Reading time ( words)
RF is becoming almost ubiquitous; how many devices in your home contain at least one antenna? Automotive, aerospace, defense, and IoT segments are all pushing the envelope for wireless communication. But designing an RF PCB is a lot different than designing typical boards.
We recently met with IPC instructor Kris Moyer for a discussion about designing RF PCBs and wireless applications in general. Kris teaches RF design, among other things, so we asked him to discuss RF design techniques, how designing for wireless applications differs from laying out traditional PCBs, and when to design your own antenna vs. using commercial off-the-shelf (COTS) antennas.
Andy Shaughnessy: Kris, tell us about the RF and wireless PCB market. How does it compare to the rest of the market for PCBs?
Kris Moyer: It's never huge, but with the increase of smart devices and IoT devices, it’s definitely growing. It’s not easy to design. Think about your modern cellphone, or your modern smart watch that has GPS, a Wi-Fi antenna, a cellular antenna, and a Bluetooth antenna.
Happy Holden: Now there’s wireless charging.
Moyer: Right. So, there are five different RF frequencies and modules running simultaneously on one board. There are probably more than just five that we're not aware of in terms of what the military, NASA, and other organizations like SpaceX are doing with hundreds of little satellites. The big question is: How do you get all of that to work together and not interfere with each other?
Shaughnessy: Our readers have a lot of questions about antennas and antenna design.
Moyer: This means not only designing the antenna design itself, but also the Bluetooth and Wi-Fi protocols, FCC compliance, and the system-level design to integrate all these different communication methods. You have all kinds of different IEEE protocols involved with that, and how they may interact or interfere with each other. How do you deal with SMA connectors, controlled impedances, wave cavities, and so on?
Back in the radio days, it was literally just an amplifier, a transmitter, and a receiver. The design of the antenna was critical because it controlled your frequency and so on. That's why we used to have TV antennas back in the day, especially the old TV antennas that were triangle-shaped. They used the triangle shape because each of those lengths of antenna picked up a different frequency. Nowadays, it's not only the design of the antenna, but it's which chip you need to use. They actually sell not only 3D antennas, but pre-packaged antennas in chip packages. You can buy a Bluetooth antenna in the equivalent of, say, a 1206 or a chip package, and just solder it onto the board.
Shaughnessy: It seems like the easiest way forward would be to use an off-the-shelf antenna, since it’s been validated, I assume, and that preliminary work has been done.
Moyer: There’s more math involved in designing your own antenna, which is technically a “copper geometry device.” If you're trying to design the antenna out of the board trace itself, there is a lot of engineering mathematics and analysis that goes into that. Specifically, to get the antenna to work right, you have to convert to the SAP or mSAP process so you don't get non-ideal geometries that you get from the standard subtractive process. You must be much more diligent about the width and the clearance, especially if you're serpentining your antenna rather than making one long straight trace, to save board space.
With a serpentine pattern, now the gap width and the length before the turns will be much more critical than it is with just the simple serpentining I would do for length matching. There's a lot more mathematics involved if you're trying to design the actual antenna. If, on the other hand, all you're doing is buying parts and trying to connect them, you're still into RF board routing techniques, but without the complex mathematics of antenna design.
To read this entire interview, which appeared in the April 2023 issue of Design007 Magazine, click here.
