Reading time ( words)
I recently ran into Dr. Bruce Archambeault at PCB Carolina in Raleigh, North Carolina. He’s retired from IBM now, but as he explains, he’s still teaching.
He taught a class at the show that focused on layout considerations and inductance, and why inductance needs to be measured. There weren’t many empty seats in his class.
I asked Bruce to discuss his class and why it’s so importance for designers and engineers to measure inductance. Bruce also discusses how he’s linked the study of inductance to the “Dirty Harry” movie franchise.
Andy Shaughnessy: Hi, Bruce. It’s been a while. I caught part of your class on layout considerations. You really had quite a crowd in there. Tell us about the class.
Bruce Archambeault: Yes, we were doing a presentation as well as a demonstration. We have a series of circuit boards that were made to demonstrate fundamental physics, things like inductance. Most of what we focused on today was inductance—the inductance of filters, connections, conductance of traces between the capacitor and the ground pad, and a number of things like that. It showed very clearly that the loop area of a wire is much more important than the wire size. The wire size made almost no difference at all. It was eye-opening for a lot of people.
Shaughnessy: So, just by proper layout and good layout practices, you can help ameliorate a lot of the problems down the road.
Archambeault: Right. The parasitic conductance has been referred to as a hidden schematic and that’s true because when you draw your schematic, you don’t draw into parasitics. You don’t even know what they are. But it can really change how things go. You think you’ve got a capacitor that’s filtered at certain frequency, but then find out it’s not at all because it’s the inductance of the connection. One of the things we demonstrated was that, if you’ve got a decoupling capacitor and the planes are right up on top, then you have a very small amount of connection inductance.
But if you’re going down into the board, even a 60-mil board, you go halfway down and that adds quite a bit of inductance—a lot more than the manufacturer's equivalent series inductance. They might claim half a nanohenry or 0.3 nanohenrys connected onto the board and you’re getting two or more nanohenrys. That’s the real thing people don’t learn in school, the concept of inductance. They don’t teach it, so I focused today primarily on inductance. I like to say inductance is everywhere. On the first day, Genesis tells us that God created light. Well before he did that, he created inductance and said, “This will really freak them out.”
Shaughnessy: That’s good. And I like what you said: “Inductance is everywhere.” And everyone enjoyed your mention of Dirty Harry’s line, “Do you feel lucky?”
Archambeault: Right. If you don’t consider it, then you better be counting on luck.
Shaughnessy: Did you hear any interesting comments or questions from the attendees?
Archambeault: The most interesting one was from a guy who said he had seen an article published back in the ‘70s where the return currents would take the shortest path. Well, that’s only true with low frequencies. At high frequencies, it takes the longer path, because that’s the least inductance and the return currents are right underneath the trace. A lot of times people don’t understand that, so that was interesting. Other than that, it was pretty much us trying to do the demonstration and explain what people were seeing. Because we’re using a network analyzer, and unless you're familiar with them, it’s not always rolling off a log easy to figure out what it means.
Shaughnessy: It seems like a lot of this just comes down to solid board design, board layout, and proper stackup.
Archambeault: It does. As an EMC engineer. I almost don’t care about the schematic. I mean, I’ll review it if someone asks me to, but there’s very little to see because the actual circuit board layout is what makes a difference. This trace is too close to the edge of the board, or to the decoupling capacitors, or you have too much trace. It’s things like that that I really care about.
Shaughnessy: It’s almost an art. We talk about how it’s art and science. You have to know the science, but you must have an artistic eye to apply it.
Archambeault: That’s true. There are some tools out there that can help you. CST sells a board check tool. That’ll read in an Allegro file or who’s ever file. It’ll check it against a bunch of EMC rules. And then it’ll warn you that this is violating, you should fix this. But not all that many people use those tools.
Shaughnessy: So, are you really retired now?
Archambeault: I’m retired, but I’m still teaching classes. Now that COVID has settled down a little bit, I’m hoping to be able to do some in-person classes. The last couple of years it’s been only by computer. I do a couple classes a year. Until COVID hit, I was teaching at Oxford University every summer.
Shaughnessy: I heard. That sounds like a good gig.
Archambeault: Yes, I go over there and teach the Brits.
Shaughnessy: Do they ask you about aluminium?
Archambeault: And they spell color wrong.
Shaughnessy: Brilliant! So, I’ve noticed that there’s a real thirst for knowledge about this content that’s not just your basic board-level design stuff. It’s more about field theory, electromagnetic theory, things like that. I see lots of veteran designers and EEs paying more attention to fields now. Are you seeing that? And why is field theory just now really kicking in with our segment?
Archambeault: Well, for a lot of reasons. The way they teach electromagnetics at the university, unless you’re a mathematician, is going to put you off electromagnetics. But it’s really not that hard. If you want to solve something, the math can get messy. But we have computers for that, so I don’t think it’s that hard to understand the concepts.
Shaughnessy: You don’t have to be a mathematician to design for EMC.
Shaughnessy: Well. thank you, Bruce. Great talking with you.
Archambeault: Good seeing you too, Andy.