Miniaturization of handsets and other wireless devices creates scores of shielding challenges as high-frequency components become more closely spaced. As printed-circuit boards (PCBs) shrink, new electromagnetic-interference (EMI) shielding solutions must provide greater levels of interference suppression, but without significantly adding mass, weight, and cost to a device. Fortunately, a new shielding technology developed by W.L. Gore & Associates called snapSHOT™ shield, replaces bulkier soldered approaches with a snap-on metallized thermoformed shell that can be attached to a PCB by means of standard ball-grid-array (BGA) solder spheres.

In understanding the shielding requirements of compact wireless designs, the shortcomings of traditional test methods (mostly based on military requirements) become apparent when applied to battery-powered portable wireless designs. Some test methods, such as ASTM D 4935¹ for planar shielding and a coaxial cell² for EMI gaskets, for example were developed to characterize the materials that would ultimately comprise an EMI enclosure. But there remain no formal published test methods to evaluate the shielding effectiveness (SE) of a shield assembled to the PCB of a portable wireless device.

At present, two primary shielding approaches are used in cellular telephones: soldered perforated cans and plated covers with EMI gaskets (Fig. 1). Both try to create a complete shield around the PCB's components to ensure proper electrical performance and comply with regulatory requirements for EMI emissions and susceptance.

The goal of an EMI shield is to create a Faraday cage around the enclosed RF components using the six sides of a metallic box. The top five sides are created using a shielding cover or metal can, while the bottom side is achieved by using the ground plane within the PCB. In an ideal enclosure, no emissions would enter or exit the box. In reality, leaks do occur, such as from holes perforated into soldered cans that allow thermal heat transfer during solder reflow. Leaks can also occur from imperfections along an EMI gasket or solder attachments. Leaks are also possible from the spaces between ground viaholes used to electrically connect the shielding cover to the ground plane.

Shields designed for portable devices must be light in weight and low in cost, but they must also meet demanding mechanical and electrical requirements. Phenomenon such as cavity resonance, aperture radiation, and planar shielding are factors RF engineers face when designing shielding enclosures. The problem is further complicated by the fact that accurate EM field prediction from complicated PCB assemblies, particularly in the near field, is virtually impossible, forcing many engineers to build custom test fixtures to evaluate their designs.

To create the Faraday cage required for proper shielding, a metallic enclosure must be placed around and in close proximity to the components on a PCB. Unfortunately, this may have adverse effects on the performance of the components and the functionality of the circuit, with the greatest concern being enclosure (cavity) resonances at any of the PCB's operating frequencies. To study this, a simple test fixture was designed to mimic the effect of placing a metallic enclosure over RF components.

The test fixture (Fig. 2) consists of two 50-Ω, 0805 resistors that are launched from SMA connectors from the opposite side of the ground plane. The spacing is arbitrarily set at 0.5 in. (1.27 cm) so that minimal coupling would occur between the two components when the shield is not in place (Fig. 3). Coupling was determined from 20logS₂₁ measurements on a microwave vector network analyzer (VNA). A perforated metal square can, with inside dimensions of 1.805 × 0.114 in. (4.584 × 0.29 cm), was soldered over the components to illustrate the effects of a metal cavity. The simple formula in Fig. 3 was used to roughly calculate the resonant modes of the EMI enclosure. The formula applies to rectangular cavities and is fairly accurate if the cavity is filled with air. However, most enclosures on PCBs will include the PCB material and components within them, raising the cavity's effective dielectric constant and thus lowering its resonant frequency.

With the shield in place, coupling between the two components is severely increased at and around the resonances, as much as 50 dB for this fixture. Peaks occur at the resonant frequencies calculated by the formula. Below the first resonant frequency, the coupling between the two components is virtually unchanged. Thus, it is critical to consider these resonant conditions when designing EMI enclosures.

A more accurate way to predict the effects of a rectangular cavity involves the use of an EM field simulator, such as the Sonnet^® Professional Planar Software Suite from Sonnet Software (Liverpool, NY). This software models planar circuits within a metallic box, and can easily be used to examine the effects of cavity dimensions, substrate material, and metal wall conductivity. (A free version of the software is available on Sonnet's website at www.sonnetusa.com.) Figure 4 shows the response of a model produced by the software, with close agreement to the actual measured data.