHz
6.6 to 8.4 GHz
8.3 to 13.6 GHz
xx−153 dBm−152 dBm−155 dBm
xx−151 dBm−150 dBm−153 dBm
xx−151 dBm−149 dBm−153 dBm
x−147 dBm−146 dBm−150 dBm
x−143 dBm−141 dBm−147 dBm
x–145 dBm –143 dBm–148 dBm
x−150 dBm−148 dBm−152 dBm
x−144 dBm−142 dBm−148 dBm
x–146 dBm–144 dBm–149 dBm
x−147 dBm−145 dBm−149 dBm
x–149 dBm–147 dBm–151 dBm
x−149 dBm−147 dBm−151 dBm
8.3 to 13.6 GHzx−147 dBm−145 dBm−149 dBm
50
PXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental
Information
8.3 to 13.6 GHz
13.5 to 16.9 GHz
13.5 to 14 GHz
13.5 to 14 GHz
14 to 17 GHz
14 to 17 GHz
16.9 to 20 GHz
17.0 to 22.5 GHz
17.0 to 22.5 GHz
20.0 to 26.5 GHz
22.5 to 26.5 GHz
22.5 to 26.5 GHz
26.4 to 30 GHz
26.4 to 30 GHz
30 to 34 GHz
x−149 dBm−147 dBm−151 dBm
x−145 dBm−143 dBm−147 dBm
x−143 dBm−141 dBm−146 dBm
x−145 dBm−143 dBm−148 dBm
x−145 dBm−143 dBm−148 dBm
x−147 dBm−145 dBm−150 dBm
x-143 dBm-140 dBm-145 dBm
x−141 dBm−139 dBm−146 dBm
x−145 dBm−143 dBm−148 dBm
x−137 dBm−135 dBm−140 dBm
x−139 dBm−137 dBm−143 dBm
x−142 dBm−140 dBm−145 dBm
x−138 dBm−136 dBm−142 dBm
x−141 dBm−139 dBm−145 dBm
x−138 dBm−135 dBm−142 dBm
30 to 34 GHz
33.9 to 37 GHz
33.9 to 37 GHz
37 to 40 GHz
37 to 40 GHz
40 to 49 GHz
40 to 46 GHz
46 to 49 GHz
49 to 50 GHz
49 to 50 GHz
Additional DANL, IF Gain = Low
e
xxx–164.5 dBm (nominal)
x−141 dBm−138 dBm−144 dBm
x−134 dBm−131 dBm−139 dBm
x−137 dBm−133 dBm−142 dBm
x−132 dBm−129 dBm−138 dBm
x−136 dBm−133 dBm−141 dBm
x−130 dBm−126 dBm−135 dBm
x−136 dBm−132 dBm−140 dBm
x−133 dBm−129 dBm−138 dBm
x−128 dBm−124 dBm−133 dBm
x−133 dBm−129 dBm−137 dBm
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. Specifications marked with an x in this column apply to analyzers with mmW frequency options (Option 544 or
550) and none of the following options that affect the signal path: MPB, LNP, B85, B1X, B2X, or B5X.
c. Specifications marked with an x in this column apply to analyzers with mmW frequency options (Option 544 or
550) and one or more of the following options that affect the signal path: MPB, LNP, B85, B1X.
51
PXA Signal Analyzer
Dynamic Range
d. 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.
e. 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.
52
PXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
DANL without Noise Floor Extension, with
Option EP0
a
Input terminated
Sample or Average detector
Averaging type = Log
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. 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.
d. 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.
54
PXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Displayed Average Noise Level with Noise Floor
Extension Improvement
a
mmW with no signal path options
b
mmW with one or more signal path optionsc 95th Percentile (≈2σ)d
RF/μW (Option 503, 508, 513,or 526)
Band 0, f > 20 MHz
Band 0, f > 20 MHz
Band 0, f > 20 MHz
f
f
f
x9 dB10 dB
x10 dB9 dB
x10 dB9 dB
Preamp OffPreamp On
Band 1x10 dB9 dB
Band 1
Band 1
x9 dB9 dB
x10 dB9 dB
Band 2x10 dB10 dB
Band 2
Band 2
x9 dB8 dB
x9 dB8 dB
e
Band 3x9 dB9 dB
Band 3
Band 3
x9 dB8 dB
x10 dB8 dB
Band 4x10 dB8 dB
Band 4
Band 4
Band 5
Band 5
Band 6
Band 6
Improvement for CW Signals
g
Improvement, Pulsed-RF Signals
h
x10 dB9 dB
x11 dB9 dB
x11 dB8 dB
x12 dB9 dB
x11 dB7 dB
x12 dB8 dB
3.5 dB (nominal)
10.8 dB (nominal)
Improvement, Noise-Like Signals9.1 dB (nominal)
55
PXA Signal Analyzer
Dynamic Range
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. Specifications marked with an x in this column apply to analyzers with mmW frequency options (Option 544 or
550) and none of the following options that affect the signal path: MPB, LNP, B85, B1X, B2X or B5X.
c. Specifications marked with an x in this column apply to analyzers with mmW frequency options (Option 544 or
550) and one or more of the following options that affect the signal path: MPB, LNP, B85, B1X, B2X or B5X.
d. 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.
e. 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.
f. NFE does not apply to the low frequency sensitivity. At frequencies below about 0.5 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 0.5 and 20 MHz the NFE effectiveness increases from nearly none to near its maximum.
g. 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.
h. 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.
56
PXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
Displayed Average Noise Level with Noise Floor Extension
mmW with no signal path options
mmW with one or more signal path options
b
c
RF/μW (Option 503, 508, 513, or 526)
Band 0, f > 20 MHz
f
xxx−163 dBm−174 dBm
a
95th Percentile (≈2σ)b
Preamp Off
Preamp On
Band 1 x−162 dBm−174 dBm
Band 1
Band 1
x−159 dBm−172 dBm
x−160 dBm−172 dBm
Band 2 x−162 dBm−173 dBm
Band 2
Band 2
x−159 dBm−172 dBm
x−161 dBm−173 dBm
Band 3 x−156 dBm−172 dBm
c
Band 3
Band 3
x−159 dBm−173 dBm
x−161 dBm−174 dBm
Band 4 x−150 dBm−166 dBm
Band 4
Band 4
Band 5
Band 5
Band 6
Band 6
x−154 dBm−169 dBm
x−158 dBm−171 dBm
x−153 dBm−167 dBm
x−157 dBm−168 dBm
x−144 dBm−158 dBm
x−149 dBm−161 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.
57
PXA Signal Analyzer
Dynamic Range
c. 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.
58
PXA Signal Analyzer
Dynamic Range
Spurious Responses
DescriptionSpecificationsSupplemental Information
Spurious Responses: Residual
and Image
Preamp Off
Option EP0 effects
a
b
(see Band Overlaps on page 21)
Residual Responses
c
200 kHz to 8.4 GHz (swept)
Zero span or FFT or other frequencies
−100 dBm
−100 dBm
(nominal)
Image Responses
Tuned Freq (f)Excitation
Freq
Mixer
Level
ResponseResponse (typical)
d
RF/μWmmWRF/μWmmW
10 MHz to 26.5 GHzf+45 MHz−10 dBm−80 dBc−80 dBc−118 dBc
−105 dBc (EP0)
10 MHz to 3.6 GHzf+10245 MHz−10 dBm−80 dBc−80 dBc−112 dBc
13.5 to 17.1 GHzf+645 MHz−10 dBm−74 dBc−80 dBc−84 dBc−102 dBc
17.0 to 22 GHzf+645 MHz−10 dBm−70 dBc−80 dBc−82 dBc−100 dBc
22 to 26.5 GHzf+645 MHz−10 dBm−68 dBc−70 dBc−79 dBc
−75 dBc (EP0)
−97 dBc
−90 dBc (EP0)
26.5 to 50 GHzf+45 MHz−30 dBm–90 dBc
(nominal)
26.5 to 34.5 GHzf+645 MHz−30 dBm–70 dBc–94 dBc
34.4 to 42 GHzf+645 MHz−30 dBm–59 dBc–79 dBc
−76 dBc (EP0)
42 to 50 GHzf+645 MHz−30 dBm–75 dBc
(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. Where the presence of Option EP0 affects the performance, it is noted with "(EP0)" text.
c. Input terminated, 0 dB input attenuation.
d. Mixer Level = Input Level − Input Attenuation.
59
PXA Signal Analyzer
Dynamic Range
e. We also support the following additional spurious responses specifications from 8 to 12 GHz at 20 to 30º C. Image
responses are warranted to be better than –82 dBc, with 95th percentile performance of –88.3 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.
DescriptionSpecificationsSupplemental Information
Spurious Responses: Other
First RF Order
Mixer Level
b
a
Response
Response (typical)
(f ≥ 10 MHz from carrier)
Carrier Frequency ≤ 26.5 GHz −10 dBm
−80 dBc + 20 × log(N
c
)
Includes IF feedthrough, LO
harmonic mixing responses
Carrier Frequency > 26.5 GHz −30 dBm–90 dBc (nominal)
Higher RF Order
d
(f ≥ 10 MHz from carrier)
c
Includes higher order mixer
Carrier Frequency ≤ 26.5 GHz −40 dBm
−80 dBc + 20 × log(N
)
responses
Carrier Frequency > 26.5 GHz
LO-Related Spurious Responses
e
−30 dBm−90 dBc (nominal)
fe
−10 dBm
−68 dBc
+ 20 × log(Nc)−72 dBc + 20 × log(Nc) (typical)
(Offset from carrier 200 Hz to 10 MHz)
Close-in Sidebands Spurious Response
(LO Related, offset < 200 MHz)
−73 dBc
(nominal)
e
+ 20 × log(Nc)
a. Mixer Level = Input Level − Input Attenuation.
b. 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.
c. N is the LO multiplication factor.
d. RBW=100 Hz. With higher RF order spurious responses, the observed frequency will change at a rate faster than
the input frequency.
e. Nominally −40 dBc under large magnetic (0.38 Gauss rms) or vibrational (0.21 g rms) environmental stimuli.
f. We also support the following additional spurious responses specifications from 8 to 12 GHz at 20 to 30º C.
Image responses are warranted to be better than –82 dBc, with 95th percentile performance of –88.3 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.
60
PXA Signal Analyzer
Dynamic Range
Second Harmonic Distortion
DescriptionSpecificationsSupplemental Information
Second Harmonic Distortion
mmW (Option 544, or 550)
RF/μW (Option 503, 508, 513, or 526)
Mixer
Level
Distortion
a
SHI
Distortion
(nominal)
bc
Source Frequency
0 to 100 MHzxx–15 dBm–57 dBc+42 dBm
d
to 3 GHz
d
x
x–15 dBm–60 dBc+45 dBm
x–15 dBm–77 dBc+62 dBm
x–15 dBm–72 dBc+57 dBm
x−15 dBm–77 dBc+62 dBm
x−15 dBm–70 dBc+55 dBm
x−15 dBm–62 dBc+47 dBm
x−15 dBm−65 dBc+50 dBm
0.1 to 1.8 GHz
1.75
1.75 to 3 GHz
3 to 6.5 GHzx
6.5 to 10 GHzx
10 to 13.25 GHzx
13.2 to 25 GHz
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 distor-
tion level relative to the mixer tone in dBc.
c. Performance >3.6 GHz improves greatly with Option LNP enabled. See Option LNP - Low Noise Path
Specifications on page 225.
d. These frequencies are half of the band edge frequencies. See Band Overlaps on page 21.
bc
SHI
(nominal)
61
PXA 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 21.
Refer to footnote
Distortion".
Verification conditionsc)
mmW Option544, or 550
RF/μW Option 503, 508, 513, or 526
20 to 30°C
10 to 150 MHzx
150 to 600 MHz
f
600 MHz to 1.1 GHzx
1.1 to 3.6 GHz
g
Intercept
x+13 dBm+16 dBm
xx+18 dBm+21 dBm
x+20 dBm+22 dBm
xx+21 dBm+23 dBm
e
Intercept (typical)
3.5 to 8.4 GHzx+17 dBm+23 dBm
d
for the "Extrapolated
3.6 to 3.7 GHz (Band 0)xSee note
3.5 to 8.4 GHzx+16 dBm+23 dBm
8.3 to 13.6 GHz x+17 dBm+23 dBm
8.3 to 13.6 GHz
x+16 dBm+23 dBm
13.5 to 17.1 GHz x+15 dBm+20 dBm
13.5 to 17.1 GHz
17.0 to 26.5 GHzi
17.0 to 26.5 GHz
i
26.4 to 34.5 GHzx
34.4 to 50 GHz
g
x+13 dBm+17 dBm
x+16 dBm+22 dBm
x+13 dBm+20 dBm
+13 dBm+18 dBm
x
+10 dBm+15 dBm
Full temperature range
10 to 150 MHzx
150 to 600 MHz
j
x+12 dBm
xx+17 dBm
h
62
PXA Signal Analyzer
Dynamic Range
DescriptionSpecificationsSupplemental Information
600 MHz to 1.1 GHzx
1.1 to 3.6 GHz
k
x+18 dBm
xx+19 dBm
Third Order Intermodulation
Full temperature Range (cont.)
3.5 to 8.4 GHzx+14 dBm
3.5 to 8.4 GHz
x+13 dBm
8.3 to 13.6 GHz x+14 dBm
8.3 to 13.6 GHz
x+13 dBm
13.5 to 17.1 GHz x+13 dBm
13.5 to 17.1 GHz
x+10 dBm
17.0 to 26.5 GHzi x+14 dBm
17.0 to 26.5 GHz
i
x+10 dBm
26.4 to 36.5 GHzx+9 dBm
34.4 to 50 GHz
l
x+6 dBm
a. See the IF Prefilter Bandwidth table in the Gain Compression specifications on page 48. When the tone separa-
tion 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
a
note
2
is often suitable for best TOI performance, because of how it affects the IF Prefilter settings Foot-
links to the details.
c. TOI is verified with two tones, each at −16 dBm (10 MHz to 26.5 GHz) and −20 dBm (26.5 GHz to 50 GHz) at the
mixer, spaced by 100 kHz.
d. 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 PXA, 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 curvature of the third-order intermodulation line in the 1 GHz graph that follows, which is representative of performance below 3.6 GHz.
e. 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.
f. For Option EP0:
150 to 300 MHz +10 dBm +19 dBm (typical)
300 to 600 MHz +18 dBm +20 dBm (typical)
63
PXA Signal Analyzer
Dynamic Range
g. For mmW PXA with Option EP0:
1.1 to 1.5 GHz +20 dBm +22 dBm (typical)
1.5 to 3.6 GHz +21 dBm +23 dBm (typical)
34.4 to 50 GHz +10 dBm +13 dBm (typical)
For mmW PXA with Option MPB:
34.4 to 50 GHz +10 dBm +15 dBm (typical)
For mmW PXA without Option MPB:
34.4 to 50 GHz +9 dBm +15 dBm (typical)
h. 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. 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.
i. Intercept performance is nominally 3 dB better in this band in those analyzers which have either Option LNP or
Option MPB installed, or both, when these options are not in use.
j. For Option EP0:
150 to 300 MHz +15 dBm
300 to 600 MHz +17 dBm
k. For mmW PXA with Option EP0:
1.1 to 1.5 GHz +18 dBm
1.5 to 3.6 GHz +19 dBm
l. For mmW PXA with Option MPB:
34.4 to 50 GHz +6 dBm
For mmW PXA without Option MPB:
34.4 to 50 GHz +5 dBm
Nominal TOI vs. Input Frequency and Tone Separation [Plot]
a
a. This plot is not applicable to Option EP0.
64
PXA Signal Analyzer
Dynamic Range
Nominal Dynamic Range vs. Offset Frequency vs. RBW [Plot]
a
a. This plot is not applicable to Option EP0.
65
PXA Signal Analyzer
Dynamic Range
Nominal Dynamic Range vs. Offset Frequency vs. RBW [Plot]
a
a. This plot is only applicable to Option EP0.
66
PXA Signal Analyzer
Dynamic Range
Phase Noise
DescriptionSpecificationsSupplemental Information
Phase Noise with Option EP1Noise Sidebands
(Center Frequency = 1 GHz
Optimization
Internal Reference
a. The nominal performance of the phase noise at center frequencies different than the one at which the specifica-
tions apply (1 GHz) depends on the center frequency, band and the offset. For low offset frequencies, offsets
well under 100 Hz, the phase noise increases by 20 × log[(f + 0.3225)/1.3225]. For mid-offset frequencies such
as 50 kHz, phase noise trends as 20 × log[(f + 5.1225)/6.1225], and also varies chaotically an additional nominally ±2 dB versus the center frequency. 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 21; f is in GHz units in all these relationships;
all increases are in units of decibels.
b. Noise sidebands for lower offset frequencies, for example, 10 kHz, apply with the phase noise optimization
(PhNoise Opt) set to Best Close-in φ Noise. Noise sidebands for higher offset frequencies, for example, 1 MHz,
as shown apply with the phase noise optimization set to Best Wide-offset φ Noise.
c. 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 setting 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. For
more information, see Phase Noise Effects, Ext Ref vs. Loop BW [Plot] on page 74.
d. Nominal phase noise was –75 dBc/Hz for instruments produced before approximately September 1, 2012.
d
67
PXA Signal Analyzer
Dynamic Range
e. 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 rapidly. 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.
DescriptionSpecificationsSupplemental Information
Phase Noise with Option EP0
Noise Sidebands
a
(Center Frequency = 1 GHzb Best-case
Optimization
Internal Reference
a. Noise sidebands around a signal are dominantly phase noise sidebands. With the extremely low phase noise of
the PXA, 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 specifica-
tions apply (1 GHz) depends on the center frequency, band and the offset. For low offset frequencies, offsets
well under 100 Hz, the phase noise increases by 20 × log[(f + 0.3225)/1.3225], where f is the larger of 0.5 and
the center frequency in GHz units. For mid-offset frequencies such as 50 kHz, phase noise trends as
20 × log[(f + 5.1225)/6.1225], where f is the larger of 1.0 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 21.
68
PXA Signal Analyzer
Dynamic Range
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.
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 setting 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. For
more information, see Phase Noise Effects, Ext Ref vs. Loop BW [Plot] on page 74.
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 characterized 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 consistent with a warranted specification of –90 dBc/Hz.
69
PXA Signal Analyzer
Dynamic Range
Nominal Phase Noise of Different LO Optimizations without Option EP0 [Plot]
70
PXA Signal Analyzer
Dynamic Range
Nominal Phase Noise at Different Center Frequencies without Option EP0 [Plot]
71
PXA Signal Analyzer
Dynamic Range
Nominal Phase Noise at Different Carrier Frequencies, Phase Noise Optimized vs Offset Frequency with Option EP0
[Plot]
72
PXA Signal Analyzer
Dynamic Range
Nominal Phase Noise at Different Phase Noise/Spurs Optimization with Option EP0 [Plot]
73
PXA Signal Analyzer
Dynamic Range
Phase Noise Effects, Ext Ref vs. Loop BW without Option EP0 [Plot]
The effect of the Ext Ref Loop BW control (Narrow and Wide) is shown in this graphic. When set to Wide, the noise from the
internal circuitry is reduced, but noise in the external reference is subject to being impressed on the LO through the transfer
function shown in the smooth curve labeled "Wide." For an excellent reference, this can give lower overall noise. When the
Narrow selection is made, the internal noise effect is higher, but external reference noise above 20 Hz is rejected.
The noise curves were measured at 1 GHz center frequency with an excellent reference and excellent RF signal, and is thus
a conservative estimate of the residual noise of the PXA circuitry. At that center frequency, the transfer function curves
approach 40 dB gain to phase noise at low offset frequencies for a 10 MHz external reference. (This 40 dB is computed as
20 × log
10 ^(fC/f
10
). The measured noise curves will scale with frequency the same way.)
REF
Example: Consider an external reference at 10 MHz with phase noise of −135 dBc/Hz at 20 Hz offset. If the Narrow setting
is chosen, the analyzer noise density will be −86 dBc/Hz at 20 Hz offset, the gain to the reference will be 40 dB, giving
−95 dBc/Hz contribution. Add these together on a power scale as 10 x log
(–86/10)
(10
10
+ 10
(–95/10)
) = −85.5 dBc/Hz. If
Wide is chosen, the analyzer noise density will be −97 dBc/Hz at 20 Hz offset, the gain to the reference will be 42 dB, giving
−93 dBc/Hz contribution. Add those together on a power scale as 10 x log
(–97/10)
(10
10
+ 10
(–93/10)
) = −91.5 dBc/Hz.
"Wide" will give a 6 dB superior result to the Narrow selection.
74
PXA Signal Analyzer
Power Suite Measurements
Power Suite Measurements
The specifications for this section apply only to instruments with Frequency
Option508, 513, or 526. For instruments with higher frequency options, the
performance is nominal only and not subject to any warranted specifications.
The measurement performance is only slightly different between instruments
with the higher frequency options. Because the hardware performance of the
analyzers is very similar but not identical, you can estimate the nominal
performance of the measurements from the specification in this chapter.
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 40.
b. See “Frequency and Time” on page 20.
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
75
PXA 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
h
level
MS (UE)10 MHz±0.09 dBAt ACPR range of −40 to −46 dBc with optimum mixer
i
level
BTS5 MHz±0.22 dBAt ACPR range of −42 to −48 dBc with optimum mixer
j
level
BTS10 MHz±0.18 dBAt ACPR range of −47 to −53 dBc with optimum mixer
i
level
BTS5 MHz±0.10 dB
At −48 dBc non-coherent ACPR
k
Option EP0
RadioOffset Freq
MS (UE)5 MHz±0.09 dB At ACPR range of −30 to −36 dBc with optimum mixer
h
level
MS (UE)10 MHz±0.11 dBAt ACPR range of −40 to −46 dBc with optimum mixer
i
level
BTS5 MHz±0.25 dBAt ACPR range of −42 to −48 dBc with optimum mixer
j
level
BTS10 MHz±0.25 dBAt ACPR range of −47 to −53 dBc with optimum mixer
i
level
BTS5 MHz±0.12 dB
At −48 dBc non-coherent ACPR
k
76
PXA Signal Analyzer
Power Suite Measurements
DescriptionSpecificationsSupplemental Information
Adjacent Channel Power (cont.)
Dynamic RangeRRC weighted, 3.84 MHz noise
bandwidth
Noise
Correction
Offset
l
FreqMethodACLR (typical)
m
ACLR (EP0)
m
(typical)
Optimum MLn
(nominal)
Off5 MHzFiltered IBW−81.5 dB−80 dB−8 dBm
Off5 MHzFast−81 dB−80 dB−8 dBm
Off10 MHzFiltered IBW−87 dB−86 dB−4 dBm
On5 MHzFiltered IBW−82.5 dB−81.5 dB−8 dBm
On10 MHzFiltered IBW−88.0 dB−87 dB−4 dBm
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 specified
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 speci-
fications table, the optimum mixer drive level for accuracy is approximately −37 dBm − (ACPR/3), where the
ACPR is given in (negative) decibels.
77
PXA Signal Analyzer
Power Suite Measurements
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.
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 pass-
band 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 accommodate 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 spec-
trum 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 testing
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.
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
79
PXA Signal Analyzer
Power Suite Measurements
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)
Description SpecificationsSupplemental Information
Spurious EmissionsTable-driven spurious signals;
search across regions
Case: Radio Std = 3GPP W-CDMA
Dynamic Range
a
, relative (RBW=1 MHz)
88.8 dB 91.8 dB (typical)
(1 to 3.6 GHz)
Dynamic Range
a
, relative (RBW=1 MHz)
86.9 dB 89.9 dB (typical)
(1 to 3.6 GHz) (Option EP0)
Sensitivity
b
, absolute (RBW=1 MHz)
−88.5 dBm−91.5 dBm (typical)
(1 to 3.6 GHz)
Sensitivity
b
, absolute (RBW=1 MHz)
−86.5 dBm−89.5 dBm (typical)
(1 to 3.6 GHz) (Option EP0)
AccuracyAttenuation = 10 dB
20 Hz to 3.6 GHz±0.19 dB (95th percentile)
3.5 to 8.4 GHz±1.09 dB (95th percentile)
3.5 to 8.4 GHz (Option EP0)±1.15 dB (95th percentile)
8.3 to 13.6 GHz±1.48 dB (95th percentile)
8.3 to 13.6 GHz (Option EP0)±1.52 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.
80
PXA Signal Analyzer
Power Suite Measurements
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 40 for more information. The numbers shown are for 0 to 3.6
GHz, with attenuation set to 10 dB.
82
PXA Signal Analyzer
Options
Options
The following options and applications affect instrument specifications.
Option 503:Frequency range, 2 Hz to 3.6 GHz
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 B1X:Analysis bandwidth, 160
Standard B25:Analysis bandwidth, 25 MHz
Option B40:Analysis bandwidth, 40 MHz
Option B85:Analysis bandwidth, 85 MHz
Option B2XAnalysis bandwidth, 255 MHz
Option B5X:Analysis bandwidth, 510 MHz
Option BBA:BBIQ inputs, analog
Option C35:APC 3.5 mm connector (for Freq Option 526 only)
Option CR3:Connector Rear, second IF Out
Option CRP:Connector Rear, arbitrary IF Out
Option EA3:Electronic attenuator, 3.6 GHz
Option EMC:Precompliance EMC Features
Option EP0:Enhanced Phase noise, DDS LO
Option ESC:External source control
Option EXM:External mixing
Option LNP:Low Noise Path
Option MPB:Preselector bypass
Standard NFE:Noise floor extension, instrument alignment
Option P03:Preamplifier, 3.6 GHz
Option P08:Preamplifier, 8.4 GHz
Option P13:Preamplifier, 13.6 GHz
Option P26:Preamplifier, 26.5 GHz
83
PXA Signal Analyzer
Options
Option P44:Preamplifier, 44 GHz
Option P50:Preamplifier, 50 GHz
Option RT1:Real-Time analysis up to 160 MHz, basic detection
Option RT2:Real-Time analysis up to 160 MHz, optimum detection
Option RTS:Real-Time wideband I/Q data streaming
Option YAS:Y-Axis Screen Video output
Option YAV:Y-Axis Video output
N9054EM0E:
N9054EM1E:
EM0E:
N9063
N9067
EM0E:
N9068
EM0E:
N9069
EM0E:
N9071
EM0E:
N9073
EM0E:
N9077
EM0E:
N9080
EM0E:
N9081
EM0E:
N9082
EM0E:
N9083
EM0E:
N9084
EM0E:
Flexible Digital Demodulation measurement application
Custom OFDM measurement application
Analog Demodulation measurement application
Pulse measurement application
Phase Noise measurement application
Noise Figure measurement application
GSM/EDGE/EDGE Evolution measurement application
W-CDMA/HSPA/HSPA+ measurement application
WLAN measurement application
LTE-Advanced FDD measurement application
Bluetooth Measurement application
LTE-Advanced TDD measurement application
MSR measurement application
Short Range Communications measurement application
N9085
N9092
EM0E:
EM0E:
5G NR measurement application
Avionics measurement application
84
PXA Signal Analyzer
General
General
DescriptionSpecificationsSupplemental Information
Calibration Cycle1 year
DescriptionSpecificationsSupplemental Information
Environment
Indoor
Temperature Range
Operating
Altitude ≤ 2,300 m0 to 55°C
Altitude = 4,600 m0 to 47°C
Derating
Storage
Altitude
Humidity
Relative humidity95% relative humidity, non-condensing at
DescriptionSpecificationsSupplemental Information
Environmental and Military
Specifications
a
−40 to +70°C
4,600 m (approx 15,000 feet)
40°C, decreasing linearly to 50% relative
humidity at 55°C.
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.
85
PXA Signal Analyzer
General
DescriptionSpecificationSupplemental Information
Acoustic NoiseValues given are per ISO 7779 standard in the "Operator Sitting"
position
Ambient Temperature
< 40°CNominally under 55 dBA Sound Pressure. 55 dBA is generally
considered suitable for use in quiet office environments.
≥ 40°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.)
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, On630 WMaximum
Power Consumption, Standby45 WStandby power is not supplied to the
frequency reference oscillator but to
the CPU.
Typical instrument configurationPower (nominal)
Base PXA instrument B85/B1X/B2X/B5X330 W
Adding Option B85/B1X to base instrument+45 W
Adding Option BBA to base instrument+46 W
86
PXA Signal Analyzer
General
DescriptionSupplemental Information
Measurement Speed
a
Local measurement and display update rate
Remote measurement and LAN transfer rate
bc
bc
Nominal
10 ms (100/s)
10 ms (100/s)
Marker Peak Search2.5 ms
Center Frequency Tune and Transfer (Band 0)43 ms
Center Frequency Tune and Transfer (Bands 1-4)69 ms
Measurement/Mode Switching40 ms
W-CDMA ACLR measurement timeSee page 76
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 integer format, markers Off, single sweep, mea-
sured with IBM compatible PC with 2.99 GHz Pentium® 4 with 2 GB RAM running Windows® XP, Keysight I/O
Libraries Suite Version 14.1, one meter GPIB cable, National Instruments PCI-GPIB Card and NI-488.2 DLL.
DescriptionSpecificationsSupplemental Information
Display
a
Resolution1280 × 800Capacitive multi-touch screen
Size269 mm (10.6 in) diagonal (nominal)
a. The LCD display is manufactured using high precision technology. However, if a static image is displayed for a
lengthy period of time (~2 hours) you might encounter "image sticking" that may last for approximately 2 seconds. This is normal and does not affect the measurement integrity of the product in any way.
DescriptionSpecificationsSupplemental Information
Data Storage
Internal TotalRemovable solid state drive (≥ 160 GB)
Internal User≥ 9 GB available on separate partition for user
data
87
PXA Signal Analyzer
General
DescriptionSpecificationsSupplemental Information
WeightWeight without options
Net22 kg (48 lbs) (nominal)
Shipping 34 kg (75 lbs) (nominal)
Cabinet DimensionsCabinet dimensions exclude front and rear
Height177 mm (7.0 in)
Width426 mm (16.8 in)
Length556 mm (21.9 in)
protrusions.
88
PXA Signal Analyzer
Inputs/Outputs
Inputs/Outputs
Front Panel
DescriptionSpecificationsSupplemental Information
RF Input
Connector
StandardType-N femaleFrequency Option 503, 508, 513, and 526
2.4 mm maleFrequency Option 544, and 550
Option C353.5 mm maleFrequency Option 526 only
Impedance50Ω (nominal)
DescriptionSpecificationsSupplemental Information
Probe Power
Voltage/Current+15 Vdc, ±7% at 0 to 150 mA (nominal)
−12.6 Vdc, ±10% at 0 to 150 mA (nominal)
GND
DescriptionSpecificationsSupplemental Information
USB Ports
Host (3 ports)Compliant with USB 2.0
ConnectorUSB Type “A” (female)
Output Current
Port marked with Lightning Bolt1.2 A (nominal)
Port not marked with Lightning Bolt0.5 A
DescriptionSpecificationsSupplemental Information
Headphone Jack
Connectorminiature stereo audio jack3.5 mm (also known as "1/8 inch")
Output Power90 mW per channel into 16Ω (nominal)
89
PXA Signal Analyzer
Inputs/Outputs
Rear Panel
DescriptionSpecificationsSupplemental Information
10 MHz Out
ConnectorBNC female
Impedance50Ω (nominal)
Output Amplitude≥0 dBm (nominal)
Output ConfigurationAC coupled, sinusoidal
Frequency10 MHz ×
(1 + frequency reference accuracy)
DescriptionSpecificationsSupplemental Information
Ext Ref In
ConnectorBNC femaleNote: Analyzer noise sidebands and spurious
response performance may be affected by the
quality of the external reference used. See
footnote in the Phase Noise specifications within
the Dynamic Range section on page 67.
Impedance50Ω (nominal)
Input Amplitude Range
sine wave
square wave
Input Frequency1 to 50 MHz (nominal)
Lock range
DescriptionSpecificationsSupplemental Information
SyncReserved for future use
ConnectorBNC female
±2 × 10
reference input frequency
−6
of ideal external
−5 to +10 dBm (nominal)
0.2 to 1.5 V peak-to-peak (nominal)
(selectable to 1 Hz resolution)
90
PXA Signal Analyzer
Inputs/Outputs
DescriptionSpecificationsSupplemental Information
Trigger Inputs
(Trigger 1 In, Trigger 2 In)
ConnectorBNC female
Impedance10 kΩ (nominal)
Trigger Level Range−5 to +5 V1.5 V (TTL) factory preset
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.
93
PXA 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 Conformity):
To find a current Declaration of Conformity for a specific Keysight product, go
to: http://www.keysight.com/go/conformity
95
PXA Signal Analyzer
Regulatory Information
96
Keysight X-Series Signal Analyzer
N9030B
Specification Guide
2I/Q Analyzer
This chapter contains specifications for the I/Q Analyzer measurement
application (Basic Mode).
97
I/Q Analyzer
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 99 in this chapter.
Video BandwidthNot available.
Clipping-to-Noise Dynamic RangeSee “Clipping-to-Noise Dynamic Range” on page 100 in this
Resolution Bandwidth Switching UncertaintyNot specified because it is negligible.
Available DetectorsDoes not apply.
Spurious ResponsesThe “Spurious Responses” on page 59 of core specifications still
IF Amplitude FlatnessSee “IF Frequency Response” on page 38 of the core
IF Phase LinearitySee “IF Phase Linearity” on page 39 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 101 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.
98
I/Q Analyzer
Frequency
Frequency
DescriptionSpecificationsSupplemental Information
Frequency Span
Standard instrument10 Hz to10 MHz
Standard B2510 Hz to 25 MHz
Option B4010 Hz to 40 MHz
Option B8510 Hz to 85 MHz
Option B1X
Option B2X10 Hz to 255 MHz
Option B5X
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)
Standard B2510 Hz to 25 MHz
Option B4010 Hz to 40 MHz
Option B8510 Hz to 85 MHz
10 Hz to 160 MHz
10 Hz to 510 MHz
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)
Option B1X
Option B2X10 Hz to 255 MHz
Option B5X
10 Hz to 160 MHz
10 Hz to 510 MHz
99
I/Q Analyzer
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
b
dB
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 “DANL without Noise Floor Extension and
without Option EP0” on page 50.
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 “DANL without Noise Floor Extension and without Option
EP0” on page 50, gives a line entry on the excess noise added by using IF Gain = Low, and a footnote explain-
ing 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.
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.