|
This
interactive Application Note provides additional
information on effectively utilizing the Summitek
Instruments line of Passive IM Analyzers. Various
typical measurement scenarios and application
information are presented. Some of this information
is based on Summitek's experience in IM measurement,
while some is based on how to best apply the unique
capabilities of these instruments to specialized
IM test applications.
Downloadable
Adobe PDF Application Notes:
| Download
the latest Acrobat viewer, it's free! |
 |
Jump
to a specific topic using these navigational links:
Cable
Assemblies
There
are two approaches to measuring the passive IM
response from a cable assembly. Figure 1
illustrates the connection diagrams associated
with each method.
Figure 1.
Typical Test Connections for
Measuring Cable Passive IM
The
most accurate measurements are made using the
connection shown in Figure 1 (a). With
this method, the cable is terminated into a low-IM
termination. This termination should be capable
of handling the combined carrier power used for
the test while generating a minimum level of IM.
The termination should present a good return loss
to the cable not only at the carrier and IM frequencies
of interest, but also at the low-order harmonics
of the carrier frequencies. Because IM is the
result of a combination of harmonics (such as
2*F1 and F2 for IM3), maintaining a near-unity
VSWR for each of the relevant harmonics enhances
measurement accuracy.
The
disadvantage to using the termination technique
shown in Figure 1 (a) is the possibility
of a frequency-dependent IM response. As the IM
sources contained within the cable assembly may
combine both in and out of phase (depending upon
the electrical length of the cable at the IM frequency),
a single-frequency test is insufficient to characterize
the cable's worst case performance across an entire
communications band. This requires the use of
the Swept Frequency Mode
to identify the IM peaks and nulls which might
occur in a reflected (reverse) IM measurement.
Figure 1 (b)
illustrates how the analyzer itself may be used
to terminate the cable assembly. Port 2 of the
analyzer presents a low-IM termination to the
cable. However, because this port is actually
a filter input, the impedance match at frequencies
other than the transmit band or receive band is
not typically well behaved. This can result in
a high VSWR value within the cable assembly at
frequencies that are harmonically related to the
carrier frequencies. This can increase the measurement
uncertainty of the cable IM measurement.
The
advantages to using the method illustrated in
Figure 1 (b) are two-fold. First, both
the reflected (reverse) and through (forward)
IM responses can be readily measured and compared
in a single connection. This can help diagnose
the location of an IM response on a long cable
assembly. Further, the through (forward) IM response
of a lossless cable assembly is theoretically
frequency independent. This is because each of
the IM sources adds in-phase (for a worst-case
measurement) at Port 2 of the analyzer.
In
summary, both techniques provide unique advantages
and disadvantages for assessing the IM performance
of a cable assembly. The one-port method (shown
in Figure 1 (a)) provides optimal measurement
accuracy, but requires an external termination
and the use of swept-frequency
mode. The two-port technique] (shown in Figure 1 (b))
provides measurement flexibility at the expense
of increased measurement uncertainty.
Back
to Top
Duplexers
Figure 2
illustrates a typical connection diagram for evaluating
the IM of a three-port duplexer. The transmit
port of the analyzer (Port 1) is connected
to the transmit port of the duplexer. The residual
IM from Port 1 of the analyzer and the interconnecting
cable to the transmit port of the duplexer should
be no greater than the expected IM level of the
duplexer plus the Rx isolation of the duplexer.
For example, if the duplexer has a 40 dB Rx-band
isolation between the Tx port and the Antenna
port, and the IM specification for the duplexer
is -120 dBm, a maximum of -80 dBm of
IM is allowable at the input port of the duplexer.
The
antenna port of the duplexer is terminated in
a low-IM termination that has a residual IM below
the expected IM measurement level.
Finally,
the Rx port of the duplexer is connected to Port 2
of the analyzer. A "Through" (or forward) IM measurement
mode is selected on the analyzer. By monitoring
the reflected (reverse) IM power, the integrity
of the Tx-port connection can be monitored to
ensure a high IM level is not being injected into
the duplexer.
Figure 2.
Typical Connection Diagram
for Evaluating Duplexer Passive IM
When
evaluating the measured IM of a duplexer connected
as shown in Figure 2, key information is
available for analysis by comparing the Reflected
and Through modes. If a high IM level is present
on both the antenna- and transmit-ports of the
duplexer, and these two levels differ by the known
S21 isolation between these two ports, a failure
is likely within the duplexer near the antenna
port connection. If a frequency-independent IM
level is measured with a low IM level present
on the antenna port, the failure is likely located
near the duplexer's transmit port.
Back
to Top
Transmit
Filters
Figure 3
illustrates a typical connection diagram for evaluating
the IM response from a transmit filter.
Figure 3.
Typical Connection Diagram
for Evaluating Transmit Filter Passive IM
By
comparing the Reflected and Through IM responses,
the IM of each side of the transmit filter can
be evaluated in a single connection. These two
responses may be significantly different as the
filter itself prevents the receive-band IM from
one port from leaking into the other port.
If
the IM specification for the transmit filter calls
for a precise power level to be present at the
port under test, it may be necessary to reverse
the filter to evaluate the response from each
port. This would be the case when the transmit
(passband) losses of the filter attenuated the
transmit carriers by an unacceptable level before
reaching the output port of the filter.
Back
to Top
Noise
Floor vs. Residual IM Floor
Determining
the amount of Averaging to use is key to maximizing
test speed while ensuring the true IM performance
of the DUT is accurately measured. Understanding
how the random noise floor of the analyzer and
the residual IM floor of the analyzer are related
is essential for setting the proper amount of
averaging.
The
noise floor of the analyzer is defined
as the mean value of the measured signal when
Port 1 and Port 2 of the analyzer are
terminated into 50 Ohms and the RF is turned
off. This noise is random and is typically due
to a combination of phase noise in the local oscillator
and kTBF noise from the receiver's pre-amplifiers.
The noise floor of the receiver varies from approximately
-127 dBm to -140 dBm depending upon
the selected averaging level. You cannot make
a meaningful IM measurement at a level below the
noise floor of the receiver.
The
residual IM level of the analyzer is caused
by internally generated IM within the analyzers
cabling, internal connectors, filters, and duplexers.
This level is typically larger than the noise
floor of the receiver for the third-order IM product.
When an IM measurement is performed on a DUT whose
true IM level is near that of the analyzer, significant
measurement errors can occur. This is because
the residual IM of the analyzer vectorially combines
with the true IM of the DUT producing a measurement
with a high uncertainty level.
Note:
Averaging cannot reduce the residual IM
level. Averaging is only useful in reducing the
level of the receiver noise floor. For the most
efficient measurement time, use only enough averaging
to maintain the receiver noise floor at least
10 dB below the expect minimum PIM level.
Figure 5
shows the classical measurement uncertainty curve.
The x-axis shows the magnitude of the error source
relative to the true magnitude of the response
being measured. The y-axis shows the maximum measurement
uncertainty in the measurement due to this single
error source assuming worst-case coherent addition
and subtraction between the error source and the
true response.
Figure 5.
Standard Voltage Error Curve Showing the
Maximum Measurement Uncertainty Due to a Single
Error Source.
Some
useful data points from this curve are presented
in Table 1.
Table 1.
Key Measurement Uncertainty Values for
10, 20, and 30 dB-Down Error Sources.
|
Error
Relative
to True Value, dB
|
(+)
Uncertainty,
dB
|
(-)
Uncertainty,
dB
|
|
-10.0
|
2.39
|
-3.30
|
|
-20.0
|
0.83
|
-0.92
|
|
-30.0
|
0.27
|
-0.28
|
Note:
When using the Passive IM Analyzer, the fluctuations
in the measured value due to the receiver noise
floor are not as great as the values in Figure 5
indicate. This is because the internal receiver
in the analyzer utilizes significant amounts of
video filtering to effectively reduce the magnitude
of random noise errors to their mean value.
Back
to Top
Measuring
Residual IM
Measuring
the residual IM of the test setup prior to testing
the DUT is highly recommended. This measurement
establishes the lower limit below which the DUT's
IM characteristics cannot be accurately measured.
For the typical passive IM measurement, it is
desired that the residual IM level be at least
5 to 10 dB below the IM level of the DUT.
From Table 1, a residual IM level 10 dB
below the DUT IM level allows the measurement
of the DUT's IM with an approximate +2.4 and -3.3 dB
uncertainty.
To
measure the residual IM of the test setup, the
high power transmitter must be terminated into
a load. This load must not generate significant
levels of IM. For example, the SI-20B Low-IM Termination
produces an IM level not exceeding -115 dBm
with 2 x 20 W carriers. Placing
this load at the end of a cable (which would normally
connect to the DUT) allows the user to certify
the residual IM of the test setup does not exceed
-115 dBm.
When
using a connecting device (such as a cable or
adapter) between the analyzer and the DUT, it
is good practice to ensure that the residual level
is stable under all conditions that might be encountered
during the DUT measurement. For example, if stress
will be placed on the cable or adapter during
the test, place a similar stress on the cable
or adapter when terminated into the load to ensure
the residual level remains at or below the desired
limit under all test conditions.
Back
to Top
Antenna
PIM Plots
Click
on the links below to see sample plots.
Back
to Top
|
