During EMC testing, many issues can be revealed for the first time. A common practice is to conduct a radiated prescan or preliminary emissions evaluation. This article examines a variety of compliance problems along with possible solutions.

Approved Components. One problem that is often identified during a prescan evaluation is failure of an assembled device that primarily consists of approved components and subassemblies. A recent specific example of this type of failure occurred during testing of an industrial printer. The main components of the printer were a processor control circuit, a laser printer head assembly, and a power supply. All three components had received an agency compliance approval. However, when combined and assembled into the printer, prescan testing showed that the system's conducted emissions exceeded the level acceptable for compliance—much to the surprise of the manufacturer.

A more-detailed analysis showed that the power supply responded differently under the real-world loading conditions of the printer compared with the load conditions of the qualification testing at the component level. This problem required the manufacturer to reevaluate the power supply and change its filtering components to ensure compliance with the conducted emissions criteria. In this case, even though the problem was discovered fairly early in the testing phase, it still delayed the introduction of the printer to market, causing lost revenue to the manufacturer.

Component Documentation. Another problem commonly confronted in the prescan stage is emissions that radiate from a communication line. Addressing this problem can be difficult when insufficient documentation is available regarding the components.

For example, on a communication card for a computing device, emissions were detected at the termination location of the connector, with an excessive amount of radio-frequency (RF) leakage present on the cable. Without the proper schematics with respect to the communication card, engineers were forced to spend unnecessary time individually reviewing each conductor line until the offending conductor was identified. If the manufacturers of the card had supplied the schematic, the problem would have been solved in much less time.

Enclosures. Testing often reveals that cosmetic features and treatments can create problems or can cause serious failures during the emissions testing process. A specific example of the effect of a product's housing was immediately evident on an arcade device under test (DUT). The device was housed in a wooden enclosure. Cables from the boards led to a fluorescent lighting structure, video monitor, and power supply.

When tested, the assembled arcade device failed the emissions test. The manufacturer was well into a manufacturing run of 50,000 units, and redesign was not a viable option. With this in mind, engineers designed and fabricated a metal Faraday-cage subenclosure and installed it within the
wooden enclosure. In addition, an entirely separate filter board was designed and built to ensure that all lines leaving the circuit board were treated to suppress unwanted electromagnetic energy.

The combination of the two additional features proved to be successful and allowed a process-capable design to be added
at a very late development stage. Such postdesign, Band-
Aid-style fixes are costly and tedious to implement as a retroactive design feature. This type of problem is much easier
to identify during early design evaluation when lower-cost
solutions can be developed.

When the electronics that are housed within an enclosure are tested and found to meet specifications, it is often the enclosure that is the problem. The location of mating surfaces is one area that consistently causes EMC testing failure. Areas of specific concern are doors, access panels, removable vents, covers, and windows.

It is critical to maintain a low-impedance, continuous-contact path at any seam or opening, which entails the use of a masking system (tape or other material) during the painting procedure. Removing the masking material from the enclosure after painting ensures a metal-to-metal surface contact.

Much testing on enclosures painted in these crucial areas has shown that the linear gaps created by the insulating properties of the paint can cause both immunity and emissions failures. For example, on a metal cabinet with a nonconductive powder paint covering all surfaces, the paint can be easily removed using masking techniques, but the exposed conductive surface must be treated to ensure long-term conductivity.

One solution to ensure long-term conductivity would be the use of conductive treatments such as iriditing. A common specification used for this treatment is MIL-STD C-5541 Class 3. It is important that the treatment is thin enough to ensure the best possible conductive path. A good way to test a conductive surface is to use an ohmmeter with the probes touching two pennies on the surface of the coating. It is
also important to remember that anodized surfaces are not conductive.

Doors or panels that are opened repeatedly require additional gasketing to ensure reliable electromagnetic compatibility. Depending on the environmental conditions, gasket choices can vary. Most gaskets offer easy installation with adhesive strips or friction-fit designs. Some offer low closure force to take up variable gaps. Nonconductive gaskets are almost never appropriate for EMC testing requirements and should be avoided if possible.

Immunity Testing

Immunity testing also can reveal serious design and compliance problems.

Lightning surge test equipment.

Surge Testing. One aspect of the immunity requirements, surge testing, involves the introduction of high-energy pulses into the power line or the input/output (I/O) lines. This test simulates the introduction of unique phenomena such as indirect lightning strikes. Failure during this test can literally destroy a product. Manufacturers must be made aware that this test is often destructive, and appropriate measures should be taken to ensure that in the unfortunate event of product failure, replacement circuit boards or components are available. The surge test should never be conducted with a one-of-a-kind prototype sample.

In one recent test, a manufacturer of an industrial gas
analyzer was required to meet this specific test criterion because the device was usually installed in an outdoor area susceptible to potential lightning strikes. During a simulated surge test, the power supply board overloaded and blew up, destroying the device. Unfortunately, the manufacturer provided a prototype device with only one set of boards.

Additional testing could not be performed, which significantly delayed marketing of the end product. Engineers developed a low-cost and process-capable surge-suppressing solution that was implemented in the redesigned analyzer. When retested with the suppression solution installed, the device passed with flying colors.

Radiated Immunity Testing. Radiated immunity is another testing requirement that can prove to be difficult to pass. In this test, the DUT is exposed to radiated electromagnetic energy. During the test, the device must continue to operate as intended, without any degradation of performance. Products or devices often malfunction only during this extreme test.

In one case, a laboratory analysis device that incorporates a temperature control module was being tested. The device performed flawlessly when exposed to RF energy up to 600 MHz, after which the temperature controller was fooled into believing that the temperature was lower than the actual programmed set point. This false temperature reading caused the heating element to engage, creating a thermal runaway condition.

The higher actual temperature caused inaccurate test analysis data within the device. Moreover, this condition could incur damage to the device itself. This damage would result in unreliable operation when the device was exposed to frequencies at or above 600 MHz, which includes radiation from cellular telephones, for example. A close inspection of the device enclosure revealed that a shielded window covering the display was not properly terminated, thereby causing the immunity failure. A simple grounding gasket was installed, ensuring 360° termination. This solution eliminated the problem completely.

Electrical fast-transient test in progress.

EFT Testing. An area that is often overlooked during early design evaluation is the effect of electrical fast transients (EFT). An EFT test simulates high-frequency disturbances on ac and I/O cables. These disturbances can be caused by a wide variety of components such as relays, switches, contacts, motors, fluorescent light ballasts, ignitions, welding devices, and certain types of thermostats.

The purpose of this test is to target digital circuitry because problems often manifest themselves in lockups, resets, data loss, or undesired operating modes. In extreme cases, EFTs can cause permanent damage to a DUT. An example of this occurred during a test of an uninterruptible power supply. During the test, the microprocessor interpreted the induced transients improperly, which caused the field-effect transistors to engage simultaneously and continuously, resulting in an overvoltage situation. This overvoltage caused the power supply to catch on fire.

After review and analysis of the problem, it was determined that the location of the line filter was the primary contributing factor. The I/O leads of the line filter were too close together, in effect bypassing the filter. Rerouting the filter leads to move them farther apart enabled the filter to function properly. It is rare that a device reaches this extreme condition, but it is worth noting and including in an evaluation checklist.

Engineer performing electrostatic discharge test.

ESD Testing. Electrostatic discharge (ESD) testing can be a destructive event. An ESD test is designed to simulate the static given off by a person handling a device or by a device touching another conductive surface. An ESD generator is used for this test. Traditionally, testing is performed to 8-kV levels. ESD failure can occur, for example, when plastic enclosures are treated with a conductive paint. Conductive paint allows for an aesthetically appealing end product.

During the ESD test in one case, however, the discharge found a path across the conductive paint to the circuit board. The electrical charge, finding the lowest-impedance path to ground, destroyed certain integrated circuits. This electrical charge results from the conductive paint not having enough conductive filler material on the surface of the housing to properly dissipate the electrical charge.

By replacing the paint with a conductive metal foil with proper termination, the discharge was able to dissipate over a larger surface and then was taken to ground through the termination point on the device. It is important to note that not all conductive paints lack the proper amount of conductive filler. However, without proper conductive filling, paint can cause a device to fail ESD testing.

With the examples described, most if not all of the problems identified could have been prevented. In nearly every case, products could have been redesigned with low-cost process-capable solutions that would not impede the requirements for introduction to market. In every case, the best possible solution is to identify and eliminate potential problems at the earliest possible design or testing stage, minimizing the impact on cost and time to market.