Why Electronic Products Fail FCC Testing

Shawn Singh

Ten of the most common mistakes and oversights committed by designers can easily be avoided, eliminating unnecessary expense and marketing delays.

FCC requires that electronic products intended for sale in the United States pass standards for electromagnetic interference (EMI). Product designers may know this, but that does not mean they always take FCC requirements into account during the design phase of a project. Too often, manufacturers do not detect problems until late in the product development cycle–perhaps only when they are getting ready to ship the product.

While its engineering staff may have design experience ideally suited for a product's intended market, a company can avoid costly fixes in later product stages if review by a competent compliance engineer and prototype prescans are part of the design phase. It costs a penny to make a change in engineering, a dime in production, and a dollar after the product is in the field. The old adage is correct.

Because FCC, like its equivalents around the world, sets emissions standards that all designs must meet, controlling unwanted electromagnetic radiation should be an important concern for every electronic device designer. Understanding typical reasons that products fail FCC testing enables a manufacturer to take control of the regulatory situation by designing and building its products to meet applicable standards and pass FCC tests the first time.

This article discusses the 10 most common causes of product failure to meet FCC requirements at first testing, which are:

• Ignoring or downplaying FCC requirements for the product during the initial design phase.
  
• Selecting the fastest components and highest clock rate.
  
• Using a single- or two-layer board instead of a multilayer PC board.
  
• Not considering emissions in clock layout.
  
• Not using enough bypass capacitors.
  
• Using unshielded cables.
  
• Using plastic connectors.
  
• Not using ferrites in cable design.
  
• Not incorporating a power line filter.
  
• Not shielding the chassis properly.

Too Little Attention to FCC Standards

The most fundamental compliance mistake designers make is giving little attention to FCC product requirements in the first place. Before even initiating the design phase of the product development cycle, savvy designers determine which FCC and other global compliance regulations are applicable. This enables them to determine the acceptable radiated and conducted limits from the outset of the project.

Proper electromagnetic compatibility (EMC) design techniques should be employed with every PCB, clock, cable, connector, enclosure, or other component that significantly affects the overall EMC performance of the product. Working with an experienced compliance engineer to review the design for proper EMC practices, along with conducting a prescan in the early stages of the design process, can dramatically increase the designer's odds of meeting FCC testing standards the first time.

Prescans for radiated emissions (noise that measures above 30 MHz) and conducted emissions (noise below 30 MHz) are performed on products to identify frequencies that pose a risk of violating the specified limits. Radiated-emissions prescans should be conducted in a semianechoic chamber. This type of chamber is free of ambient noise and suppresses the reflections experienced in a shielded room, thus ensuring accuracy of measurement.

The right reviews and prescans performed early in the design cycle can save substantial time and money, enabling a company to meet its budget and schedule. Correcting an EMC problem early on may be as simple as moving components and traces around on the PC board, something relatively easy and inexpensive to do during the design phase.
 

Logic Devices That Are Too Fast

When selecting a digital logic device for a particular application, design engineers are generally most interested in functionality and operating speed. Unfortunately, faster components tend to increase radio-frequency (RF) currents, crosstalk, and ringing.1

The main sources of radiation in digital circuits are the clocks and other fast-rise-time signals that are widely distributed in the system. Generally, lowering the clock speeds of the system will reduce emissions.

In addition, components with fast rise times are key contributors to radiated interference, even if they operate at lower clock frequencies; therefore, it is critical to control rise times when a product is designed.2 Almost all digital components have internal logic gates that operate at a faster edge rate than the propagation delay required for functionality. Designers should consider using slower-logic component families that meet the required propagation delay for their design function in order to minimize EMI. A good general rule is to select the slowest-logic family that fulfills the actual functional timing requirements of the circuit.

If fast-logic families are necessary for timing purposes, the designer should address individually the issues of special decoupling, routing, and handling of the clock traces.3 The simplest and least-expensive way to do this is to use a series resistor terminator to add resistance and thus slow the rise time of the signal.

A good design approach for minimizing EMI at the functional circuit level is to use the slowest clock speed compatible with the design and to control rise times by selecting the slowest possible logic family that will nevertheless maintain adequate timing margins.
 
The Wrong Type of PCB

Good product design employs multilayer printed circuit boards rather than single-layer boards wherever possible. Top and bottom ground planes with controlled impedances can reduce radiation from multilayer boards by 10 dB or more. Single-layer PCBs, although cheaper, demand more precautions in controlling EMI because loop sizes are necessarily large in all circumstances. Multilayer boards are the ultimate answer to PCB noise in general and to radiated EMI specifically.

The use of multilayer boards delivers several benefits. One is that they afford easy access to signal traces for repairs and temporary wiring changes. Multilayer boards minimize crosstalk, and therefore EMI, between the signal layers, as well. And their minimal spacing between the VCC and 0-V planes–only one layer of thickness–holds EMI to a lower level than would be the case with a single- or double-layer design.4
 
Careless Clock Layout

Several clock layout considerations are key. Clock generators, associated components, and distribution lines account for nearly all of the emissions generated on a PCB.

A clock circuit area is defined as the functional area that physically contains the clock oscillator or its buffer, drivers, and associated components, both active and passive. RF emissions are directly related to the rise and fall times (edges) of active components. Situating clock circuits near the center or a ground stitch location (to chassis ground) on the PCB rather than along the board perimeter or near the I/O section will minimize EMI. Separating the I/O lines from clocks on circuit boards can minimize unwanted coupling and the resulting higher emissions. Additionally, it is best to keep clock traces as short as possible in order to keep the lead inductance and loop area small, which in turn minimizes radiation. Again, to reduce EMI it is important to keep the return-path impedance low and to properly terminate clock lines to avoid excessive ringing.3
 
Insufficient Bypass Capacitors

Bypass is a technique that prevents energy transfer from one circuit to another. Bypass capacitors are used to reduce noise on PCBs. Designers selecting bypass capacitors should calculate the frequency of concern based on the logic family and clock speed used, then select capacitance value based on the reactance that the capacitor presents to the circuit. Above self-resonance, the capacitor becomes increasingly inductive, and this minimizes RF decoupling.

When determining how many capacitors to use, and of what type of technology, designers should consider the type, speed, purpose, and quantity of components on the board. Several capacitor manufacturers offer simple simulation programs to help in determining which capacitor will best meet the needs of a particular design.
 
Unshielded Cables

Since most of the noise generated within a system finds its way out through cables, the cables within a design often become unintentional noise antennas that create huge EMI problems. Suppression of EMI in a cable generally is achieved by means of shielding. Shielding cables can reduce emissions by up to 20 dB and minimize the problem of crosstalk. If high-speed clocks or signals with fast rise times travel across the cables, then care should be taken to ensure that proper shielding and termination are in place.

A pigtail is a connection in which the shield is brought down to a single wire and extended through a connector pin to the ground point. Because they are easy to assemble, pigtails are commonly used to connect the shields of data cable. However, it is advisable to avoid using cables with excessive pigtails, because those connections are hard to terminate properly.4

Many types of shielded cables are available. A shielded cable that consists of a single thin layer of aluminum foil loosely wrapped around the conductors may be suitable for some applications, but double-shielded cables that include a layer of tinned braid over the aluminum foil are the optimal choice for most applications.

Usually, the biggest problems are caused by the way the cable shield is connected at the ends rather than by the quality of the cable shield itself. Maximum benefit from a well-shielded cable will be realized only if the shield is properly terminated. Proper shield termination requires very-low-impedance ground connection and 360° contact with the shield.
 
The Wrong Type of Connectors

Connector leakage is a major source of cable problems. Metal or conductive plastic connectors offer protection against EMI at the termination point on the connector shell. The most effective shielding integrity is obtained by bonding a heavy metal cap from the cable shield to the equipment shield. Shielding integrity between a shielded cable and a shielded enclosure is maintained by using shielded connectors at both ends of the cable assembly.

Most styles of connector are available in shielded versions, allowing 360° contact on braid. A copper foil shield wraparound, which is soldered to the connector and the cable shield, provides a simple and effective technique for reducing radiation leakage. Generally, designers should choose metal or conductive plastic connectors when using shielded cable. Other plastic connectors will perform inadequately.
 
Cables Designed without Ferrites

Common-mode noise occurs in cables when the PCB signal connections and returns form a common impedance. This type of noise can be minimized through the use of proper PCB design techniques to reduce the common-mode impedance or by placing a ferrite bead around the cable.

A ferrite should be placed as close to the source as possible. Ferrites absorb energy and can reduce emissions by 10-20 dB. They are available in cylindrical and flat core shapes (see Figure 1). Split ferrites are designed for quick installation during troubleshooting or manufacturing. Best performance will be obtained if the inside diameter of the ferrite fits the cable sheath snugly.

In selecting a ferrite, the designer should consider the frequency at which maximum attenuation is required; ferrite permeability characteristics as they relate to the frequency range in question (that is, initial permeability); and the installation environment and mechanical attachment requirements. If a choice of shape is available, longer is better than thicker.

Figure 1. Ferrites come in a variety of shapes: split type, round, and flat. Photo courtesy of Steward (Chattanooga, TN).

Absence of a Power Line Filter

Power-line conducted radio-frequency interference (RFI) can be brought to a tolerable level by including a power line filter in the system. The filter suppresses conducted noise leaving the unit, reducing RFI to an acceptable range and lowering the susceptibility of the equipment to incoming power line noise that can affect performance.

A typical power line filter includes components to block both common-mode and differential-mode noise (see Figure 2). Common-mode noise in this case is EMI noise that is present on the line and neutrally referenced to safety ground. Differential-mode noise is EMI noise present on the phase line referenced to the neutral. The power line filter should be installed at the electronic system input to limit the level of EMI conducted along the power cord into the supply system.
 

Figure 2. A typical power line filter. Courtesy of Schaffner EMC Inc. (Edison, NJ).

Improper Chassis Shielding

A properly grounded and enclosed chassis design is key to controlling EMI. Designers should keep openings to a minimum, using the 1/20 wavelength rule: no opening larger than 1/20 the wavelength of the highest frequency.5 Seams become leaky even at 1/20 of the wavelength. Designers should avoid using a chassis with oxidized or painted steel pieces, and should make sure that the various chassis pieces make good electrical contact. In general, the shield should be tied to the circuit ground in as many places as possible. The enclosure should be a fully shielded six-sided box.

Several other guidelines for designers are worth noting. They should aim for 40 dB of shielding effectiveness at the highest frequency tested for emissions. When slots are necessary, materials should be overlapped. Designers should employ the most conductive material available for the highest frequency of interest. And they should use a milliohmmeter to check continuity across seams, different types of plating, and ground connections.
 

Conclusion

While FCC regulations and those of other global organizations do present design challenges, there is no reason they should cause project delays or added expense. Knowing and understanding the preeminent reasons for FCC test failure enumerated in this article enables a manufacturer to take full control of its own product development project. If a product is designed and built so as to meet the applicable standards starting from the earliest stages of the development process, making necessary design changes is relatively easy and inexpensive. Incorporating proper compliant-design techniques and testing from the outset gives the company the best chance to meet FCC and other regulatory requirements at the first challenge so that the project stays on schedule and within budget.
 
Shawn Singh is a senior compliance engineer at Percept Technology Labs (http://www.percept.com) in Boulder, CO. He can be reached at 303 444-7480 or via e-mail at shawns@percept.com.
 

References

1. Michel Mardiguian, Controlling Radiated Emissions by Design, 2nd ed. (New York: Wiley, 1992).

2. Kimmel Gerke Associates Ltd., "Designing for EMC," in Practical Tools, Tips and Techniques for Bullet Proof Designs, rev. A (West St. Paul, MN: Kimmel Gerke, 2000).

3. Mark I Montrose, Printed Circuit Board Design Techniques for EMC Compliance, 1996.

4. Tim Williams, EMC for Product Designers, (Oxford: Newnes, 1992).

5. Kimmel Gerke Associates Ltd., "Systems Grounding and Shielding," in Practical Tools, Tips and Techniques for System EMC, rev. A (West St. Paul, MN: Kimmel Gerke, 2000).