Proof of compatibility indicates that your product has survived certification testing. A major player in all this testing is system hardware. The hardware relies heavily on detectors to gather signal measurement data for a variety of interference types.
Types of Interference
The RF interference measured by an EMC test station can be divided into three generic categories:
- Discontinuous Noise—as created by a commutator motor. When viewed on an oscilloscope, these signals have the appearance of a fundamental frequency and random noise. They cause a broadband spectrum with erratic characteristics.
- Pulsed Interference—creating a transient every time a device comes on or goes off. For example, a light dimmer will produce transient pulses at a 120-Hz rate corresponding to every half cycle on a power line. This produces a broadband spectrum that generally is flat.
- Continuous Noise—as radiated by an electronic device with an oscillator, clock, or switched-mode power supply. This produces a peak with numerous harmonics in the frequency spectrum.
The goal in EMC testing is to find a measurement technique that copes with each of these signals. You want a meaningful display and a report indicating the level of unacceptable interference caused by the unwanted signals.
If you are testing for compliance to one or more regulations, the applicable government documents provide test guidance and acceptable limits. Commercial products to be sold in the United States are controlled by the Federal Communications Commission (FCC) as defined in Part 15, Subpart B. Class A applies to equipment for business use, and Class B relates to products for homes. The European Community (EC) and many other parts of the world use the EN 50xxx, EN 55xxx, and EN 60xxx standards that are established and maintained by the European Committee for Standardization (CEN) and related European agencies.
Test-Station Architecture
Several differences exist between the test stations used for conducted interference testing and those measuring radiated signals. “From a functional standpoint, however, the systems operate in the same manner,” according to Greg Senko, business manager for test equipment at Schaffner EMC. “The type of transducer makes the difference. For radiated tests, an antenna is required. For conducted tests, a coupling network, an injection probe, or a clamp is used.”
Radiated Interference
Radiated interference testing has two facets: either a product radiates objectionable signals or it is susceptible to radiated interference. Radiated emissions and immunity tests accommodate both circumstances. In the European Community (EC), generic standard EN 50081 covers emissions, and EN 50082 relates to immunity.
Emissions
For pickup of interfering signals, you need a sensitive receiving antenna with a radio receiver or spectrum analyzer to amplify, process, and display your measurement results. The active region or focal point of the receiving device should be 3 m or 10 m from the DUT.
Immunity
For immunity, perhaps the more difficult test to perform, an RF field of 3 V/m or 10 V/m is generated and swept across the spectrum of concern. The DUT is placed at the prescribed distance from the radiating element, typically 1 m or 3 m, in a shielded chamber or TEM cell.
Uniformity of signal strength surrounding the DUT is specified in two ways. In the EN 61000-4-3 radiated susceptibility method, most useful at frequencies above 80 MHz, 16 field strengths are measured at equally spaced points in a 1.5 m × 1.5 m area. If 12 or more are within 6 dB of each other, the field is considered uniform. The IEC 801-3 approach, typically applied at frequencies between 26 MHz and 80 MHz, uses a closed loop to adjust the generator output level.
Often the radiator for immunity tests is an E-field generator rather than a conventional antenna. This enables you to get a concentration of power with a small element.
Conducted Interference
Even with excellent shielding and the use of gaskets at all the seams, your product still has input and output wires or cables that can interfere with other equipment or be affected by conducted interference. This is where you need the ability to perform conducted emissions and immunity tests.
Emissions
For measurement of conducted emissions on power connections, the measurement system uses a line-impedance stabilization network (LISN), also called an asymmetrical artificial network (AAN), to provide a 50-W RF connection. This network passes the differential-mode signal of interest almost unattenuated while decoupling the test circuit from external effects. This virtually eliminates any measurement uncertainty.
Immunity
For measuring immunity to interference, a coupling/decoupling network (CDN) is used in the test system to inject RF from a generator and an amplifier onto input and output cables. The coupler is chosen to cover the range of interest, typically 150 kHz to 80 MHz.
Detector Choices
The detector in an EMC test system accepts the down-converted carrier from the receiver’s intermediate-frequency (IF) amplifier and demodulates it to present a meaningful baseband signal. Typical IF bandwidths are 200 Hz, 10 kHz, and 120 kHz. Averaging, quasipeak (QP), and peak detector types are available, and each has a specific assignment.
- An averaging detector, taking its signal from the receiver through a simple filter, is optimum for discontinuous noise. This is what CISPR-16 specifies for such signals.
The output is a good representation of unacceptable interference. If you use an instantaneous detector on such a signal, you will get random and meaningless fluctuations. - The QP detector generally is used for pulsed interference. In this circuit, the input passes through a leaky peak-hold circuit where charge and discharge time constants are set to match CISPR specifications. The output from this leaky detector is taken to a critically damped electronic averaging circuit.
The output of the QP detector is a good representation of the composite of a pulsed signal’s peak amplitude, duration, and repetition rate. This is the only detector type defined by the FCC, but CISPR calls out both QP and averaging limits.
The averaging detector can’t be used for measuring pulsed interference. When pulses have a low duty cycle, the detector will give a very low output reading even if the interference is high. The peak detector is not appropriate for these signals because it produces an output corresponding to the magnitude of the pulses, regardless of their repetition rate. - A peak detector is good for measuring continuous noise. The result is a true representation of the magnitude of the interference, and it develops its output instantaneously.
In contrast, the averaging and QP detectors are quite slow. They must dwell at each frequency increment for a relatively long period before producing results.
“If the interfering signal is continuous or has a repetition rate above the IF filter bandwidth,” according to David Mawdsley, managing director at Laplace Instruments, “all detectors produce the same result. When performing radiated emissions tests, it often makes sense to use the peak detector because radiated signals tend to be continuous or have a repetition rate well above 10 kHz.”
As the repetition rate is reduced, the averaging detector output drops more rapidly than the QP detector. In all cases, the peak detector gives the highest reading and has the fastest sweep. If the peak reading is below the limits specified for average measurements, the product is acceptable without further tests. Only if the peak detector readings are above these limits must further testing be done with the averaging and QP detectors.
Comparing the results of different detectors at any given frequency can provide details about the nature of a signal. For instance, if the peak, QP, and averaging detector levels are similar, the signal must be effectively continuous. If they differ by significant amounts, the signal is very impulsive. In between these points, you can develop rules of thumb for signal characterization.
What to Expect in the Near Future
Building on the existing technology and responding to the needs of users, you can expect continuing improvements in EMC test equipment. For example, the top-of-the-range interference frequencies, now typically up to 18 GHz, will increase, according to Karl-Heinz Weidner, product manager at Rohde & Schwarz. This will stimulate the need for and development of higher-performance receivers, antennas, and preamps.
“Test time will be reduced in the next few years,” is the opinion of Travise Miranda, manager of Com-Power. “This will be accomplished by further automation of test stations, use of more sophisticated software, and harmonization of the test standards.”
More attention will be given to test-site capabilities for radiated emissions testing. “Users need the convenience of an in-house setup with the performance of a good open-area test site (OATS),” suggested Mr. Mawdsley of Laplace Instruments. “The system must provide emissions measurements that are nearly equivalent to free-space results. This includes fully anechoic chambers, suitably calibrated test cells, and ad hoc open sites. On-site calibration capability also is on the shopping list.
“Because of experience and EMC education,” he continued, “more engineers will become familiar with emissions test technology. This will lead to more self-testing for precertification as well as a better understanding of the preparation for official certification tests.”
Acknowledgement
Some of the material for this article is from Mawdsley, D., Technical Notes, Laplace Instruments, September 2000.
Published by EE-Evaluation Engineering
All contents © 2000 Nelson Publishing Inc.
No reprint, distribution, or reuse in any medium is permitted
without the express written consent of the publisher.
December 2000