Circuit Board Housing Cavity Resonances
What is going on here? Elementary microwave engineering states that that any enclosure will have certain resonant frequencies which will oscillate. These occur roughly when a dimension of a cavity is one half the free space wavelength of a multiple thereof. For high frequencies or large cavities multiple resonant frequencies can oscillate increasing the likelihood that your device will radiate extraneous signals that will cause the cavity to resonate.
Why is this bad? A standing wave inside a cavity will vary the impedances in the cavity. At the walls (assuming metallic or otherwise conductive wall) the transverse E field and hence the impedance will be zero while a quarter wavelength away the impedance becomes infinite (assuming no loss). Depending on where the VSWR peaks and nulls occur they would affect the input or output impedances of your device.
The calculation of the resonant frequencies and field distribution of a rectangular cavity is fairly straightforward. However, a large cavity will have multiple resonance modes and frequencies. Knowledge of how the resonance is excited would be needed to fully determine the field distribution and this is often difficult if not impossible.
Adding microwave absorbing material to a cavity has proven to be a quick and inexpensive way to eliminate cavity resonances. From basic field theory, the insertion of a high permeability/permittivity material in a cavity will cause the field distribution to shift such that the vast majority of the energy is oscillating within the material. If that material is also lossy (high magnetic or dielectric loss tangent) then the energy will also be attenuated.
Determining the actual field solutions and excitation mode for a particular component in a cavity might be an interesting academic exercise but most engineers do not have the time for this. Cutting and pasting absorber pieces in various places on the circuit board cover and testing circuit performance has been the most common method of choosing the best solution. Absorber manufacturers can give recommendations on material types and thicknesses that experience has shown to work but it is extremely difficult to determine a particular solution that will work without detailed knowledge of the circuit which would be difficult or impossible to determine.
Emerson & Cuming Microwave Products offers a free Problem Solver Kit for just this purpose. Enclosed are samples of various materials complete with pressure sensitive peel-and-stick adhesives that you can try in your circuit immediately. The materials cover a broad range of frequencies.
Considerations in choosing a cavity resonance absorber
- Material type- Magnetically loaded, dielectrically loaded, elastomer, ,epoxy, foam etc
- Material thickness- In general thinner material can be used at higher frequencies
- Method of adhesion- Peel-and-stick adhesive, liquid contact adhesive, mold-in-place
- Outgassing considerations- In hermetic packages, outgassing can be very important, particularly hydrogen outgassing in a circuit containing Gallium Arsenide components
Material Type- Basic field theory states that the transverse electric field is zero at a conducting wall and the magnetic field is at a maximum. For this reason magnetically loaded (with either a iron based material or a ferrite for lower frequencies) material is usually recommended for these applications. These materials are usually elastomers (silicone or urethane) but are sometimes epoxies. These materials have the advantage of being relatively thin (0.010"-0.100") and nonconductive so they will not short out a circuit if they come in contact. A less expensive solution can be had with a dielectrically loaded foam material. The tradeoff is that these materials are thicker (0.125"+) and conductive and hence could short out a circuit.
Material thickness- It is often difficult or impossible to determine exactly how much attenuation is needed. A rule of thumb is to use the thinnest material that will do the job as that will minimize costs. ECMP can make most materials at virtually any thickness (>0.010") to minimize your absorber expense.
Method of adhesion- In the past, liquid adhesive systems were used to adhere the absorber to the cavity walls. This was mess, time consuming and not cost effective. With the development of pressure sensitive peel-and-stick absorbers, a large step was taken in increasing the ease of application. Material could be supplied precut to shape with the adhesive for quick and sturdy adhesion. For a more permanent solution ECMP offers its Mold-in-Place technology. This is a service whereby the circuit board housings would have an epoxy based absorber material applied directly to the cavities. The material is then virtually impossible to remove without machining but might be the application of choice when a long lifetime is desired.
SOURCE: Emerson & Cuming Microwave Products