Text preview for : 5965-8166E.pdf part of HP 5965-8166E HP Publikacje 5965-8166E.pdf
Back to : 5965-8166E.pdf | Home
Agilent
10 Hints for Making Better
Network Analyzer Measurements
Application Note 1291-1B
Contents Introduction Scalar network analyzers
A scalar network analyzer (SNA)
HINT 1. Measuring high-power This application note contains a vari- measures only the amplitude portion
amplifiers ety of hints to help you understand of the S-parameters, resulting in
and improve your use of network measurements such as transmission
HINT 2. Compensating for analyzers, along with a quick gain and loss, return loss, and SWR.
time delay in cable summary of network analyzers and Once a passive or active component
measurements their capabilities. has been designed using the total
measurement capability of a VNA,
HINT 3. Improving reflection Overview of network analyzers an SNA may be a more cost-effective
measurements measurement tool for the production
Network analyzers characterize line to reveal out-of-specification
HINT 4. Using frequency-offset active and passive components, such components. While SNAs require an
for mixer, converter and as amplifiers, mixers, duplexers, external or internal sweeping signal
tuner measurements filters, couplers, and attenuators. source and signal separation
These components are used in sys- hardware, they only need simple
HINT 5. Increasing the accuracy tems as common and low-cost as amplitude-only detectors, rather
of noninsertible device pagers, or in systems as complex and than complex (and more expensive)
measurements expensive as communications or phase-coherent detectors.
radar systems. Components can have
HINT 6. Aliasing in phase or one port (input or output) or many
Network/spectrum analyzers
delay format A network/spectrum analyzer elimi-
ports. The ability to measure the
nates the circuit duplication in a
input characteristics of each port, as
HINT 7. Quick VNA calibration benchtest setup of a network and
well as the transfer characteristics
verification spectrum analyzer. Frequency cover-
from one port to another, gives
age ranges from 10 Hz to 1.8 GHz.
designers the knowledge to configure
HINT 8. Making your measure These combination instruments can
a component as part of a larger
ments real-time, accurate be economical alternatives in design
system.
and automated and test of active components like
amplifiers and mixers, where analy-
Types of network analyzers sis of signal performance is also
HINT 9. Making high dynamic
range measurements needed.
Vector network analyzers (VNAs)
are the most powerful kind of net- Incident Transmitted
HINT 10. Simplifying multiport work analyzer and can measure fre- DUT
measurements quencies from 5 Hz up to 110 GHz. SOURCE Reflected
Designers and final test in manufac-
turing use VNAs because they SIGNAL
SEPARATION
measure and display the complete
amplitude and phase characteristics INCIDENT
(R)
REFLECTED
(A)
TRANSMITTED
(B)
of an electrical network. These char-
acteristics include S-parameters, RECEIVER / DETECTOR
magnitude and phase, standing wave
ratios (SWR), insertion loss or gain, PROCESSOR / DISPLAY
attenuation, group delay, return loss, Figure 1. Network analyzer block diagram
reflection coefficient, and gain
compression.
VNA hardware consists of a sweep-
ing signal source (usually internal), a
test set to separate forward and
reverse test signals, and a multi-
channel, phase-coherent, highly sen-
sitive receiver (figure 1). In the RF
and microwave bands, typical mea-
sured parameters are referred to as
S-parameters, and are also com-
monly used in computer-aided
design models.
2
HINT 1.
How to boost and attenuate
signal levels when measuring
high-power amplifiers
Testing high-power amplifiers can The frequency-response effects of
sometimes be challenging since the the attenuators and couplers can be
signal levels needed for test may be removed or minimized by using the
beyond the stimulus/response range appropriate type of error-correction.
of the network analyzer. High-power One concern when calibrating with
amplifiers often require high input extra attenuation is that the input
levels to characterize them under levels to the receiver may be low
conditions similar to actual opera- during the calibration cycle. The
tion. Often these realistic operating power levels must be significantly
conditions also mean the output above the noise floor of the receiver
power of the amplifier exceeds the for accurate measurements. For this
compression or burn-out level of the reason, network analyzers that have
analyzer's receiver. narrowband, tuned-receivers are
typically used for high-power
When you need an input level higher applications since their noise floor
than the network analyzer's source is typically 90 dBm, and they
can provide, a preamplifier can be exhibit excellent receiver linearity
used to boost the power level prior over a wide range of power levels.
to the amplifier under test (AUT).
Some network analyzers with full
two-port S-parameter capability
enable measuring of the reverse
characteristics of the AUT to allow
full two-port error correction. In this
configuration, the preamplifier must
be added in the signal path before
the port 1 coupler (figure 3).
Otherwise, the preamplifier's reverse
isolation will prevent accurate mea-
surements from being made on port
1. If attenuation is added to the out-
Figure 2. High-powered foward measurement put port of the analyzer, it is best to
configuration use a higher power in the reverse
direction to reduce noise effects in
By using a coupler on the output of the measurement of S22 and S12.
the preamplifier, a portion of the Many VNAs allow uncoupling of the
boosted input signal can be used for test-port power to accommodate
the analyzer's reference channel different levels in the forward and
(figure 2). This configuration reverse directions.
removes the preamplifier's frequen-
Network analyzer
cy response and drift errors (by
ratioing), which yields an accurate
measurement of the AUT alone. RF amplifier Attenuator
Attenuator
"R"
When the output power of the AUT Channel in
exceeds the input compression level
of the analyzer's receiver, some type Coupler
Switch
of attenuation is needed to reduce Attenuator
the output level. This can be accom- DUT
plished by using couplers, attenua- Power
tors, or a combination of both. Care divider
must be taken to choose components Attenuator
that can absorb the high power from
the AUT without sustaining damage.
Attenuator
Most loads designed for small-signal
use can only handle up to about one Figure 3. High-powered forward and reverse measurement configuration
watt of power. Beyond that, special
loads that can dissipate more power 3
must be used.
HINT 2.
Compensate for time delay for
better cable measurements
A network analyzer sweeps its The lower trace of figure 5 shows an
source frequency and tuned receiver even more confusing result when
at the same time to make stimulus- measuring the same cable on an
response measurements. Since the 8753ES with 100-msec sweep time.
frequency of a signal coming from a Not only is there an error in the
device under test (DUT) may not be data, but the size of the error makes
exactly the same as the network ana- some sharp jumps at certain
lyzer frequency at a given instant of frequencies. These frequencies are
time, this can sometimes lead to con- the band-edge frequencies in the
fusing measurement results. If the 8753ES, and the trace jumps because
DUT is a long cable with time delay the network analyzer's sweep rate
T and the network analyzer sweep (df/dt) changes in different bands.
rate is df/dt, the signal frequency at This leads to a different frequency
the end of the cable (input to the shift through the cable, and hence, a
vector network analyzer's receiver) different amount of error in the
will lag behind the network analyzer data. In this case, instead of increas-
source frequency by the amount ing the sweep time, the situation can
F=T*df/dt. If this frequency shift is be corrected by removing the R-
appreciable compared to the net- channel jumper on the front panel of
work analyzer's IF detection band- the 8753ES and connecting a second
width (typically a few kHz), then the cable of about the same length as the
measured result will be in error by DUT cable. This balances the delays
the rolloff of the IF filter. in the reference and test paths, so
that the network analyzer's ratioed
Figure 4 shows this effect when transmission measurement does not
measuring the transmission have the frequency-shift error. The
response of a 12-foot cable on an upper trace of figure 5 shows a mea-
8714ET network analyzer. The upper surement of the DUT using the same
trace shows the true response of the 100-msec sweep time, but with the
cable, using a 1-second sweep time. matching cable in R channel.
The lower trace uses the default
sweep time of 129 msec, and the
data is in error by about