Ask the ASC Expert Ask the ASC Expert: RF/Microwave decoupling

Written by: John Bushie on January 27, 2019

< Back

In the first of a series of Ask the Expert video blog posts, John Bushie, ASC’s Director of Technology, explains the basic principles of decoupling in RF/Microwave PCBs. If you’re designing for RF/Microwave, the knowledge you gain by watching this 5-minute video could save you hours in your next design cycle.

Video transcript:

SW: Hi, I'm Steve Williams with TRAC and I'm here today with John Bushie, director of technology for American standard circuits. Hey John, thanks for making the time.

JB: My pleasure, Steve.

SW: Alright, so we're going to do a little tech talk here, right. So rf microwave, that's your area of expertise, one of your areas, right? So, one of the things that comes up a lot is micro strip versus strip line and the whole coupling thing. Can you explain it down to a level that somebody like me can even understand?

JB: Sure Steve. Basically, when we start talking about couplers, we talk about a method of being able to sample a signal without contacting the trace directly, basically passively taking a sample of it to either control the power level or to control some other aspect of the performance of a device, like let's say a power amplifier. In general, we have two basic types; we have a micro strip, which is basically on the top side of what is generally a double sided or it could be a multilayer circuit board. And then the other type that we have is a strip line, where the lines are basically buried internally. The major difference is how the lines couple, when we talk about micro strip lines, couple basically at the edges, the critical parameters and those types of designs is how high are your traces and how far apart are they spaced.

SW: So John, when you say couple, what exactly is coupling?

JB: Basically, electromagnetically these lines are coupling one with the other and the other one is basically sampling a portion of the energy based on that.

SW: Got It.

JB: So that's basically what the term itself means. In this particular case, I show an example of some of the critical aspects, obviously the type of dielectric material that the circuitry is placed on, the thickness of that material, the circuits, proximity to the adjoining ground plane as well as their distance and height relative to each other. These are some of the aspects that we have to control fairly tightly. And I don't know if you can pick it up on the video, but I have basically shown the difference between what is an ideal circuit ,or I will call it a theoretical, a trace profile and what the reality is, and these are all things that we need to take into account and the designers should take into account when they're trying to design a coupler for a specific band or power level.

JB: When we go over to a strip line, it gets a little bit more complex. We're definitely talking about a multilayer structure, the traces that are coupling our broad side coupling in that they're now positioned above or above and below each other. In this particular instance, we've got three things that were controlling from a thickness standpoint, aside from the circuit height itself, we've got three independent dialectics. There's the dielectric that would basically separate these two, what could be an individual double-sided circuit board. And then we've got their locational position relative to one another. So, controlling the dielectric constant and these features as well as controlling the width, the height and the distance and the position of offset relative to each other. All those things have an impact on how they couple, what frequency they couple at, how much power they can actually a passively couple.

JB: The key aspects for us is our ability to be able to control the line widths and the positional accuracies are not such a big deal when we're talking about a small device. A lot of these tend to be relatively small in the half inch by three quarter, some as large as several inches square. The bigger issue for us is when we're doing it over an entire panel. Some of these are manufactured regularly on a 12 by 18 or an 18 by 24. So, when we start talking about the critical alignment, particularly on a strip line structure, we've got to be able to control the position of these on two separate, either two separate cores or ideally on the same core, buried internally on this type of a multilayer structure. See, it's not just important how well they line up over the top of each other in the center of the panel, it's how well did they do it on the outside edges and corners of what is a relatively large manufacturing panel.


SW: Excellent. One last question. Is there anything that you would tell a designer to avoid when you're talking about this type of coupling?

JB: When we're talking about micro strip, we need to understand exactly what our finished plated thicknesses and a lot of designers start off with the base coppers and think that that's going to be their ultimate yield. So, when they're trying to figure out what the height of these actual traces is going to be, they need to take into account if there is one, two, or three additional plating operations that may take place.

When talking about strip line, this is actually a worst-case scenario where I've placed these traces on independent cores. Ideally you would make it what would be considered a six-layer structure, whereas these would exist on layer three and four on the inside. That way they're printed on the top and the bottom of the exact same core. That's where your alignment systems in general for imaging will have their best opportunity at alignment. Otherwise you're relying on a core to core registration, or multilayer registration, to be able to achieve what is ultimately a critical alignment. A lot of these aspects, or the critical design parameters, are also important when we start thinking about filter structures very similar in how they function. They obviously have different functions, but the control of the same types of parameters are also critical in those designs as well.

SW: Great. Well, Hey, thanks John. I think it makes sense to me, which means that it probably makes more sense to anybody watching this video, so I appreciate your time.

JB: My pleasure, Steve, thanks.


This column was originally published by