Tektronix RSA5100B Wide Band Acquisition Bandwidth User Guide
- June 7, 2024
- Tektronix
Table of Contents
Tektronix RSA5100B Wide Band Acquisition Bandwidth
TECHNICAL BRIEF
Acquisition Bandwidth| Option B16x 165 MHz Acquisition Bandwidth|
B16xHD, 165 MHz High Dynamic Range Acquisition Bandwidth
---|---|---
40 MHz
< 80 MHz
| -76 dBc| -80 dBc
80 MHz
< 165 MHz
| – 73 dBc| -80 dBc
Introduction: Tektronix RSA51xxB wide band acquisition bandwidth options Tektronix offers a choice of dynamic range in the high acquisition bandwidth options for the RSA5100B series real time signal analyzers. There are advantages and disadvantages for each solution, and this white paper examines the difference between these options, and how to choose between them. There are options from 85 MHz to 165 MHz bandwidth, and this paper focuses on results from the 165 MHz options to present the most difficult case. Both options provide up to 165 MHz of real time analysis bandwidth over the frequency range of the RSA model. The differences are in the SFDR (Spurious Free Dynamic Range) with signal, other than CF (Center Frequency). All comparisons in this paper are of the published typical performance. See the data sheet for details; they are simplified here to cover all frequency ranges of the instruments. For spans less than 80MHz and for instruments up to 6.2 GHz there is some improvement in these specs The B16x option is competitive with other manufactures providing up to 165MHz of real time bandwidth. This option has great performance for analysis of 802.11ac Wi Fi signals and for wideband real time spectrum analysis with DPX spectral displays and triggering. If the option B16x specifications meet your objectives then this option will work well for you.
If you desire a bit more SFDR and a cleaner spectrum, you may want to consider
the B16xHD option. The B16xHD option provides a cleaner and lower SFDR than
the B16x option. By cleaner, we mean that the option B16xHD option has fewer
distortion products of any type, and overall lower SFDR allows the user to
look deeper and see signals without the instrument specifications limiting the
measurement. Better SFDR also provides more headroom when measuring small
signals in the presence of large signals without the need to consider
distortion products created in the measurement tool. In this paper, HD2 and
HD3 refer to HD – Harmonic Distortion, 2 – second harmonic and 3 – third
harmonic. These are generated in the intermediate frequency (IF) stage and the
analog to digital converter (ADC). When harmonic distortions are generated in
the IF/ADC, their location in the acquisition band can be difficult to
predict. At any particular operating frequency, the location of the signal
will depend upon the Nyquist zone selected for operation, whether high-side or
low side mixing products are used and the sampling rate. Changing the input
frequency can cause the signals to move in opposite directions, at different
rates. The example shown in this paper is repeatable, but you will get
different interactions at other input frequencies. This IF distortion should
not be confused with the instrument specification for front-end second and
third harmonics. RF-input harmonic distortion is predictable based on the
input frequency. For example, if the input is 1GHz, HD2 of the front end is 2
GHz and HD3 is at 3 GHz.
Figure 2. Performance of Option
B16x showing IF/ADC-generated harmonic distortion at -75.72 dBc. The signal at
M1 is the combination of HD2 and HD3 combine.
B16x Option
Figure 2 shows a signal at the reference level of -10dBm and the spectrum that
you would typically see for a center frequency (CF) of 400 MHz with a 380 MHz
tone applied to the input. The span is 165 MHz. There are usually two or three
signals created by the instrument that limit SFDR. The main component shown
above by marker M1 at -75.72 dBc is the point where the HD2 and HD3 of the
input signal in the IF are on top of each other. Remember, these HD2 and HD3
terms are generated only in the IF and are not part of the RF harmonic
distortion specification. The above illustration has no averaging so the
signal at M1 moves around many dB with noise. This screen shot has captured a
high peak since HD2 and HD3 “beat” together and cause fluctuations in the
level of the spur.
Figure 3. Performance of Option
B16x showing IF/ADC-generated harmonic distortion at -78.62 dBc, after 10
averages to reduce noise variations on the signal.
Figure 3 is the same conditions as in Figure 2, except with 10 rms averages of
the traces. The result is a less noisy measurement of about -79 dBc at
M1.
Figure 4. Tuning the input signal slightly causes the HD2 and HD3 products
to separate in frequency, seen at M1.
Moving the signal even a few hundred kHz to 380.3 MHz away splits these
signals into their two respective harmonic distortion terms. HD3 is shown at
marker M1 at -81 dBc. HD2 is the smaller term just to the right of M1. M2 is
the other smaller spur in the system. These signals limit the SFDR to -73 dBc
typical. Moving the input signal across the band produces spurs that move 1:1,
2:1, 3:1 and 4:1 across the band until they fall off the end of the screen.
Increasing the frequency further will bring these terms in band at roughly
these same levels as it aliases back into the band and moves in the opposite
direction across the displayed span.
Figure 5. When the input signal is at the low end of the acquisition
bandwidth (320 MHz), the IF/ADC harmonics are at M1 and M2. The second
harmonic is dominant at about -75 dBc, with the third harmonic at -86 dBc.
For signals at the low side of the screen, a 320 MHz signal produces the
signals seen in Figure 5. M1 is HD2 at about -75 dBc in the IF, M2 is HD3 at
about -86 dBc.
Figure 6. An example of when the
third harmonic has ‘folded’ back into the analysis bandwidth. Here, a signal
slightly lower than the center frequency of 400 MHz creates IF/ADC 2nd
harmonic at M1 and the 3rd harmonic at M2.
With the input signal near the center of the screen at 390 MHz (CF=400 MHz) we
see HD2 in the IF at M1 near the right hand side of the screen at -81 dBc and
HD3 on M2 at about -80 dBc (Figure 6).
Figure 7. With the input
signal slightly higher than the center frequency, the second harmonic appears
at M1, and the third harmonic has now moved below the input frequency at M2.
As the input frequency increases to 410MHz we see HD2 at
M1 of -83 dBc and other spurs M2 for HD3 at -80 dBc.
Figure 8. Moving the input signal to the upper edge of the acquisition
bandwidth moves the 2nd harmonic to below the input frequency, and the 3rd
harmonic to between the input and the second harmonic. The second and third
harmonics travel in different directions and at different rates, making them
difficult to identify on screen.
For signals at the right hand side of the screen at 480 MHz we see HD2 at M1
of -80 dBc and other spurs HD3 at M2 at -81 dBc. These spurs are relatively
stable with a small amount of noise if averaging is used. The performance of
option B16x is similar to that of other real time spectrum analyzers available
today. All available products use similar 14 bit digitizers applied to an
intermediate frequency. While the exact input frequency required to see the
harmonic distortion may vary, the levels of harmonics in the IF is very
similar for all instruments.
Figure 9. Option B16xHD,
high dynamic range option, shows a 10 dB improvement over the standard
acquisition system for IF/ADC harmonic distortion. Here, the 2nd and 3rd
harmonics are at the same frequency at M1, but are nearly in the noise level.
The new dominant spur, shown at M2 at -86.06 dBc is caused by the interleaving
distortion of the two 16-bit digitizers used in the high dynamic range option.
B16xHD Option
Option B16xHD improves the SFDR to -80 dBc typical, a significant improvement
over -73 dBc typical for the B16x option. The nature of the limiting
components for this option have changed though. This design uses an improved
digitizer with very low HD2 and HD3 which can be seen on the right hand side
of the screen as indicated at M1 in Figure 9. This is comprised also of the
two HD2 and HD3 terms on top of each other. The dominant spur is shown on
marker M2 which in this screen shot is -86 dBc. This image term moves up or
down with no averaging. This screen shot illustrates a case where this spur is
higher than normal but it is still at -86 dBc. The system used is an
interleaved ADC that produces interleave images that show up equidistant about
the center of the screen at 1:1 and on the opposite side as shown at M2 for a
single tone input. If the input is brought in on the left side of the screen,
the spur will appear on the opposite side equidistant from the center of the
screen to the input. As the signal moves toward the center of the screen, the
spur on the right side moves toward the center of the screen. They cross over
at the center and continue to move toward the edges of the screen. The
spurious frequency is quite easy to predict so if they fall on a suspected DUT
signal one can move the CF a bit and determine if the components are from the
DUT or from the instrument. The majority of the time this option provides
better than -80dB of SFDR. Due to the nature of the vector cancellation in the
correction process and the noise on top of those signals, the peaks of this
spur can range up to an 8 to10dB peak to peak value. There may be times that
the peaks can reach to -78 dBc, but this is statistically likely to occur only
at a 3 Sigma probability. This term is temperature sensitive so a fresh
alignment should be done prior to a very sensitive measurement. The instrument
will align itself for this spur for smaller temperature variations than the
B16x option which is very stable but has worse SFDR.
Figure 10. Option 16xHD showing the where the image spurious at M2 is nearly in the noise. This signal varies with time and temperature, but is lower than -80 dBc in nearly all cases.
Figure 10 illustrates a case with no averages where the M2 spur is very close
to the noise floor. HD2 and HD3 are nearly non-existent. This case catches a
very low M2 spur. In normal operation with no averages this signal will vary
in a noise-like way between mid -80 dBc to -90 dBc. The signal on the far left
edge of the screen is a non-dominant system spur that is well below the -80
dBc specification Figures 11-16 show the performance of the high dynamic range
options as the carrier frequency is tuned throughout the input signal range of
the instrument. The test conditions are similar to those shown when using the
standard dynamic range options, but the resultant harmonic levels and spurious
are substantially improved. See the caption headings for test conditions and
results.
Figure 11. The image spurious
after 10 averages, now at -91.5 dBc.
Figure 12. With an
input of 380.3MHz where the HD2 and HD3 would be at M1, but are nearly non-
existent due to the high dynamic range of Option B16xHD. M2 indicates the 1:1
image spur symmetrical about the center of the screen and lower than -90
dBc.
Figure 13. Option 16xHD with an input of 380.3MHz where the HD2 and HD3
would be split at M1 but they are still nearly non-existent. M2 indicates the
1:1 image spur symmetrical about the center of the screen and lower than -90
dBc.
Figure 14. With an input
at 390 MHz, M1 is the HD2 that is in the noise and M2 indicates the 1:1 image
spur at -96 dBc.
Figure 15. The input is
at 410 MHz. M1 is HD2 at -89 dBc and M2 is the 1:1 image at -90 dBc.
Figure 16. HD2 at -92 dBc (M1), with the 1:1 image at -87 dBc (M2).
Figure 17. Probability of image level measured for 4 instruments. The 95% probability is at approximately -80 dBc, with a mean (50%) level of -82 to -83 dBc.
Other information about this B16xHD option
The image is temperature sensitive. To counteract this, the B16xHD option uses
a thermally-stabilized element that maintains temperature within 2 degrees C
over nearly the entire operating range of the instrument. In the case of large
(> 5 degrees C) ambient temperature changes, the instrument may require
alignment to optimize performance of the ADC system. This limited alignment
takes only a few seconds, and can be turned off if operating conditions
require no interruptions. If this is not desirable, a better choice might be
the B16x option that has stable but worse spurs if the application doesn’t
demand the better SFDR of the B16xHD option. The image level also varies with
noise that occurs during the normalization and normal operation. The variation
is consistent with statistics where a greater than 3 Sigma variation may be
had very occasionally. During our performance evaluation testing on multiple
units, we measured the spurious signals for each 500 kHz step across the 165
MHz acquisition bandwidth, a total of 330 measurements per test. The results
of this testing were that 99% of the time, the worst case capture over many
sweeps will remain below -79 dBc. 95% of time, the worst case capture will be
under -80 dBc on our units. At the mean or 50%of the time the worst case will
be below -82 dBc. Under most conditions you will not see a worst case unless
you are at the exact frequency and time that a >3 sigma variation may happen.
The Typical-Mean will be between -80 dBc and -90 dBc except for >3 sigma
points of the noise. Figure 19 shows the image rejection probability of four
of the instruments tested during our performance evaluation.
Preselection and RF/IF Images
This paper has focused on the distortion and spurious products present in the IF/AD system of wideband analyzers, including digitizer-related images. An important contributor to overall performance is the suppression of images that result in the RF and IF down conversion process. Tektronix real time spectrum analyzers are always preselected to avoid images. Most spectrum analyzers use narrowband tunable band pass filters, often YIG tuned filters (YTF) to serve as a preselector. These filters provide image rejection and improve spurious performance in swept applications by limiting the number of signals present at the first mixing stage. YTF’s are narrow band devices by nature and are usually limited to bandwidths less than 50 MHz. These analyzers bypass the input filter when performing wideband analysis, leaving them susceptible to image responses when operating in modes where wideband analysis is required such as for real time signal analysis. Unlike spectrum analyzers with YTF’s, Tektronix Real Time Signal Analyzers use a wideband image-free architecture guaranteeing that signals at frequencies outside of the band to which the instrument is tuned don’t create spurious or image responses. This image-free response is achieved with a series of input filters designed such that all image responses are suppressed. The input filters are overlapped by greater than the widest acquisition bandwidth, ensuring that full-bandwidth acquisitions are always available. This series of filters serves the purpose of the preselector used by other spectrum analyzers, but has the benefit of always being on while still providing the image-free response in all instrument bandwidth settings and at all frequencies.
Conclusions
CHOOSE B16X OPTION WHEN:
- -73 dBc typical meets your requirements
- Expected temperature environments may vary quickly
- It is necessary that spurious be relatively constant and stable, and the standard spurious performance meets your requirements.
CHOOSE B16XHD WHEN:
- Operation in a normal lab temperature environment is expected
- There is the need for another 10 dB of SFDR for -80 dBc typical-mean spec
- Some noise variation in SFDR in the worst case is acceptable but most of the time below -80 dBc.
Appendix: Definitions of Specification Types
Specifications vary widely from manufacturer to manufacture. Tektronix
supplies several categories of published performance for spectrum analyzers,
outlined below. The specifications discussed in this article are P-Typ, or
published typical performance.
Nominal: Characteristics that are described in terms of nominal
performance—guaranteed, but without tolerance limits. Examples are number of
trigger inputs, or physical dimensions and weight.
Specified Characteristics: Characteristics described in terms of specified performance with guaranteed tolerance limits, such as reference frequency accuracy and frequency response. Specified characteristics always include a description containing tolerance limits. Specifications are warranted to the customer, and are tested either directly or indirectly in manufacturing or by type-testing of instruments prior to manufacturing. Specifications that are checked directly are included in the Performance Verification document for the instrument.
Typical Characteristics: Characteristics described in terms of typical, but not guaranteed, performance. The values given are not warranted, but 80% of measured values will meet the specification with 80% confidence, unless otherwise noted, between 18C and 28C ambient, immediately after performing a full alignment. Typical characteristics are not tested in manual performance checks.
Typical-95: These are specifications that are impractical to verify in
the field as they are based on statistical performance data across a large
sample of instruments. Examples of this are Adjacent Channel Leakage Ratio
(ACLR) dynamic range, and Absolute Amplitude Accuracy at all Center
Frequencies. A specification classified as (P-TYP-S95) denotes that the
specification will be met on any particular instrument with 95% confidence.
95th percentile values indicate the breadth of the
population (≈2σ) of performance tolerances are expected to be met in 95% of
the cases with a 95% confidence, for any ambient temperature in the range of
18 to 28°C, immediately after performing an alignment. In addition to the
statistical observations of a sample of instruments, these values include the
effects of the uncertainties of external calibration references and aging over
the course of the published calibration interval. These values are not
warranted. These values are updated occasionally if a significant change in
the statistically observed behavior of production instruments is observed.
Find more valuable resources atTEK.COM
Copyright © Tektronix. All rights reserved. Tektronix products are covered by
U.S. and foreign patents, issued and pending. Information in this publication
supersedes that in all previously published material. Specification and price
change privileges reserved. TEKTRONIX and TEK are registered trademarks of
Tektronix, Inc. All other trade names referenced are the service marks,
trademarks or registered trademarks of their respective companies. 07/2021 SMD
37W-60135-1
