Dan Beeker Discusses Field Physics and Ralph Morrison
We met with Dan Beeker, Senior Principal Engineer at NXP Semiconductors, during PCB West 2019 to discuss field physics and the legacy of Ralph Morrison.
Dan Beeker – The Interview
Can you tell us about your background?
This is my 39th year in electronics. I started out working for Motorola in 1980, working on 6,800 microprocessors. So I’ve basically grown up with the technology as things became faster and faster, and was forced to learn the methods to achieve EMC and signal integrity.
The most fortunate event happened when I was basically strongly encouraged by one of my friends to come to this conference, and I took my first training class with Rick Hartley and I knew very quickly that he was right. I needed the training and that everything I’d ever designed before only worked by accident rather than by my plan. So that’s not a very happy place to be as a senior engineer.
As a result of that, I started on a journey where I wanted to learn more about signal integrity and EMC design rules, and through that, I met Ralph Morrison, who’s become my mentor. He basically adopted me as his apprentice and he was one of the pioneers in EMC and high performance for signal integrity. He started as a radio repair technician in the Army during World War II. When he came out of the war, he went to Caltech and got his degree in physics. So he became an electromagnetic field physicist and then started working in the electronics industry, designing instrumentation where he had signals that were basically millivolts, so very small signals over miles of wire, using transformers and vacuum tubes. They didn’t have silicone back then. So that led him to write the first textbook, which was called Grounding and Shielding in 1967, and that was sort of the foundation texts that most of the current generation and the previous generation of EMC specialists started cutting their teeth on, that first book. And Ralph, over the course of his life, wrote 13 books with over 2 million in sales.
Mr. Morrison passed away on August 2nd, so we’ve definitely lost somebody that was a very important contributor to the industry and probably one of the least known and least appreciated. His legacy lives on in me and several of the other instructors here. What he taught me were some extremely simple rules that if you follow them, you’re going to have designs that are going to work, and that’s really what’s important.
How did Ralph Morrison influence your life?
We had a conference where there was a panel and my job was to come up with panelists to talk about the challenges of signal integrity. That particular week that we had our conference, there was also another conference in the United Kingdom and most of the big names in EMC were all going to that conference. So I called Howard Johnson and other people and they were all going to this other conference. I was taking books off my shelf and calling the authors.
When I got to Ralph’s book I called and there was a voice message, so I left my name and he called me back and said, “Who are you and why are you calling me?” Because he’d been retired for a bit at that time already. And I said, “Well, I’m with Freescale,” at the time, “and we’re going to have this panel and I’d really like you to be one of my experts.” He agreed to do that, so he came to the conference. We had several other signal integrity specialists and materials people but by the time they … He went first. So Ralph was first, he gave the foundation of the electronics and talked about fields and spaces, which is really important. And the rest of the people talked about current and things that are circuit theory. And when they got done … and Ralph was in his ’80s at this point already … here’s this little, old man jumping up and down going, “They’ve got it wrong. They’ve got it wrong. That’s not how it works.”
And so, I went up to him and said, “Okay, let’s calm down a bit and let’s go have lunch.” And we started talking and he was telling me the story of physics and how things really worked. And I said, “Okay, that’s something I think is really important. You’re right. What you’re saying is different than what everybody else is saying, and it explains some of the things that were confusing me.” So I had him come to my environment where we have automotive customers and we would have conferences or seminars with Ralph teaching it. Everybody would come. I would have to get an auditorium at a local university in order to hold all the people.
And from that, we became friends and then he adopted me as his apprentice because he was going to teach me field physics or he was going to kill me doing it. He worked really hard to convert my thinking from the current flow and the basic understanding of circuit theory into understanding that energy flow was fields in a space or a dielectric. That took him awhile. It took more than six years of coming to PCB West and spending time with Ralph and finally getting the picture. And from that, I’ve come up with some simple rules of electronics. It’s been presented in such a manner that people feel that it’s a complex, complicated subject, but it really boils down to some very simple things.
What are these simple things to store field?
There are only three things required to store field. You have two conductors that are separated by a dielectric. Very simple. Laws of three. You can only do three things with field. You can store it, you can move it or you can convert it to kinetic energy. There is nothing else you can do with it. It either is moving something or it’s heating something up by making the molecules vibrate, and lights will be emitting that. You only get three components. You get a conductor, a dielectric and a switch. With those components, you build even the most complicated electronic control systems. When you turn on a switch, you’ve added a space that doesn’t have any energy in it and the energy behind the switch moves forward into that new space just like water going into a hose.
If you know those rules and you design systems with those guidelines, you’re going to find that you have compliance in almost every case. And that’s from what Ralph told me. His saying was, “Buildings have halls and walls. People travel in the halls, not the walls. Energy and signals travel in the spaces, not in the traces.” So it’s all about the energy moving in a space, and that was hard for me to accept. Geometry connected to electronics just didn’t compute. I wasn’t an RF engineer, it shouldn’t matter, but it did. Everything’s RF now and you really have to think about that movement of energy, and that changes the whole design philosophy.
Can you give us an example of a problem with signal integrity you can solve with Ralph Morrison’s teaching?
What I see typically is the foundation of the board is compromised, is the board stack, where you put your signal copper relative to the ground copper. And in Ralph’s teachings, you always have the signal copper directly next to the ground copper, so one dielectric spacing. So it’s either a layer of signals over a ground plane or it’s a signal trace next to a ground trace on the same layer where the dielectric is separating them. And where people run into problems with both EMC and signal integrity is they fail to follow that rule.
Often I see people will have a board stack-up where there are one or two signal layers separating the power layer from the ground. That means that everything in the space between the power conductor and the ground conductor are part of the power supply. It doesn’t know the difference between a signal trace and power. The field just knows it needs two conductors, and if there’s energy in that space and you’re trying to pull energy from that space, it’s going to come from every signal in there as well. So you’ll see increased noise on the signal lines and you’ll also see increased noise in the power supply as a result of the signals changing their values, and the foundation fixes it all. Going back to the traces and spaces and the energy, you put the energy in the place where it needs to be and that’s where the signal integrity problems start to disappear.
How can we learn more about Ralph Morrison’s teaching?
We were able to put together a website for Ralph, ralphmorrison.com. On that, we had listed all of his publications and you can get most of them from Wiley or on Amazon or whatever, and a little bit of his history. But also, about a year and a half ago … well, two years now, I was able to get him to come to my location and we hired a videographer to have him record three two-hour sessions where he was actually teaching his latest book, and those are available for people to watch and learn directly from Ralph. I think that’s something that’s going to provide at least some lasting legacy for him because he really avoided that.
He didn’t want anybody to video him unless it was done professionally. So I assured him that we would do it and he would have all editing rights, and so he was pleased with the results and I think it’s going to be something that generations are going to be able to take advantage of. Because the laws of physics, as he said many times, they didn’t change. They were the same when he was doing this when he was young, they’re the same now. It doesn’t matter whether you’re working on high energy transmission lines for AC power from the generators to the home or if you’re working on an IC with geometries that are in the 14 nanometers or smaller. The rules are the same. The beauty of the physics is it’s ratio metric. So as you change the size of the transistors, they go faster. When they go faster, the structures that are needed to manage these energies correctly change in their geometry. It’s very simple math. It’s algebra.
DOWNLOAD OUR PCB TRANSMISSION LINES eBOOK:
What is the biggest problem you see today with signal integrity?
The biggest problem still is the foundation or philosophy they use. When you think of the spaces, then you make good choices in how you design it. But a lot of people in their processes are still about connecting the wires and there’s not enough thought about what does the space look like? You have to create a bounded space for the fields to be in. And the problem is that this isn’t taught very clearly in the universities. So people aren’t learning this connection between the physics and the electronic behavior, and that’s leading people to come into the industry without knowing the right things to do.
That’s compounded by the fact that there are plenty of engineers in senior positions like I was, who didn’t believe in the field energy either. And so, you move to a location where there’s senior engineers who still don’t fully understand the importance of managing fields in spaces and you bring fresh engineers into the situation who may remember a little bit of their field physics but probably don’t want anything to do with it, and the result has been failure, EMC. The status quo in the industry has become, “We expect to fail EMC.”
I’ve got customers at three, four, five spends for their circuit board. That’s not engineering. And each time they redo the design, they do it in an expedited mode so they have higher costs than they did the first pass. They’re wasting time that they could be using to develop new products, cost thousands and thousands of dollars and delays them getting to marketplace. And each time they do the design review, they’re not confident that it’s going to pass when they send it back for testing. So this whole problem is universal throughout the entire industry worldwide, that failures become the status quo. What we have to do is find a way to change that.
There are conferences like PCB West where you can go here and you’ll find out the right methods to do this. If you take what you learn at PCB West and go back and apply those methods in your designs, they’re going to start to see more and more first-time compliance. But as it is now, most companies aren’t willing to send their engineers to get additional training because they’d rather pay for them to redo designs three or four times. It’s really a sad thing that happened to our industry. Our rules were generated based on things that didn’t really make things break but weren’t normally or necessarily correct. And that was because the geometry of the integrated circuits was so large relative to the physical PC boards and systems they were used in that the things they did wrong weren’t so bad they caused them not to function.
At this modern age where we have geometries that are extremely small and signals that are switching in tens of gigahertz, you can’t ignore the laws of physics anymore. So the methods have to change to match the reality. We ignored that as an industry and still do. There’s no demand from the consumers of engineers, the companies that hire them to the universities, to change their curriculum so they can address this problem. And until somebody asks them to, they’re not going to change what they’re doing. My customers would rather pay to redo a board than for their engineers to be … to pay for them to go to training, and that has to change.
Are any software companies working with you to correct their simulation tools?
I don’t know of any hardware or software design company that has, whether PC board design or even integrated circuit design tools that are looking at a simulation tool that will accurately model the AC behavior of the transmission lines and the switches. They look at circuit. They’re looking at resistance, inductance and capacitance. That doesn’t help you when you’re trying to design a system where you’re moving energy from one place to the other.
They don’t tell how large the space is because they don’t know where the bounding conductor is, because it takes two conductors to manage the signal. They don’t care how far it is between switches, so from transistor to transistor, so you have no idea how much energy it takes to fill that space or how long it takes to move through that space. That’s absolutely critical. You have to know those things to do a good design. It’s the same on a PC board as it is on an IC device. Nobody’s working on it. It takes a full … like a finite element analysis tool to be able to have all of the data necessary to automate the evaluation of the transmission lines on the circuit board.
As an engineer, how do you help the industry follow Ralph Morrison’s teaching?
This is one of the most powerful ones that I found. I try to come here as an instructor every chance I get. Within my corporation, I teach classes at our customer events when I can. I teach classes to my customers whenever I can. There are several conferences now. Sierra was great about hosting a seminar for me last month and we had to get attendance and I think that the teachings are well received. It’s just very difficult for most single engineers who are at a conference to go back to a team of four or five other ones and have them be willing to accept the change in philosophy. It takes a while. It took me a long time.
Can you give us a specific example of what engineers do wrong?
The biggest mistakes are with the basic board stack-up, that if you don’t have a solid ground plane adjacent to each of the signal layers, that you’re going to make the design more difficult. You can make transmission lines with ground next to a signal on a single layer. It’s not the best transmission line but it still works as a transmission line. But people will still throw two or three signal layers against a ground plane and expect the current to flow in the way they think it is. The old rules are not being tossed out yet, and if you change the board stack philosophy, you have a chance of making a board that’s going to be compliant.
What would you recommend engineers to do to get more knowledge?
Sierra Circuits has been sponsoring several of the signal integrity and EMC specialists in seminars around the country. If you get a chance, those are really excellent ways to get the important knowledge that you need to be successful. A nanometer geometry is switching at a very fast speed. So everything now is RF and you have to worry about designing a space.