Are 3D-printed RF passives and assemblies finally ready?

Article By : Jean-Jaques (JJ) DeLisle

Metallized 3D printed dielectric structures can be used in place of full metal structures for resonators, housings, and transmission lines/waveguides without taking a substantial hit on performance.

RF electronics components, assemblies, and parts that aren’t solid state, generally require rather exacting manufacturing tolerances to perform adequately. In the large mix of trade-offs for RF design, there usually isn’t much wiggle room to allow for poor manufacturing processes or materials. This is usually why RF parts are expensive and come with long lead times. The limited number of RF designers and experts competent enough to source RF designs and facilitate the production of RF parts also explains why there are very few vendors for quality RF parts.

This outlook is certainly frequency and application dependent. For relatively low frequency (sub-1 GHz) and relatively low power (microwatts to only a few milliwatts), there are more part options and utilities. This is also true for some applications, such as parts for the 2.4 GHz ISM band—supporting Wi-Fi, Bluetooth and Zigbee—which are readily available due to the popularity of these frequency bands.

It isn’t surprising then that new manufacturing methods, especially if they offer greater design freedom, material options and other fabrication benefits, would perk the ears of RF designers. This was likely the case for many when 3D printing first started becoming more accessible in the mid-to-late 2000s. In those early years, fused deposition modeling (FDM) was the talk of the town. Though many exciting concepts and parts can be made with FDM 3D printed objects, available filaments and early tolerances, repeatability, and material properties fell short for a vast majority of RF applications, especially those beyond a few GHz.

Also, FDM printing has a limit of feature size related to the nozzle size, so higher resolution printing leads to much longer printing times. Generally, FDM filaments of the time had relatively high complex dielectric properties, dielectric constant (Dk) and loss tangent (tand). Eventually, conductive filaments were introduced, which again perked the ears of hopeful RF designers. But alas, these conductive filaments were not really all that conductive and often introduced unfortunate manufacturing characteristics.

When stereolithography (SLA) and digital light projection (DLP) 3D printing came of age in the early 2010s, there was again the hope of tighter tolerances, smaller feature sizes, and technical- or engineering-grade resins. Though higher Dk and tand than generally desired, many SLA/DLP compatible resins did enable early prototypes and some options for lower frequency parts. Typically, too lossy for use beyond a few GHz, SLA/DLP technologies did offer the resolution needed to meet tolerance requirements for higher frequency applications. All that was needed were resins that had much lower Dk and tand to make 3D printed dielectrics possible.

Around the same time, metal powder bed fusion (MPBF)—in which direct metal laser sintering (DMLS), selective laser melting (SLM), and electron beam melting (EBM) are all included—and other forms of 3D printing of metals started to kick off. One method that has certainly advanced is 3D printing of materials that can be metal sintered. These technologies made it possible to fabricate fully metal parts with cavities, posts, and other interesting geometries. Even metals useful for RF manufacturing—such as aluminum, stainless steel, copper and brass—can be printed with these methods. Though promising, surface finish is a challenge for many of these methods, and secondary manufacturing steps are often needed to refine the parts before they are viable.

Figure 1 The end-to-end additive manufacturing solutions comprising materials, 3D printing technology, software, and applications expertise now produce the components required for the large antenna array on satellites. Source: 3D Systems

In the last few years, commercially available systems that can 3D print useful RF dielectrics and conductors with the resolution, precision, and reliability needed for parts to operate in the tens of gigahertz have emerged. It’s pretty huge for satellite communications and aerospace applications where the need to reduce part count, size, weight, and manufacturing complexity are high on the bucket list. This is likely why NASA and the ESA have sponsored projects to prototype 3D printed RF parts, and why many prime military contractors are releasing tidbits on their RF 3D printing technology partnerships and capabilities.

Figure 2 The first 3D printed RF filter tested and validated for use in commercial satellite communications reduces weight by 50% over the previous design. Source: 3D Systems

Moreover, there have been some exciting developments with 3D printable resins and 3D printing systems that can print these complex resins with embedded materials. Some of these 3D printed RF dielectric systems have low enough Dk and tand—also known as dissipation factor—to yield RF lenses and other dielectric structures that can operate at W-band. Furthermore, these materials can be metal plated, deposited, or otherwise metallized to yield complete passive RF components such as filters, couplers, and dividers/combiners. Another recent advent is 3D printable ceramics that are suitable for RF applications. Technical ceramics with good RF properties open the doors to much higher temperature RF passives, and in some cases, more compact RF features as technical ceramics typically have higher Dk, better thermal conductivity than resins, and still have relatively low dissipation factor.

I would argue that these new 3D printing systems and materials make it possible to fabricate viable RF passives, assemblies, and other parts that are capable of tackling many of the challenges in common RF applications. For instance, in some cases, metallized 3D printed dielectric structures can be used in place of full metal structures for resonators, housings, and transmission lines/waveguides without taking a substantial hit on performance. The question now is just which companies are agile enough to embrace this new technology and brave enough to pitch it to potential customers.

This article was originally published on Planet Analog.

Jean-Jaques (JJ) DeLisle, an electrical engineering graduate (MS) from Rochester Institute of Technology, has a diverse background in analog and RF R&D, as well as technical writing/editing for engineering design publications. He writes about analog and RF for Planet Analog.


Leave a comment