Rogers’ John Coonrod on Insertion Loss
John Coonrod of Rogers Corporation gave a keynote presentation at the recent Geek-A-Palooza trade show, concentrating on printed circuit board fabrication’s influences on insertion loss. I sat down with John to learn more about his presentation and what OEMs and designers need to be aware of to avoid insertion loss.
Barry Matties: John, why don't we start with a little background about what you do and where you’re from?
John Coonrod: I'm part of Advanced Connectivity Solutions at Rogers Corporation. We generally make high-frequency materials normally used in PCBs—specialty boards for the high-frequency microwave and millimeter wave range of frequencies. I do electrical characterization on our materials or competitor materials, if I can get my hands on them (laughs), and I also do evaluations on any kind of new materials that we are developing for electrical performance.
Matties: Nice, and you've been at Rogers for many years?
Coonrod: A long time, since 1987.
Matties: We are here at Geek-A-Palooza and you're doing a keynote. What is the topic of your presentation today?
Coonrod: The topic is insertion loss, and more specifically printed circuit board fabrication’s influences on insertion loss. Whenever you build a printed circuit board, just due to the nature of building the board, you will have some insertion loss variables. Really, what I'm going to be showing today are some of the studies I've done looking at that, and the effects of different PCB fabrication processes on insertion loss.
Matties: And in this presentation, what's going to be the takeaway for your audience?
Coonrod: More than anything else, I'm hoping this is going to be educational, because I start off with some really basic ideas about insertion loss and why it is important, and then I get into how these different PCB fabrication influences can have an effect. One of the reasons I'm doing this is for people who may not know much about insertion loss, and that way they get a good introduction and overview of it. Then the other idea, I'm hoping, is that people who do know something about insertion loss and do circuit designs will now know more about some of these variables that they may not have known about before.
Because I've run into that from time to time, where an OEM or someone will come to us and say, "Hey, your material is not working right," and then we find out it's not really the material, it's something special or something funny that happened between the design and the fabrication. The more the designers know about these things the better, I think.
Matties: Communication is key. Could you give a quick overview of insertion loss for those of our readers that may not know anything about it?
Coonrod: Insertion loss, in an RF sense and in a frequency sense, is really the total loss of the circuit. The way to think about it is when you apply RF power on one end of the circuit, and you try to take power off the other end, you don't get the same amount of power you put into it because there's some loss there, and that's really what the circuit is doing. It's kind of a complicated issue where there are several other things that come into play, but it's really just how much loss the circuit really causes to an applied RF signal.
Matties: Is the variable the design, primarily?
Coonrod: That's definitely part of it. There's really a big interaction between three important things: the materials, the design and the PCB fabrication. And that’s because, for one particular design, the PCB fabricator can build it multiple different ways and one way may impact the insertion loss differently than another way. So that is kind of my thought process here, and what I’ll be trying to show are some of these variables.
Matties: You've talked about different manufacturing processes. How does an OEM or buyer of that technology verify this before they get into a problem situation?
Coonrod: That's a good question, actually. Because normally an OEM designs a board and sends it off for the fabricator to build it, and they put on there what they are interested in, of course, for conductor widths, what kind of finish they want and things like that. That gives the PCB fabricator a lot of room to do a lot of different things. Normally, the OEMs don't specify some of the things that I'm going to be talking about today, and in some ways I hope what I'm going to talk about will be enlightening for the OEMs, but also for the PCB fabricators too—that they need to be more aware of how they build the boards sometimes for the OEMs. I'm not sure if that really answers your question or not though... (laughs)
Matties: Well, it sounds like there's no real answer.
Coonrod: No, not really.
Matties: Other than that they can be aware of the problem. If the OEM is aware of the nature of the problem is there a specification where they can come in and say, "We want it manufactured like this"?
Coonrod: There are some cases, as you will see when I go through my presentation, where they can put a spec on things and say, "Okay, it has to be within this window," and then that takes care of the problem. In other cases it is the nature of the beast, and either the OEM has to realize that their design is going to have this wider range than they expected or they have to re-design and do something different.
Matties: Is this a prevalent problem in the industry?
Coonrod: It's a pretty big deal, I think. Really what happens is a lot of the fabricators that have been in the industry a long time have learned these lessons along the way, so they are already doing a lot of this automatically and you just don’t run into this problem with some of the really experienced people. Then, in some cases, the OEMs are doing things that they haven't tried to do before because they are pushing the limits of technology and they're trying do different things with materials and processes that haven't be done before. So every now and then when you get the right combination, unexpected things can happen.
Matties: What does a designer need to be aware of in this process?
Coonrod: What I'm going to show today is about four major topics in the PCB fabrication process that can affect the performance of the RF board. One of them is copper plating thickness, because it's very common, obviously, to plate copper through the circuit. But copper plating thickness variation, how much it varies from one circuit to another, has a pretty big role on one type of design but not as much on another type of design. So I kind of split that up today, showing that copper plating thickness is more important with some designs than in other designs.
Then there is the plated finish, which is used a lot, where they put electroless nickel immersion gold (ENIG) on copper in order to have a good reliable board for a long period of time. ENIG is a good process, but there are some things that it does that causes more insertion loss. But with certain designs it causes more insertion loss than in other designs, so that's another thing where there are certain designs that can use ENIG and not have as much loss as others. The other two topics are copper surface roughness, which is a really big topic and I’m just going to scratch the surface today, and then the last one is solder mask, and that is another topic that comes up from time to time.
Matties: It sounds like once you find the right recipe, stick with it.
Coonrod: Yes, that is true. I think that's how lots of engineers are. Once you've got something that works, don't mess with it.
Matties: But isn't that a trap?
Coonrod: It can be, yes. Especially if you are trying to deal with new technologies. A lot of times when you're breaking out a new technology you have to do something different and you have no choice but to go out on a limb. And that's kind of what some of this presentation is. I'm really sharing my learning experiences of different things I've heard and seen in the industry or helped troubleshoot, and just trying to make sure everyone is aware of these potential hazards.
Matties: John, thanks for spending time with us. I look forward to your presentation.
Coonrod: Great, thank you.
