The Case for Combining EMC and Environmental Testing
The Impact of Extreme Environments Must be Factored into EMI Testing.
WILLIAM H. PARKER AND WAYNE TUSTIN
Equipment Reliability Institute
Santa Barbara, CA
Garwood Testing Laboratories
Pico Rivera, CA
Most Electromagnetic Compatibility (EMC), problems are just nuisances, such as an electric shaver interfering with radio reception. But some interference problems are far more serious, so that, for example, airline passengers may not use their laptop computers at certain critical times during each flight. To control interference problems, the U.S. Federal Communications Commission (FCC), the Canadian Standards Association (CSA), and the European Union (EU) require tests proving that equipment meets intersystem EMC standards; specifically, commercial aircraft and military EMC standards. Such testing is traditionally conducted under rather benign and clean laboratory climatic conditions. Temperature, humidity, and altitude are moderate. There are no condensation, contamination, or dynamic stresses (mechanical shocks, vibration).
Still, equipment, especially military equipment, is operated under field conditions where any or all of these, and possibly more, adverse environments can temporarily or even permanently degrade protections against interference.
This article explores the persistent concern that traditional (benign environment) EMC testing fails to reveal certain intersystem problems. It is proposed that all future test standards call for climatic and dynamic environmental tests to precede EMC tests or, for greater realism, to be conducted simultaneously with EMC testing.
QUICK REVIEW: EMI/EMC/EME AND TESTING
Problem: You are involved with electronic equipment that you discover is vulnerable to electrical interference.
Solution: You might place your equipment inside a metal box (i.e., Faraday cage). If your equipment is battery-powered and if the metal box is completely closed, your equipment will be able to withstand quite high levels of radiated electromagnetic fields from any direction.
Problem: You are involved with electronic equipment that, unfortunately, along with “doing its job,” radiates strong electromagnetic fields. These fields can interfere with other services.
Solution: If your equipment is battery-powered and, if you place it inside a metal box and, if the box is completely closed, the radiated electromagnetic fields from your equipment will be significantly reduced and, most likely, will not affect even nearby sensitive equipment.
Unfortunately, in a “real world” environment, Solutions 1 and 2 cannot be fully achieved. Most equipment operates from power brought in via cables from external power sources. Most equipment communicates with other equipment via interface cabling. Test standards dictate the measure of radiated electromagnetic emissions and electromagnetic susceptibility under which equipment must operate properly. Such tests may be performed in an EMI chamber. Most EMI chambers will attenuate RF signals as much as 80 dB over the frequency range of 10 kHz to 10 GHz. Other test standards (Example 2) limit how much electromagnetic radiation—both the wanted signals and the unwanted “noise”—you are permitted to generate.
BOXES MAY LEAK
It is very difficult to build a box that keeps 100% of the RF energy inside (Example 1) or outside (Example 2), particularly if the box has openings for cables, access doors, etc. Instructors often cite the somewhat comparable difficulty you would have in building an airtight house for your family. Even Columbia University’s famed “sealed” Biosphere 2 near Oracle, AZ had an annual atmosphere exchange rate near 10%.1 A practical house needs doors for human and equipment access. It needs ventilation for life support and cooling. It needs penetrations for pipes and cables, for fluids, and for electrical and communications connections.
Doors to RF enclosures never fit perfectly, so EMI gaskets must be added to them. But climatic stresses such as solar radiation, salt fog, sand, and dust can degrade gaskets. Further, metal surface treatments (intended to conduct current) can corrode. Finally, dynamic environments (vibration and shock) can degrade EMI gaskets.
PENETRATIONS AND CABLING
Cables are usually necessary, often with disconnects from connectors that actually pass through the skin of boxes or racks of equipment. It is undesirable for cables to conduct or radiate unwanted RF energy. Ferrite attenuators or braided shielding over individual conductors or entire cable harnesses can be used to attenuate this energy. Filters that pass useful signals but that block unwanted RF “noise” can be installed. Just how well do these shields, filters, and attenuators actually work in service? What will be the reliability of individual components and manufacturers’ materials in a “real world” that includes climatic and dynamic stresses? Are components mounted securely or is potting used? Actual environmental testing is far more useful than any theoretical attempts at reliability prediction.
BOXES AND RACKS DEFORM
Our box (or rack supporting several boxes) may be attached to a vehicle or to a fixed structure. Vehicles, such as automobiles, tanks, ships, or helicopters, deform as they move. Less obviously, even supposedly fixed structures move and deform. Think about the effects of building machinery, elevators, wind loading, and nearby highway inputs. Any structural deformations can exert force on a box or rack and could result in relative motion along originally tight metal seams. Further, a stressed box may move relative to its unstressed doors. Ultimately, apertures will lose their “RF tightness.”
VIBRATION AND RESONANCES
Electrical engineers use the letter “Q” as a measure of voltage magnification at resonance. Mechanical engineers use it as a measure of vibration magnification at resonance. Resonance occurs when a forcing frequency ff matches a natural frequency fn.
As a very simple example of resonance, place a favorite child onto a playground swing. By varying the timing of your pushes, you will very quickly find that you get greatest magnification “Q” when your ff matches the swing’s fn.
That playground swing is deceptively simple. It only has one natural frequency fn. Vehicles and buildings respond strongly at their several twisting and flexing natural frequencies ( fns) to vibratory inputs at various forcing frequencies ( ff s).
In an automobile, those forces come from the engine, drive train, wheels, etc. In an Army tank, these forces are joined by the “clanking” of the treads. A ship’s propeller is a strong vibration source. A helicopter’s rotors create considerable vibration.
Recognize that a box or rack has numerous twisting and flexing natural frequencies. If any of these fn s is excited by any of the ff s, a resonance that significantly magnifies RF leakage problems—perhaps by as much as 100 times—will occur.
Rather than waiting for failures to surface when the equipment is put into actual service, many potential problems can be uncovered by performing the vibration and EMC testing simultaneously.
COMBINED ENVIRONMENT TEST EXAMPLE
A client recently asked a well-known test laboratory to pass 100 watts of RF power (band reaching from 12 to 18 GHz) through a cable that was heated to 110C and to do this at a simulated altitude of 60,000 feet. The cable survived. The EMC technicians had some difficulty sealing the cable penetration into the thermal test chamber, but the altitude test specialists knew exactly what to do.
LOSS OF COOLING
Electronic systems depend upon cooling to prevent heat-caused performance degradation or outright failure. These systems are designed to operate under certain ambient altitude and air flow conditions. Heat received from an external source may compound the difficulty. Temperature conditions can degrade the performance of components and circuits. As an example of Problem 1, component values in a power supply filter may change at certain temperatures or altitudes. This change may compromise the efficacy of the filter, which performed properly at ambient conditions, so that power line noise is no longer attenuated effectively. RF interference from outside may disrupt the normal operation of the device. As an example of Problem 2, component value shifts can increase radiation from a device and can interfere with other systems nearby or which share the same power bus. These possibilities call for combined EMC (radiated and conducted emissions and susceptibility tests) and thermal testing.
INTERMITTENT CHASSIS GROUNDS
Unwanted RF signals (noise) are usually filtered to chassis ground, but what happens if the impedance of that ground connection degrades from near-zero Ohms to some higher value? This degradation can be caused by climatic variables such as temperature, moisture condensation, corrosion, etc. It can also be caused by vibration loosening, say, a bolted connection. The result, depending upon where the mishap occurs, can be an increase in RF susceptibility or in unwanted RF emissions.
Current flow through an intermittent ground can cause arcing, which can generate large amounts of broadband RF energy and can disturb the operation of nearby equipment. Two examples follow.
A PILOT COMPLAINED about navigation and communication equipment on his plane becoming inoperable when flying through rain. Studies have shown that when an aircraft flies through rain, static electricity on the aircraft skin can exceed 100,000 volts. In the hanger, testers simulated this situation by isolating the aircraft from ground. Using a high voltage power supply, they charged the aircraft skin to approximately 100,000 volts. A portable RF receiver was used to locate the source of broadband RF noise. It turned out to be arcing between poorly bonded aircraft surfaces.
AUTHOR MASONE RECALLS a flight test to document a similar problem. The test involved flying EMI specialists into a storm. Immediately upon entering the storm, Masone heard a high-pitched squealing sound from the pilot’s headset. Its intensity was such that the pilot had to remove his headset. The navigation display then went black. All navigation and communication equipment was inoperable in a whiteout condition with freezing rain and snow! Fortunately, the pilot was quite experienced, and there were no other aircraft in the flight path. Upon exiting the storm, all navigation and communication functions returned to normal operation. Back at the hanger, high voltage testing led to the discovery of a poor bond between two surfaces on the horizontal stabilizer.
It is not proposed that climatic or vibration tests be performed on the entire aircraft. These particular examples were used because they were known to one of the authors. But please recognize that similar poor bonds can occur in and on your boxes and racks of equipment, resulting in similar difficulties. Bonding problems will be exacerbated by vibration, shock, and corrosion. Corrosion will, in turn, be exacerbated by combinations of temperature, humidity (condensation), salt fog, etc. Better to uncover your product’s bonding problems in the test lab than in service.
On-off power switches are very common devices. They all work in ambient conditions; but under temperature, altitude, or vibration situations, contact arcing may occur and can generate broadband RF noise.
INTERMITTENT COVER PLATE BONDING
RF emissions were once observed to increase intermittently during a vibration test. The difficulty was traced to a cover plate losing its continuous RF bond. A somewhat similar loss of cover plate bond can be caused by differences in coefficients of thermal expansion (CTE) between box and cover plate materials. Such difficulties are relatively easy to fix, if the difficulties are known. Again, it’s better to find difficulties in the test lab than to find them in service.
COMMERCIAL AIRCRAFT AND MILITARY ENVIRONMENTAL AND EMC TEST STANDARDS
As an example of test procedures that factor in environmental stresses, consider RTCA DO-160E, the current issue of “Environmental Conditions and Test Procedures for Airborne Equipment.” 2 This is the aerospace standard for commercial aircraft. EMC concern ranges from DC to 18 GHz. Other standards are extended to 40GHz. Within the military, MIL-STD-461E3 is most commonly cited today.
Commercial or military EMC testing is seldom combined with climatic or dynamic (vibration and shock) testing. The authors decry this lack. This article encourages a comprehensive approach approximating actual in-service conditions. RF tests (Audio Frequency Conducted Susceptibility-Power Inputs Test, Induced Signal Susceptibility Test, Radio Frequency Susceptibility Test, Emission of Radio Frequency Energy Test) should be combined with climatic (temperature, altitude, humidity, waterproofness testing, fluid susceptibility testing, sand and dust testing, fungus resistance testing, salt spray testing) and with dynamic (sawtooth mechanical shock, sine and random vibration, explosion-proof) tests.
1. See www.bio2.edu.
2. Radio Technical Commission for Aeronautics (RTCA), Washington, DC. Information can be found at www.rtca.org.
3. Texts of MIL-STD 461E and other military standards can be found at www.astimage.daps.dla.mil.
WILLIAM H. PARKER holds a BSEE from North Carolina State University. He has more than 26 years of experience in EMC theory, analysis, specifications, design, testing, fixes, teaching, management, and marketing. Bill can be reached at [email protected] or visit www.equipment-reliability.com.
WAYNE TUSTIN has 50+ years experience in vibration and shock measurement, analysis, calibration and testing. His training and consulting interests lie primarily in the measurement, analysis and simulation of mechanical shock and vibration (particularly random vibration) that can cause equipment failure. Contact Wayne at [email protected] or visit www.vibrationandshock.com.
TONY MASONE has worked in EMC testing for over 15 years. His extensive background includes testing to Military, Aerospace, Medical, Commercial, NEBS and EPRI specifications. Tony is currently the EMC/EMI Lab Manager at Garwood Laboratories, Inc. at Pico Rivera (Los Angeles) California. He can be reached at [email protected] or visit www.garwoodtestlabs.com.