Introduction to Higher Signal Frequencies

Higher signal frequencies are becoming commonplace as many applications in diverse markets continue to fill the frequency spectrum. Fifth Generation (5G) wireless cellular systems demonstrate the trend, at one time operating only below 6 GHz but more recently moving to bands at 24 GHz and higher for increased wireless communications capacity. Signals operating in those higher bands are referred to as millimeter waves (mmWaves) in the electronics industry. The reason for this is that at frequencies from 30 to 300 GHz, the wavelengths of the electromagnetic (EM) waves are between just 1 to 10 millimeters (about 0.39 in) in length. mmWaves hold great promise in various technologies, two of the most extensive being automotive radars and wireless communications systems, provided that printed circuit boards (PCBs) operating as such high frequencies can be designed and fabricated cost-effectively. Applications that require reduced size, weight, and power (SWaP) only add to the list of challenges. With the latest design tools, suitable experience, and expertise, however, practical PCBs can bring mmWave solutions to growing numbers of users. 

5G and mmWave Design Considerations 

Every successful PCB engineering team with design and manufacturing experience has developed proven design cycle methodologies for advancing PCBs from concept to production. Reliable inputs are essential to that process, especially when designing at 5G and mmWave frequencies. Computer-aided design (CAD) and computer-aided manufacturing (CAM) software have become ubiquitous and can juggle the growing number of variables that must be considered as designs move into higher frequency bands. With these tools and a solid understanding of limitations in the manufacturing process, experienced engineers can develop prototype circuits capable of approaching a target set of performance goals.  Further analysis with the latest inspection equipment and automatic-test-equipment (ATE) systems makes it possible for prototypes to meet or exceed those goals. By incorporating techniques such as design for manufacturing (DFM) and design for test (DFT), a prototype design can transition to cost-effective mass production, even for applications operating at mmWave frequencies. 

PCBs containing RF/microwave and mmWave circuits are needed in many sizes and configurations for various applications, including automotive radars, satellite communications (satcom) systems, and 5G cellular wireless communications equipment. Depending upon the functionality required, those PCBs consist of as few as two to more than 30 layers, with analog, digital, power, and RF circuits often side by side. Higher-frequency circuits are typically based on microstrip, stripline, or grounded coplanar waveguide (GCPW) transmission-line technologies. PCBs operating at mmWave frequencies carry densely packed circuit configurations because of the microscopic wavelengths at those high frequencies. The general rule is that as wavelengths decrease, so must everything else in the signal path. Circuits feature short, closely spaced transmission lines with high current densities in small PCB sizes. Nearly everything contributes to signal fidelity in the art and science of design. 

Controlling Losses in PCB Design

Experienced designers know that signal loss is one of the primary challenges when designing at mmWave frequencies compared to lower frequencies, setting a solid need for thoughtful circuit designs that only minimally diminish signal energy. A useful metric for characterizing that loss in RF design is insertion loss, which is a measure of the energy lost to the circuit as the signal travels towards its intended destination. Many factors contribute to that loss, including conductor, dielectric, radiation, and leakage losses. Insertion loss captures the combined effect of these, but at mmWave frequencies it is essential to understand each in its own right.

Conductor Losses

Metal conductors such as copper suffer from increasing conductor loss with increasing frequency. This is primarily caused by a phenomenon known as skin effect, which is a tendency for current to migrate to the surface of the conductor as signal frequency increases, squeezing more and more current into a smaller and smaller cross-sectional area. The crowding experienced by the current near the surface is the source of the loss. Making matters worse, conductor materials with perfectly smooth surfaces are not available for existing manufactured circuit materials. The surface terrain across which current is forced to travel, known as surface roughness, can contribute significant amounts of loss, especially at mmWave frequencies, unless great care is taken in choosing materials with adequately smooth surfaces.  The effects of surface roughness include variations in signal amplitude and phase, which are harmful, for example, to the accuracy of mmWave weather or automotive radar systems. Using circuit materials with smooth copper surfaces is essential for these technologies to function.

One additional consideration is conductor plating. While copper is commonly used as a conductor base metal, due to concerns with oxidation, it is typically either covered with a nonconductive protective coating or plated over with an oxidation-resistant surface finish. Various surface finishes are common, such as electroless nickel immersion gold (ENIG), but what they all have in common is a tendency to further reduce signal integrity by increasing loss or altering a signal’s phase characteristics. The effects of plating cannot be ignored when estimating the performance of a mmWave PCB.

Dielectric Losses

Loss from a PCB’s dielectric material increases with decreased thickness. There is therefore a tradeoff between loss and trends to reduce PCB size and weight. Such loss is defined by a dielectric material’s essential characteristics, such as dielectric constant (Dk) and dissipation factor (Df). Circuit materials with low Dk and Df values exhibit low dielectric loss. For mmWave circuits, it is desirable that Dk and Df values be consistent, both spatially across the length of the dielectric material and across the frequency range of the intended application. Even small variations in Dk and Df can contribute to unacceptable variations in a signal’s amplitude and phase as it travels along the transmission lines printed on the material. Thermal stability is also a concern, and a good circuit material should be characterized by Dk and Df values that remain stable with temperature, at least across the temperature range of the PCB’s intended application. 

Radiation Losses

Radiation losses occur at disruptions in a transmission path such as the transition between a connector pin and a PCB transmission line. At the discontinuity, the characteristic impedance changes, signal reflections occur, and standing waves develop. Eliminating these losses requires that a transition be tuned as closely as possible to the circuit’s characteristic impedance, usually 50 Ω, or that transition features be kept small, no larger than one-twelfth of a wavelength of the operating frequency. This is especially important at mmWave frequencies. Unless feature sizes are sufficiently small, a designer may be designing in an antenna without even realizing it.

Leakage Losses

Leakage losses can occur from unwanted coupling to surrounding circuitry. While leakage losses at lower RF and microwave frequencies are generally only a concern for higher-power circuits, the same is not true for the limited signal strength available at mmWave frequencies. 

Minimizing Circuit Variations at mmWave Frequencies

Because of the tiny wavelengths of mmWave signals, variations in circuit materials and circuit features can distort the signal’s amplitude and phase, and those variations should be minimized. Phase responses, for example, are commonly used for modulation in communications signals and target information in a radar system. The phase length for a signal at a given frequency will increase as the Dk of the circuit material increases with more significant phase changes at higher frequencies. A phase aberration for a PCB’s signal at 30 GHz will be one-half the phase distortion for the identical PCB at 60 GHz, where two wavelengths are equivalent to one wavelength at 30 GHz.

With the shrinking wavelengths at mmWave frequencies, the effects of mechanical circuit variations on signal amplitude and phase must be considered during the design stage and in the choice of a manufacturing process so that the fine circuit features and tight tolerances needed for mmWave frequencies can be produced in the volumes required for an application.  

Circuit Layout for mmWave PCBs

Once a circuit material or materials have been selected as the optimum foundation for a mmWave PCB, a circuit layout that reflects the physics of mmWave frequencies must be created for the appropriate portions of the PCB. For example, for circuits where phase matching of transmission lines is critical, such as in phased-array antennas, some space should be left on the PCB to allow tuning trace lengths to required values. In addition, as dense as circuits become at mmWave frequencies, enough room should be allowed to isolate any radiating components and to prevent capacitive and inductive coupling or crosstalk between transmission lines that may be too closely spaced. 

A schematic diagram, creatively applied, can help speed, and simplify the layout process for mmWave circuits by clearly identifying transmission lines and components that may require special attention. Components that generate noise, for example, should be away from board edges and as closely centered as possible to minimize noise. Component placement should attempt to reduce transmission distances traveled by high-frequency and high-speed-digital (HSD) signals to minimize degradation of those signals. 

Manufacturing Cost of PCBs

Layout choices can have a great deal to do with the final manufacturing cost of a PCB since drilling holes and extra plating and lamination are often unexpected steps in preparing a circuit design for the manufacturing process. For multilayer PCBs, plated through holes are typically used for signal and ground interconnections, and the placement of these mechanical structures will depend upon the type and location of a PCB’s components. The placement and layout of the interconnections will also differ according to the choice of transmission-line technology. 

Plated through holes exhibit much higher impedance than the typical characteristic impedance (50 Ω) of the transmission lines they link. Any impedance mismatch must be corrected to minimize insertion and return losses at those junctions. The via stubs, which are the extensions of the via holes beyond the target layer in the PCB stack up, also introduce parasitic effects to the transmission lines, which increase with increasing frequencies. While necessary for the multiple layers of a compact PCB, plated through holes introduce mechanical and electrical aberrations to a circuit, which must be considered during the design stage or corrected later when a PCB prototype is being evaluated during the inspection and test stages. 

CAD and CAM Software

Simulations with commercial CAD and CAM software provide close estimates of the impact of various layout configurations and the effects of component choices on PCB performance at mmWave frequencies. A CAD tool, for example, can accurately calculate the transmission line and package losses between surface mount technology (SMT) components on a PCB for any number of circuit-board placements or even the added signal losses contributed by coaxial connectors. CAD tools can also estimate the results of lower practical Dk values on return loss and insertion loss when evaluating a choice of circuit materials. More advanced 3D EM simulation software can even model the effects of added materials and structures, such as surface finishes and plated through holes. While CAD software cannot be expected to eliminate the need for building a prototype as part of the strategy for developing a PCB layout that meets electrical, mechanical, and environmental requirements, it can help reduce the number of prototypes that are needed for achieving a PCB’s expected performance levels.  

Design and Development

As an example of PCB design and fabrication capabilities into the mmWave frequency range, design engineers at Benchmark have pursued the most demanding needs for both HSD and mmWave PCBs for several years, tracking sets of electrical and mechanical requirements from simulation through production with the aid of internal CAD tools, manufacturing systems such as automated SMT assembly systems, electrical and optical inspection equipment, and ATE systems. Backed by the plethora of available tools, these engineers’ experience makes practical PCB solutions possible even at mmWave frequencies, which also comply with the latest DFM and DFT requirements. 

From radio frequency (RF) to millimeter wave, optical, and HSD, translating design specifications into PCB hardware requires the right experience and toolset to interweave the numerous circuit technologies in modern high-frequency electronics. It requires a solid understanding of first principles and the ability to juggle many component inputs and material considerations. Current CAD programs are invaluable in developing realistic design models for reliable predictions, achieving performance optimization, and meeting reduced SWaP needs. Reinforced by extensive test-and-measurement capabilities, innovative (but practical) PCBs can be developed and manufactured repeatably and reliably to meet modern needs for increasing circuit functionality in shrinking PCB sizes.

References

1. John Coonrod, “PCB Design and Fabrication Concerns for Millimeter-Wave Circuits,” Printed Circuit Design & Fab, March 2021, pp. 28-35, www.pcdandf.com. 

2. Benchmark Technology, www.bench.com, “Arming Antennas for Transition to 6G,” company blog, 2022.