In the rush to complete RF immunity testing on schedule, it is not all that unusual to overlook inherent test equipment limitations. While some test equipment characteristics such as power amplifier harmonics are obviously a limiting factor, the broadband characteristics of antennas, directional couplers, power meters and isotropic field probes can hardly be considered a limitation for most applications. However, when used with power amplifiers exhibiting significant harmonic distortion in Immunity test systems, the broadband characteristics of these devices can result in measurement uncertainty and unacceptable errors.
A case in point is the ubiquitous broadband isotropic field probe that provides an E-field reading representative of the total energy from all frequencies within its operating band. Given the ideal, albeit rare, case of a pure sinusoidal signal, field probes provide an extremely accurate reading. To the extent that additional frequencies are present, errors are introduced and depending on the number and strength of the additional signals, a point is reached where the field reading is totally unrepresentative of the required test level at the desired frequency. The most troublesome unwanted frequencies are harmonics generated by RF test system nonlinearities. Often power amplifiers, especially those driven into saturation, are a major source of harmonics. To a lesser extent, signal sources, directional couplers and antennas exhibit some degree of nonlinearity and also contribute to the system level harmonics. Accordingly, the IEC 61000-4-3 has instituted system requirements intended to limit the allowable harmonic levels in the test field.
While it is imperative to consider instrument harmonic levels supplied by instrument vendors, test engineers must also confirm manufacturer’s data by testing. While this article specifically addresses ways to check for harmonic levels mandated by IEC 61000-4-3, the procedures can be readily applied when testing to other RF immunity standards.
Harmonics are unwanted frequencies generated by system nonlinearities. They are multiples of the fundamental test frequency, and generally, the higher the multiple, the less the amplitude of the harmonic. All “real” test systems have a finite amount of nonlinearities, and thus, exhibit some degree of harmonic distortion. The test engineer must ultimately determine acceptable levels of harmonics. His determination is primarily based on test standard mandates. In EMC testing applications, RF power amplifiers are responsible for most of the unwanted harmonics.
Understanding Harmonics in an Amplifier
All amplifiers exhibit harmonic distortion to some extent. While some applications like industrial RF heating and plasma generation are not affected by harmonics, high levels of signal distortion will introduce unacceptable errors when testing for EMC immunity. Accordingly, harmonic distortion is a key power amplifier specification. It has been proven that properly designed Class A amplifiers when operated in their linear region have acceptable levels of harmonics and are an ideal choice for EMC test applications.
Keep in mind that even a properly designed, robust Class A RF power amplifier does not guarantee a distortion free test field. Care must be taken to operate within the linear range of the amplifier, even at the sacrifice of a smaller output signal. While driving the amplifier harder will indeed provide greater field strength, the inherent signal distortion resulting from a spike in the harmonic levels will introduce uncertainty and error in the resultant E-field. Ultimately, the question becomes, “Just how much input signal is required to ensure the desired signal purity in any given application.” It can be seen that an EMC amplifier should not be operated beyond the 1dB compression point. In fact, operating in a more linear region below the 1dB will drastically minimize harmonics. Another less desirable option is the use of harmonic filtering at the output of the amplifier. Since this approach adds cost, insertion loss and complexity to the system, it should only be considered when there is no other practical option. For example, some TWT amplifiers are best served by the use of harmonic filters.
Since it is all but impossible to predict the cumulative effect of all the system devices on the purity of the E-field, a system level measurement must be taken. While vendor data should be consulted and relied on when selecting a power amplifier, there is no substitute for actual system measurements when it comes to validating the viability of a system design.
How do multiple signals influence power measurement?
Most field probes and power heads use diode sensors with broadband characteristics. These devices are not frequency selective and will measure all signals within their operating range. The resultant reading is the square root of the sum of the squared amplitude of the fundament and all harmonics present. Clearly, harmonics will add proposition. Thus, the conundrum is determining what would be an acceptable level. Fortunately IEC 61000-4-3 provides guidance in this area. The latest version of IEC 61000-4-3 states the following: “For all frequencies where harmonics are produced at the output of the amplifier, the rejection of these harmonics in the field by more than 6 dB below the fundamental is adequate.” In other words, there is now a 6dBc harmonic requirement in the test field. Note that dBc is a measurement of a specific harmonic level in relation to the carrier. A measurement of -6dBc by definition means that the amplitude of the harmonic is 6dB less than the carrier amplitude. Past IEC 61000-4-3 standards have specified the output harmonic level from the power amplifiers. The latest version of the standard considers the entire system when it mandates a 6dBc requirement. This level takes into account the fact that the transmitting antenna operates more efficiently at the 3rd harmonic than at the fundamental. It is not uncommon to see as much as a 5dB gain variation. As discussed in IEC 61000-4-3 annex D, limiting all harmonics in the test field to -6dBc will result is no more than a 10% field strength error. Figure 1 graphically plots this relationship. Note that with a -6dBc harmonic level a field probe reading of 10V/m actually represents about a 9V/m carrier level. If the test calls for more accuracy, the harmonics must be further reduced. For example, a 5% error in field strength requires the harmonic to be at least -10dBc. Standards that do not take into consideration the effect of the transmitting antenna concentrate on the power amplifier harmonics. For example, older versions of IEC 61000-4-3 limited amplifier harmonic levels to -15dBc. When compared to the new -6dBc total field specification, the -15dBc results in slightly less field level error.
Figure 1: Single Harmonic Contribution to Measured Field
There are two generally accepted methods used to determine the harmonic content of a test field. In both cases a frequency selective device is required to measure the level of the fundamental frequency as well as the harmonics. The most popular instrument used for this purpose is a spectrum analyzer. The required frequency range of the spectrum analyzer is determined by the frequency range mandated in the EMC standard. For example, since IEC 61000-4-3 covers 80MHz to 6GHz, the spectrum analyzer should have a minimum bandwidth of 80MHz to 18GHz in order to respond to at least the 3rd harmonic. For the rare occasion where there is significant harmonic content beyond the 3rd harmonic, a higher frequency analyzer is required. In most cases harmonic levels are inversely proportional to frequency and are not a factor outside the operating band of the amplifier. Since there are some exceptions to this general rule, it is prudent to always verify harmonic levels by testing. One needs to look no further than to some TWT amplifiers which exhibit significant harmonics well beyond the frequency band of the amplifier. The message here is to be keenly aware of the predicted harmonic levels as published by the amplifier manufacturer, but always test to verify the published data.
The test setup used for this method replicates that used for the actual test. Since the harmonics are measured directly without the need for calculations, it is the preferred method providing the most accurate data.
As noted above, the spectrum analyzer used is primarily determined by the test frequency range of the EMC test standard. The IEC 61000-4-3 covers 80MHz to 6GHz. To measure out to the 3rd harmonic, the spectrum analyzer must cover 80MHz to 18GHz. An ideal solution for the receive antenna would be one that covered the entire frequency range of 80MHz to 18GHz. Since typically this is not possible, the next best approach is to break the overall band up to coincide with the band breaks of the transmit antennas.
Recommended frequency assignments for both transmit and broadband biconical receive antennas are shown in Figure 2. This is an ideal solution since each receive antenna covers the harmonics from each transmitting antenna. Since there is no need to switch in additional antennas, this is a rather simple solution. While not as elegant as a single receive antenna, it is the next best thing and quite amenable to control via software.
In the event that a single receive antenna were not available to respond to the 3rd harmonic of each transmitting antenna, one could opt for a less desirable, overlapping approach as shown in Figure 3. This setup is commercially available by combining a Biconical Log-Periodic with a double-ridge antenna. It can be seen that the lower frequency transmit antenna requires both the receive antennas to adequately cover all the harmonics. This is a much more difficult setup to implement either manually or via software control.
Figure 4: Basic Setup Diagram for Receive Antenna
The Directional coupler method can also be used to measure system level harmonics. This approach is more complex than the receive antenna method and given the following inherent uncertainties, it is the least desirable choice.
Based on an assumption that harmonics should fall off at the top end of the amplifier band and not reappear at points outside the band of the amplifier, one can limit the extent of measurements taken. However, tests should be run to backup any assumptions made.
In addition to the considerations noted with the receive antenna method covered above, additional directional couplers must be compatible with the power amplifier in terms of power handling capability as well as frequency range.
Figure 5: Basic Setup Diagram for Directional Coupler
Note: Care should be taken that if an additional directional coupler is used it does not add significant losses to the test system. Pat Malloy has been the sales application engineer at Amplifier Research, now AR, since 1987. Previous work experience includes four years with the U.S. Navy as a guided-missile electronic technician, seven years in an engineering group at AT&T Bell Laboratories, and 16 years as a senior sales engineer for Tektronix. He graduated from Lafayette College in 1972 with a B.S.E.E. Mr. Malloy can be reached at pmalloy@ar-worldwide.com.
Jason Smith has been the applications engineering manager at AR since 2004. Previous work experience includes test engineer and EMC lab manager at Radiation Systems and EMC lab manager at Analalb, LLC. Jason has over 10 years experience in EMC testing experience with military, avionics, commercial, medical, telecom and automotive applications. He is a member of the USNC to SC77B and SC77C and a participating member of WG10 (IEC 61000-4-3, -6). He graduated from the University of Delaware in 1997 with a B.S. in Engineering Technology. Mr Smith can be reached at jsmith@ar-worldwide.com.
complianceharmonicsiec 61000-4-3immunityradiated immunitystandardstest equipment
In Compliance is a leading source of news, information, education, and inspiration for electrical and electronics engineering professionals.
Aerospace Automotive Communications Consumer Electronics Education Energy & Power Industrial Information Technology Medical Military & Defense