The third edition of IEC
61000-4-3, published in 2006, has mandated changes in some key areas that may
require the replacement of your radiated immunity test amplifier. Along with
extending the test frequency range to
6 GHz and better defining the calibration of the uniform field, the standard
requires verification of the power amplifier linearity and output harmonics. As
a result, some existing systems may fail to meet the new requirements because
the amplifier in use is being run near saturation, and consequently, the
modulated field is distorted.
Amplifier Linearity
A linear relationship exists
between the input power and the output power over most of the operating range of
an amplifier. However, as the power level rises, limitations in the supply
voltage and current cause the output power to increase at a slower rate than the
input. This initiates amplifier saturation.
The relationship between input power and
output power starts to roll off and eventually reaches a point where increases
in the input power cause no change in the output power—the saturation point.
This point is the maximum power that can be supplied by the amplifier at a given
frequency (Figure 1).
Note that this saturated power level will vary with frequency.
|
Figure 1. Amplifier Gain Curve
|
Most amplifier suppliers use this saturated
power figure to define the performance of the amplifier; however, in the case of
IEC 61000-4-3 testing, this figure is not useful for selecting an amplifier or
predicting compliance with the standard. Since the test requires that the RF
test field is sine wave amplitude modulated, the input level is effectively
increasing and decreasing around the nominal test level.
If the amplifier is operating at any
point above the linear portion, the output-modulated signal will be distorted (Figure
2). For a true reproduction of the modulation
envelope, the amplifier must be operating in its linear region.
|
Figure 2. Amplifier Operated Beyond
Linear Range
|
Calibration
The IEC 61000-4-3 third
edition requires that the extra power needed for modulation be taken into
consideration. Previously, calibration was performed at the nominal test level
and modulation added later, so the user may not have been aware that saturation
was occurring. The latest version requires the user to show that the amplifier
will be operating linearly, at the nominal test level, including the peak of the
modulation envelope which is 1.8 times the nominal test level.
The new edition does not insist that the
output modulation waveform be completely undistorted but requires that the
amplifier be no more than 2 dB into compression at the peak of the modulation
envelope. To show this, the chamber must first be calibrated with the actual
equipment that will be used for testing and at a level 1.8 times the specified
test level; that is, for 10-V/m test, the calibration is run at 18 V/m. This
calibration produces a list of powers required at each frequency to achieve at
least 18 V/m at one location in the uniform field with another 11 in the range
of 0 to 6 dB above this level. Based on 16-point field uniformity, this requires
that 75% of the points must fall inside the 0 to 6-dB range.
Subsequently, at each frequency, the
calibration power level is generated by the amplifier, and then the
signal-generator level is reduced by 5.1 dB. The amplifier power level must fall
by at least 3.1 dB for the amplifier to meet the 2-dB linearity criteria and be
considered linear (Figure 3).
Figure 3. 2-dB Linearity Criteria
|
If this cannot be achieved, improvement to the
system is required, and it is necessary to operate the amplifier further down
its gain curve. This may be accomplished in a number of ways:
• Reducing the losses between the amplifier and antenna using shorter or lower loss cables.
• Using a higher gain antenna.
• Improving the performance of the test chamber.
• Using a higher power amplifier.
Test Distance
The original issue of IEC
61000-4-3 defined the test distance as 3 meters from the uniform plane to the
tip of a log periodic antenna or to the balun of a biconical antenna. This did
not allow for combination antennas such as bilog.
Many users opted to set the distance to the
phase center of a bilog antenna, which is approximately half way along the log
periodic section. This meant that the balun and the part of the antenna
responsible for transmitting the lower frequencies were approximately 3.5 meters
from the uniform plane.
The third edition of IEC 61000-4-3 defines the
test distance for combination type antennas as 3 meters to the tip of the
antenna. This can increase the distance to the lower frequency radiating points
of the antenna by up to 0.5 meter, which could add 1.16 dB to the power required
at the lower frequencies or 30% more power.
Harmonics
All amplifiers will produce
input signal harmonics. The level of these harmonics is dependent on the design
and quality of the amplifier and will worsen as the amplifier approaches
saturation.
However, the use of broadband
combination antennas leads to a potential problem. Since the gain of the
antennas typically increases rapidly between 80 MHz and 200 MHz, harmonics
produced by the amplifier in this frequency range have a disproportionate effect
in the field. See the sidebar Why Harmonics in the
Field Are a Problem.
Rather than define the harmonics from the
amplifier, the third edition of IEC 61000-4-3 requires the user to show that the
harmonics in the field are at least
6 dB down from the fundamental. As most test laboratories do not have
frequency-selective field measuring equipment, the standard allows the user to
measure the harmonics from the amplifier at each frequency and then calculate
the level in the field based on the performance of the antenna.
The harmonics from the amplifier must be
measured at the level required to create 1.8 times the target test level as
established during the field uniformity calibration. Again, it is important to
ensure that the amplifier is not operating near the saturation level.
Figure 4 shows the gain vs. frequency
curve of a typical combination type antenna. Comparing the gain at 100 MHz, 200
MHz, and 400 MHz, it increases by around 6 dB between 100 MHz and 200 MHz and by
around 7 dB between 100 MHz and 400 MHz. Harmonics would need to be at least 12
dB and 13 dB, respectively, below the carrier frequency of 100 MHz to achieve a
6-dB margin in the field.
|
Figure 4. Anetenna Gain vs. Frequency
|
Most amplifier manufacturers specify
harmonics of at least 20 dB down from the fundamental but only in the linear
region of the amplifier. Harmonics will increase once the amplifier starts to go
into saturation. See the sidebar Measuring Harmonic
Content in the Field.
Software
As with all aspects of RF EMC immunity testing, the amount
of data required to confirm the field uniformity, amplifier linearity, and
harmonic performance makes the use of automated test software essential. For
example, Teseq has recently released Version 4.00 of its Compliance 3 RF
Immunity test software that allows a test laboratory to perform the field
uniformity calibration at 1.8 times the nominal test level and then determines
and records the linearity and harmonics at the test level.
Conclusion
The third edition of IEC
61000-4-3 better describes the test conditions and clarifies some previous
anomalies. For instance, the test distance for combination antennas now is
better defined, and the quality of the test field in terms of modulation shape
and harmonic content is specified.
Although the purpose of the new standard was
to create more consistency and repeatability in testing, it may result in some
test facilities failing to meet the new requirements. Some improvements can be
achieved by using better antennas or cables. But in some cases, there will be no
alternative but to increase the power available from the amplifier.
About the Author
John Dearing is product
manager for the Teseq range of RF broadband power amplifiers and Compliance 3
test software. Educated to degree level in electronic and electrical
engineering, he went on to study RF and microwave electronics while working on
the design and development of a range of TWTs for the defense industry. Teseq,
Ashville Way, Molly Millars Lane, Wokingham, Berkshire RG41 2PL, U.K., +44 845
074 0660, e-mail: john.dearing@teseq.com
|
Why Harmonics in the Field
Are a Problem
Harmonics in the field are explained in detail in
Annex D of IEC 61000-4-3 and fall into three main areas:
1. Inaccuracy of Test-Level
Measurement
Virtually all field measuring
instruments commonly used in EMC test laboratories are broadband
devices; that is, they have no frequency selectivity and measure the
total field present at all frequencies. If there is a significant
harmonic content in the field, the measurement of the fundamental
field strength will be incorrect.
Since there is no way of predicting if the
various signals present will be added or subtracted due to phase
differences, there is no way to make any corrections for this error.
As a result, it is vital that the harmonic content of the field be
kept to a minimum.
2. False Failures in the EUT
When testing at a particular fundamental
frequency, it is possible that the EUT could have susceptibility not
at the fundamental but at the harmonic frequency. This would
generate false failures at the fundamental frequency that could
result in wasted time and money trying to design-out the problem.
3. Testing Intentional Receiver Devices
Commonly, when receiver
devices are tested, the test software is configured to skip the
intended operating frequencies of the receiver. However, if there
are significant harmonics in the test field, high levels of the
receiver’s operating frequency will be generated during the testing
of the lower frequency range that could adversely affect the EUT.
|
|
Measuring Harmonic Content In the Field
Test
Equipment
Most EMC test labs use only broadband
measuring equipment such as isotropic field probes and power meters
for immunity testing. To measure the harmonics, access to a
frequency-selective instrument such as a spectrum analyzer or
measuring receiver will be required.
The standard is not clear on how
many harmonics should be measured. But from a practical standpoint,
the lower-order harmonics will dominate, and broadband amplifiers
will not generate significant harmonics beyond their defined
operating frequency range.
For that reason, the laboratory
would be required to have access to an instrument capable of
measuring to at least the highest achievable test frequency. It is
only required occasionally to verify the harmonics and would not be
needed to run tests on a daily basis.
Antenna Calibration
In the past, it was not necessary to calibrate the antenna used for
immunity testing. As long as the antenna was efficient at the
required frequency, there was no need to know its actual performance
because the system was calibrated as a whole.
Since frequency-selective field measuring
instruments are not commonly available and expensive, the standard
defines a method of determining the harmonic field strength by
measuring the harmonic power from the amplifier and calculating the
field strength based on the performance of the antenna at that
frequency. As a result, the performance of the antenna must be
determined.
|