3D printing using metallization and dielectric inks with suitable RF characteristics are providing new options for fabricating GHz-range devices.
You’re undoubtedly familiar with additive manufacturing (AM), often referred to as “3D printing” (which is a casual but reasonably-accurate description). The performance of AM systems hardware has increased substantially since its introduction a few decades ago with respect to dimensional precision, accuracy, finish, and cost; no surprise there. At the same time, many new materials have been added to the list of available resins, polymers, liquefied and powdered metals, and other specialty “inks” (as they are often called). There are enough choices that you can likely find one with the desired combination of mechanical properties.
The broad term “additive manufacturing” actually refers to any one of seven classifications per “ASTM Committee F42 – Additive Manufacturing” (ASTM previously stood for the American Society for Testing of Materials, but they legally shortened their name, as so many other organizations and companies have done). These classifications are:
Among the many applications for AM are:
In recent years, AM has moved into the electronic-component area in addition to purely mechanical parts, especially for higher-frequency RF applications. Many of this RF-centric AM opportunities involve more than metallic surfaces alone, but also require suitable dielectrics, so the AM challenge is to find and implement a suitable geometry plus layering of conductors and dielectric. There has been activity in four primary areas:
AM when used for RF has another dilemma: many of the applications that could benefit the most from the integration are at higher frequencies going into the single- and double-digit gigahertz range. However, as frequencies increase, the corresponding dimensions shrink while the impact of tolerance issues increases. A seemingly-insignificant dimensional imperfection or micro-roughness in surface smoothness that is tolerable at 500 MHz can be a serious problem at several gigahertz.
Further, the detailed RF specifications of AM dielectrics need to be characterized and consistent, whereas they are irrelevant for mechanical-only AM parts. Factors such as dielectric constant (permittivity) and loss tangent are important. Resins and polymers with attractive insulating properties may have poor RF characteristics. On the other hand, it is also possible to deliberately vary the material mix to enable creation of specialized dielectrics with useful properties, somewhat analogous to stepped- and even graded-index optical fibers.
Despite all these issues – or perhaps because of them – the potential benefits of AM for higher-frequency RF are getting a lot of attention from university and corporate researchers. In some cases, they are exploring and exploiting it to make smaller, better components; in other cases, they are looking to use AM to fabricate designs that are impractical or impossible to achieve using conventional techniques. Some of the projects use AM to create the entire device and its metal surfaces, while others use AM for the non-conductive body and then use standard electroplating of copper or silver to add the needed conducting surfaces.
Two examples illustrate the efforts underway. One team varied the dielectric permittivity of the resin “infill” to create a slab waveguide in order to increase the cutoff frequency of the second electromagnetic mode yet with almost no effect on the fundamental mode. The device had a central portion with higher dielectric permittivity, where the electric field of the fundamental, quasi-TE10 mode is more intense, while the two side portions had lower permittivity, where the second, quasi-TE20 mode is stronger (Figure 1 and Reference 1).
Figure 1 The top photograph shows a 3D-printed substrate integrated slab waveguide (SISW) interconnect, prior to pasting the aluminum foils and adding the metal vias (a). The bottom photo shows a SISW interconnect after pasting the aluminum foil (b). Source: Radio Engineering
Another interesting project used AM to produce a integrated SMA-to-waveguide transition for X-band energy (Reference 2). They created a dielectric-filled unit for 8.6 to 10.4 GHz and an air-filled unit for 9.4 to 10.7 GHz, each with somewhat different performance specifics (Figure 2).
Figure 2 This diagram illustrates the manufacturing and assembly process of the dielectric-filled integrated waveguide design. Source: State University of New York/New Paltz
These are just two of many examples of the work being done in this area. A basic Google search of “additive manufacturing RF” will turn up dozens of papers and projects, but many are behind the paywalls of the various IEEE societies or other sources who published them; most of the references cited below are open and not restricted.
Have you been following the use of AM for non-RF and RF-focused electronic applications? Have you considered using AM for RF components? Or have you already done so, whether for quick breadboards and prototypes or even low-volume production?
References
This article was originally published on Planet Analog.