All companies striving to survive and grow in today’s worldwide electronics industry recognize the value of performing design, development, and troubleshooting tests. These tests should be done early and often during a program to minimize overall costs to meet the formal EMC compliance requirements. It’s not necessary to test everything with every procedure every time the sample is tested, only those that may be a problem.
EMC tests can be categorized as conducted emissions (CE), radiated emissions (RE), conducted susceptibility/immunity (CS/I), and radiated susceptibility/immunity (RS/I) which can be further grouped by frequency: (a) DC to 1 GHz, (b) 1 GHz to 18 GHz, and (c) 18 GHz and higher. All companies need to test frequency range (a). Companies building equipment that falls within frequency ranges (b) or (c) are stuck with it.
Part 1, which appeared in the August issue, presented a comprehensive list of suggested EMC test equipment for frequencies to 1 GHz: a receiver/spectrum analyzer; a multipurpose ESD generator; a low-noise amplifier (LNA); a clamp-on current probe ranging from 10 kHz to 100+ MHz; assorted magnetic field (HF) and electric field (EF) probes; a line impedance stabilization network (LISN), 50-?, one for each power line; and a biconical antenna usable from 10 to 500 MHz.
Also helpful would be a laboratory DC supply, an oscilloscope, an EUT test stand, a ground plane, a DMM with LCR measurement capabilities, power line filters, an isolation transformer, a parallel plate line, coax cable, aluminum foil, 50-? terminations, a 3- to 2-wire AC adapter, and assorted ferrites, capacitors, and shielding parts. And, a shielded enclosure would be nice.
Part 2 addresses design/development testing and troubleshooting of conducted emissions, conducted susceptibility, radiated emissions, and radiated susceptibility EMC problems.
Design/Development and Troubleshooting Testing
As soon as prototype hardware becomes available, design/development testing should begin. Using the equipment on the suggested equipment list, tests could be run exactly like the test standard, but they do not have to be.
Meaningful results can be obtained from tests that will likely not look like a formal EMC test. Still, they certainly will pay for themselves. Tests should be picked based on the characteristics of the hardware, company history of the kinds of problems similar equipment has had, and the current state of the equipment.
If it’s a haywired prototype running on lab power supplies on a bench in the middle of the development lab, radiated testing is not all that reliable and may interfere with other work. As a result, tests generally are restricted to CE and CS/I. Even then, the resonances of the haywiring will make differences in the emissions/susceptibility levels but not nearly as much as the radiated tests. When the physical size of the hardware begins to contract into something more like the final package, the radiated emissions tests have greater meaning.
Emissions Tests
Interference problems start out as conducted energy and end up the same way. Conducted emissions currents originating from the test sample may exit the envelope of the test sample as a desired conducted signal, an unwanted signal, or even a combination of both depending on the frequency. If the current amplitude changes as a function of time, the accelerating electrons create a changing magnetic field that, in turn, generates a changing electric field.
Conducted emissions create radiated emissions. This makes it important to perform the CE tests first, and the test sample setup should approximate the test specification requirements if possible. Figure 1 illustrates a test setup for emissions testing.
Figure 1. Design and Development Emissions Testing
Emissions tests should be run as high in frequency as your equipment will permit. If there are any out-of-spec CE readings, fix those first and reduce the amplitudes of any others. Since the radiated emissions are a result of conducted emissions, the CE problems are solved before addressing radiated emissions problems.
Next, perform the RE tests. Note any frequencies that are out of spec and switch to the current probes, then the HF/EF probes, and check all cables for the problem frequencies. After that, remove the cables from the test sample and check the box emissions. Solve the RE problems before continuing on to the susceptibility tests. Because emissions and susceptibility fixes are reciprocal, the techniques done to reduce the conducted and radiated emissions help control conducted and radiated susceptibility/immunity problems.
Whenever possible, while performing CE and RE measurements, use a receiver or spectrum analyzer with an audio output to listen to the AM-detected RF. It is amazing how much design and troubleshooting information is contained in the sounds from EMI. Tune slowly through the RF spectrum while testing. If DC to daylight is tuned in a microsecond, everything passes.
Intermittent problems may be the result of bad connections on the PCBs or connectors or caused by some nearby source. This generally will require an in-depth investigation about what is happening in the RF environment before turning on the test equipment.
Susceptibility/Immunity Tests
External radiated RF energy creates conducted currents in exposed wires and cables outside the test sample envelope. Because of poor antenna efficiencies in the conducted susceptibility frequency ranges, it takes large radiated susceptibility signals to produce the same signal levels as the conducted susceptibility/immunity tests. Therefore, conducted tests normally are performed first and solved before continuing on to radiated susceptibility.
Unfortunately, CS/I tests are only a partial replacement for RS/I testing because CS/I tests are performed on single cables whereas an RS/I test illuminates everything at once. However, a CS/I test does not require a shielded enclosure. Running an RS/I test in the open generally is illegal, plus it gives the regulatory agencies such as the FAA and FCC heartburn when RF starts squirting into the ether without permission.
Most modern EMC specifications require susceptibility testing which necessitates a large, expensive assortment of signal generators; RF power amplifiers; antennas; transient generators; and specialized coupling networks. There are a large number of transient-type susceptibility tests. An ESD generator does an OK job approximating those test procedures, and it will duplicate an ESD test exactly.
It’s not possible to emulate all required susceptibility tests using an ESD generator because some are narrowband tests. As an example, the tuned RF susceptibility tests, CS114 and RS 103 and the EU equivalents, sweep a single modulated carrier frequency slowly through the spectrum and look for a response. An ESD generator excites a large portion of the total frequency range simultaneously with a broadband signal and looks for a response. If something responds when using an ESD generator, the frequency is unknown, but the data will indicate a problem.
Here’s a brief overview of how the ESD test is used to simulate the primary susceptibility tests:
•?ESD: Figure 2 shows the test setup for direct and indirect ESD measurements. This setup illustrates ANSI C63.16, but IEC 61000-4-2 is nearly the same. All ESD-based susceptibility tests use a similar approach.
The test represents a charged human touching a conducting surface on the sample. A 4,000-V or 8,000-V ESD transient is applied to all exposed metal switches, screws, connector bodies, or other metallic surfaces. Because the charged human might contact a conducting surface near the test sample, the requirement also specifies an indirect test to an adjacent metal plate.
•?Radiated RF Susceptibility: The indirect test called out as a part of the ESD test is a radiated RF susceptibility test. For a larger test sample with cables, the coupling can be enhanced to make this a better simulation of a radiated electric field. One way is to tape a 1.5 m x 30 cm loop made from insulated wire directly to the box and cables as called out in MIL-STD-462 Notice 6 (USAF-1987) for the RS06 Chattering Relay Test. Figure 3 shows the test setup.
Figure 3. RS06 Chattering Relay TestMIL-STD-462 Notice 6 (USAF) 15 Oct 1987
The RS06 test is a relatively good and inexpensive susceptibility test. It has fallen into disfavor because of its lack of repeatability. Using an ESD generator as the drive minimizes the repeatability problem but reduces the modulation frequency. If an organization does not have an ESD generator, the RS06 test can get the susceptibility test program started.
•?Conducted RF Susceptibility and the Electrical Fast Transient (EFT): These tests are performed by applying the ESD transient directly to cable shields in a manner like the RTCA DO 160 lightning test. The cable transfer impedance couples the transient to the internal wires. Using an ESD generator for conducted RF susceptibility does not duplicate the swept narrowband RF character of that test, but the ESD current is higher. The test sample is grounded to the ground plane at one point, and the ESD transient is applied to the cable shield at the end farthest from the test sample. The current travels down the cable shield to the ground point and returns to the ESD generator.
The ESD simulation of the EFT is almost a duplicate of the test. Instead of the capacitive coupling fixture, a length of aluminum foil is positioned parallel with and wrapped around the cable forming a tube. The tube then is grounded at the end closest to the test sample. ESD is applied to the other end of the aluminum foil tube. In this case, the ESD transient can be adjusted to be nearly the same as the EFT requirement. The conducted susceptibility tests are repeated for all cables.
Conclusion
This article briefly covers testing for design, development, and troubleshooting of CE, CS, RE, and RS EMC problems. For more information, some excellent books can help fill in the blanks:
•?EMI Troubleshooting Techniques, Michel Mardiguian, McGraw Hill, 1999.
•?The Technician EMI Handbook: Clues and Solutions, Joseph Carr, Newnes, 2000.
•?High-Frequency Measurements and Noise in Electronic Systems, Doug Smith, John Wiley, 1992.
About the Author
Ron Brewer currently is a senior EMC/RF engineering analyst with Analex at the NASA Kennedy Space Center. The NARTE-certified EMC/ESD engineer has worked full-time in the EMC field for more than 30 years. Mr. Brewer was named Distinguished Lecturer by the IEEE EMC Society and has taught more than 385 EMC technical short courses in 29 countries and published numerous papers on EMC/ESD and shielding design. He completed undergraduate and graduate work in engineering science and physics at the University of Michigan. e-mail: [email protected]
September 2009