Rockstar PCB Designer and SI Guru Rick Hartley
Table of Contents
Warner: Welcome back, Rick. The first time we met, we were ranting about how we both wished more designers would spend some time visiting their board suppliers and learn more about how boards are manufactured. What other pieces of advice would you give to an engineer or PCB designer today?
Hartley: It's interesting you say that. Back in the '77-’78 time frame, when I first became a printed, one of the fabricators that our company used sat me and the other printed in the company down and said, "We understand that the engineers in the company love you guys because you're both engineers who chose to become printed designers,” and we smiled and nodded knowingly like, "Yes, we're wonderful people, we know that.” Then they looked at us and said, "You guys may understand circuits and may understand how they work, but you know nothing about manufacturing. You people are so clueless, it's amazing. Everything you design is impossible for us to build." My ego fell through the floor. It didn't fall to the floor, it went into the floor! It literally took me a week to recover from that. But once I thought about it, I realized, they're right. I don't know anything about manufacturing. We made a conscious decision as a company to hire Norm Einarson. Anybody reading this interview who is an old-timer will recognize the name and they're smiling right now as they're reading because Norm was a PCB god back in the day. He was a manufacturer who understood design. He came to our company for a week. Attendance was mandatory for everybody in engineering, EEs included, to sit through this entire week of training. He taught us how we affect manufacturing and how manufacturing affects us.
Warner: What a great and unique opportunity for all of you!
Hartley: It was the best week of training I've ever had in my life. Ever since then, I've subscribed to multiple manufacturing magazines because I want to know what the manufacturing world is doing differently than they did last year, what's new, what's coming. I can't design circuit boards correctly unless I design to the processes they're using.
Warner: I know a lot of fabricators that wish more designers took that proactive approach. It makes perfect sense.
Hartley: It does, doesn't it? And that's one of the things I would say, if you're an EE or a printed circuit designer; if you're involved in any way in the decisions that go into designing the printed circuit boards, you need to understand the manufacturing processes. You need to understand board fabrication, board assembly, and testing, and what impact you have on all three disciplines.
Warner: I couldn't agree more. It resonates with me because I come from the manufacturing and assembly side of the electronics industry. I consistently experienced issues and delays that could have been so easily avoided if we would have just had a conversation before things became problematic.
Hartley: Another suggestion would be for PCB designers and EEs who work together to get closer. Get to know what each other does. Engineers need to understand how complex board design really is because I don't think a lot of them do. They both need to understand each other’s disciplines deeply enough to function well together. If they don’t, then they're in a toss-it-over-the-wall environment where the engineer says, “Here's your schematic, go do your thing.” That doesn't work well-- I learned that years ago. Most of the engineers I worked with at L-3 and in the telecom world, and even going back into the late seventies and eighties, worked well with me and I tried to always work well with them. I guarantee if any of them read this article when it's published, they will be nodding and saying, "Yes, we did work well together." That’s because we understood each other. And that's important.
Warner: Which appears to be rare these days. When they don't work together, does it create some tension between those two disciplines?
Hartley: Of course, it does. It becomes an ‘us and them’ environment, and that's never healthy.
Warner: Absolutely not. So, a third piece of advice, Rick?
Hartley: Don't do anything in the way of printed circuit layout that's stated on an IC application note.
Warner: You said that at IPC APEX this year, which made me laugh!
Hartley: It's true. I start every class I teach with a quote from Lee Ritchey, a quote he made in his class in 1993. When he said it, I giggled, as I thought I was the only one in the world who had that attitude. I was so glad to hear a really knowledgeable guy make that statement. Lee’s quote was, “IC App Notes should be Assumed Wrong until Proven Right.” He is not saying they are wrong, rather that enough of them are wrong that you are wise to NOT assume they are correct. Dan Beeker, a friend and an app engineer at NXP / QualComm, believes that deeply. He tells people in his classes, "When you read a piece of circuit advice in an app note, it’s probably right. But app Engineers typically don't understand board design, and they make statements in app notes that are not based in solid physics, like, split the ground plane into little pieces and put one piece here, another here and one of them over there. Put decoupling in places where it absolutely makes no sense. Don’t use 90 degree corners. They make these statements that are just not based in solid physics. And I'm not saying all app engineers are wrong, I'm saying that many of the app notes that are out there are just very poorly done with regard to circuit board layout advice. Be cautious about what you believe!
Warner: That's funny. I remember you put that quote up at your class and I chuckled. It seems very counterintuitive since we all rely on data sheets.
Hartley: Well, the information in data sheets is generally correct because they give you timing figures, they give you max current draw, and these sorts of things. Those are generally good pieces of data. One thing I wish data sheets would publish is rise and fall time. One of the most important things engineers and designers need to know to make intelligent circuit decisions is the rise and fall time. Truth is, the clock frequency, the frequency at which the clock is driven in a circuit, has very little, if anything, to do with the frequencies where things go wrong. It has almost nothing to do with EMI problems and little to do with signal integrity problems. Signal integrity and EMI problems are related much more to rise and fall time. The problem is, rise and fall time are a bit nebulous. They're not easy to pin down. But if they would at least publish the min and max rise and fall times, the worst-case conditions, so engineers and designers would know what to expect from any given IC when they're doing a layout, when they're designing the circuit and the PC board, and planning for what frequencies are important.
Warner: Why do you think they don’t publish it on data sheets?
Hartley: Because it is hard to pin down.
Warner: Is that due to environmental impacts, or what makes it nebulous?
Hartley: Temperature changes impact rise and fall time, as do loading and other factors. It will either speed up or slow down, depending on conditions. Loss factors in the board also impact rise and fall time. The information is available in IBIS and SPICE models. Models are in ASCII format, easily readable. When you use the model in a simulation tool, like Hyperlynx, for example, that model has to have rise and fall time to accurately determine how transmission lines will behave. So, it's in there, you just have to look for it.
Warner: That's interesting. I didn't know that. Okay, next question: what do you wish EDA tool companies would do differently? Or, what do they do now that either helps or hinders PCB designers today?
Hartley: That's a good question. I think all the EDA tools I've ever used, once I learned them, were fairly powerful. Many years ago, I used a tool called ‘Theda’ from a company named Incases. I found it to be horrendously non-user-friendly. It was also a horrible tool to learn, but once I figured it out, it was quite powerful. I found that to be true of Mentor Board Station as well, hard to learn but powerful.
I found original P-CAD® (Master ) to be powerful, especially since it was the first PC-based tool, but to me it was not user-friendly. Pads PCB came out in 1985. I found it to be a very user-friendly tool right from the get go. All of these tools, once I learned them, were powerful. The problem, they all have vastly different learning curves and user interfaces. Some of them were designed by people who truly understand what printed designers do, some were designed by programmers who know nothing about printed design. The ones that were designed by programmers typically aren't user-friendly because they simply do not grasp what we need. I just wish they would all step back and ask themselves, "Okay, what could we do to improve the user interface and ease of use of our tool?"
In the 1990's Tango purchased , restructured the tool entirely and renamed it Accel. When they were done it functioned more like PADS PCB than like Master . It was so user-friendly, the redesign was masterful. When I went to the telecom world, they had Accel. Having never used it, I sat down with the manual and in three days became completely proficient. Eventually, the tool was renamed 2001 and later 2003, etc. Many of the features from that tool are in Altium today.
Accel was ridiculously easy to learn. Pads I found to be almost equally that easy. Board Station took me six months and I was still scratching my head. Now Board Station's gone, it has been folded into Xpedition under Mentor's domain. They've done a good job. I like the Mentor tools, I like Altium's tools, I like a lot of the tools that exist today but I wish they would all work on their ease of learning and ease of use. I have never used Cadence tools, so cannot comment.
Warner: Makes a lot of sense. What do you think about licensing models or did that ever impact you since you were working for large OEMs who managed tool acquisition?
Hartley: At L-3, the IT guys took care of all the licensing issues. Before that, I took care of my own licensing. Some of the tools were very easy to fire up and use and their license was very easy to install, but some tools were an absolute nightmare. That's another thing that I think many of them need to work on. I know they're trying to protect their proprietary material. I understand that need. But some of these companies have figured out how to do this with easy to use license structures and others have very complex, hard to use licenses and I just wish those companies would think about it a bit more.
Warner: Good advice. Lastly, I know you have been on the executive board of the IPC Designers Council Committee, so would you talk a little bit about how and when you got involved? Will you also please comment on the value and challenges you see for regional designers councils going forward?
Hartley: Sure. I've been on the IPC DC board since 1996 or ‘97, about 20 years. Gary Ferrari recruited me. Gary and I met in '93 at PCB West and we just hit it off. When they were starting the designer council, I was curious about it and I asked him what it would take to start one in our region and he helped me with the information as did other people from IPC. He was very heavily involved back then and was very helpful. Pete Waddell of UP media (Printed Design, Fab and Assembly magazine) was also very helpful. Pete offered to send out our initial letter for recruitment of members to everybody within a 100 or so mile radius of Columbus, OH, basically to all the readers he had in that area.
Warner: Doesn’t Pete have a design background?
Hartley: Pete was a , yes. The original Designers Council was located in Atlanta. Pete was part of that, so he understood the value and the need. Anyhow, I've been involved for a long time. We had a designer chapter in central and southwest Ohio, starting in about the '93-’94 timeframe. Right around the time, the whole thing was taking off. Our designer council was strong for about 12 years. Our model was different than the ones you guys have in San Diego and in Orange County, we basically learned from each other. We had about 30 to 40 regular members. We met every month for about 12 years. Two of those meetings every year were parties, the other ten meetings of the year were educational, we basically taught each other what we knew. We had people in the group who were manufacturers, EEs, and designers, people from all walks of the business.
For one to two meetings each year we would invite an outside speaker to do a one-day seminar. We charged our members a nominal fee, for that matter, anybody could attend. Non-members paid twice as much. Basically, we would charge everybody some small amount so we could pay the speaker a stipend for their expertise and time. After about 12 years, we just ran out of things to tell each other. Our designer council basically fell to the way side and hasn't been in existence since about 2005.
The advantage of a regional designer council is the chance for people to learn. The chance for people who can't afford to go to PCB West, who can't afford or have the time to go to IPC APEX or go out and buy 100 books like I've done. To be able to go to a meeting once every month or two and learn something that might be valuable to them in their ongoing strife in life as an engineer or as a board designer is the real value of a regional designer council. The education is invaluable.
Warner: I have found the networking to be invaluable over the years as well.
Hartley: To me networking is absolutely essential, it is the key to success. There are a lot of people who want to see more online training. I like online training, don't get me wrong, but I don't think there's any substitute for standing in front of a room and letting them see you and you seeing them. And then comes the networking during and after the training. You are right, the value of networking is massive. My career has benefitted so much from the face-to-face contacts I’ve made, it’s hard to measure the value.
Warner: There is no substitute for face-to-face interaction at industry events. Which, by the way, is how I had the good fortune to meet you, Rick!
Hartley: Exactly--and look where we are today.
Warner: Well, I feel very fortunate to know you and I can’t thank you enough for taking the time to share your illustrious history in PCB design and electrical engineering, as well as your hard-won wisdom on a very complex subject.
Hartley: My pleasure, Judy, happy to do it.
Hartley in the early days of PCB design doing mylar tapeups by hand.
Judy Warner: Rick, will you start by giving us a brief history of your background, career, and what kind of products you have designed?
Rick Hartley: I started life in electronics in 1965 at the age of 20. I was a technician for several years in an R & D department, going to school at night, studying electrical engineering. After a while, I moved to field service, spent a couple years there, then moved into an engineering department as a “designer”. “Designers” in those days designed circuit boards, packaging, wire harnesses, cables, anything that was in the product, except for the circuit itself. I spent several years doing that and then, with enough education under my belt, was moved to an EE role where I spent several years designing circuits.
At one point, my boss came to me and said, "Look, you have a history of PCB design, we have a lot on the table in the way of circuit boards. Would you mind splitting your time 50/50 between circuit design and circuit board design for the next six months?" I said, "Sure, I wouldn't mind at all.” So, I did and after six months realized that my real love was in circuit board design, not circuit design. I made a conscious decision to move from circuit design back to circuit board design. A lot of designers, when they heard that, thought I was a little crazy.
Warner: I bet they did! It seems like a counterintuitive career move.
Hartley: Yes, they said, "What a masochist, who is this guy?" But I really enjoy circuit board design more than circuit design. It was a conscious decision. Fortunately, it was a lateral move, they kept me at the same pay scale. When I left and moved on to other companies I was fortunate to be able to continue to be paid at the EE level. For me, it worked out very well.
Warner: What kind of products have you worked on over the years?
Hartley: Products I've worked on over the years ... Most of my early experience was in industrial controls for factory floors that use closed loop feedback systems to control what went into a product. For example, we measured plastics as they were being made. We would control how much resin and how much air went into the mix to make sure it was distributed properly, that it was the right thickness and the right mixture of all materials. So our system basically measured the plastic as it was being made and controlled its flow. We did that with paper, with all sorts of products. Probably the first 17 or so years of my career was in that industry. From there, I moved into the design of computers because by then it was the eighties.
Warner: I remember that time well—it was an explosion of electronics for personal computers and peripherals.
Hartley: I first worked for a company that did ruggedized computers that could be put into all sorts of environments. We had a computer you could drop off the top of a three-floor building and it would still work.
Warner: What were the applications? Like construction sites and such?
Hartley: All sorts of sites where things got bounced around a lot and they wanted them to be able to work even under heavy stress. So, factory floors, military field applications, that sort of thing. From there I moved into avionics in the early nineties, where I spent several years. In the late nineties, I moved into the telecom world, spent five years there and then in 2002, the telecom world, as you know, just collapsed.
Warner: It certainly did.
Hartley: I then moved back to avionics in 2003 and spent the rest of my working career, until 2014, in the avionics world. So, the majority of the products I've worked on have been industrial controls, computers, avionics, and telecom equipment. Now I am semi-retired, teaching and doing some consulting.
Hartley teaches regularly at conferences around the US.
Warner: And the last place you worked full time was at L-3 Communications, right?
Hartley: That's right. It was called L-3 Avionics Systems. We designed and manufactured avionics for both commercial and military aircrafts. They're still thriving today, that's a great company, they're doing really well.
Warner: That's always good to hear these days. When I last saw you at IPC APEX in February, I was sitting in one of your classes and you asked people in the room to raise their hand if they were an EE who also designed boards. Then you asked them which was more challenging, EE work or laying out circuit boards. Why did you ask that question, and why do you think that most of us were surprised that they said PCB layout was more challenging?
Hartley: The first time I ever asked that question was in 2008, at PCB West. I met two people in my class who had been doing both circuit design and PCB design for around 20 years. I asked them: "In your opinion, which is most challenging?” This guy literally scoffed and he said, "Oh, printed circuit design's way harder than circuit design." Every EE in the room turned and looked at him with such incredulousness--you cannot imagine the looks he got!
Warner: Oh, that's funny! What do you think was going through their minds?
Hartley: They were upset, basically these two guys were shooting holes in their paradigm. The reason was because there was a time when circuit board design was truly connecting the dots and it was more art than science. It wasn't very technically challenging in those days. Some engineers believe it is still that way, but over the last 25 to 30 years, printed circuit board design has become a true engineering task. We have always had to solve the maze of placement and routing, while maintaining manufacturability. Today, you need do all that and understand signal integrity and field theory, as well as have some understanding of how circuits work. As you heard me say in that power distribution class, at IPC APEX this year, the energy is in the fields. Once you grasp that, you start to really understand where the energy moves in a circuit. That makes all the difference in the world as to how you should design. People who get the true nature of both professional positions do understand that circuit board design really is the harder of the two jobs. That's why I ask the question and that's why I think most of you were surprised, because you didn't expect to hear the answer they gave. They, just like the two guys in 2008, they said, "Oh yeah, board design's harder."
Warner: And they were clearly speaking from a place of experiential knowledge because they'd been at it for a long time, just like you have. I think most of us might think, if we put design professionals in a hierarchy, based on difficulty, that EEs would be higher in the pecking order.
Hartley: Exactly. EE work requires a deeper level of technical education than board design, but once that knowledge is acquired, daily challenges are greater in the PCB design arena. I do occasionally meet dual functioning people who think that EE work is harder, but they are the exception.
Warner: To your point, PCB design has become so much more complex. Especially when you get into RF, microwave, millimeter-wave, and high-speed digital, it seems to get incredibly complex.
Hartley: Well the reality is, circuit board designers today need to have a little bit of circuit theory under their belt. They need to be able to look at a schematic and know what's going on. They don't need to know the intimate details of the circuit but do need to understand it well enough to know what the processor and FPGA are doing, how memory plays into the circuit, how things are being driven and accessed. These are all items they need to understand at least well enough to know how components should be placed relative to one another to get routed transmission lines to behave as expected. And they need to understand field theory and signal integrity issues like impedance control, and proper termination. They need to know how to set up the power bus. Where do you put the decoupling caps, under the BGA or next to the BGA? The answer to that question is completely dependent on where the planes are in the board. Many engineers don't know that, most printed circuit designers do. There's a lot of engineering involved in printed circuit design these days.
Warner: I've noticed, since I returned to the industry around 2009, that many very bright engineers are being asked to design their own boards. When I took a break from the industry in 1998—those two roles were distinct most of the time. EEs just don't know what they don't know in regards to layout, particularly when it comes to high-performance boards, and sadly it causes a lot of unexpected problems.
Hartley: Yes, but you know what's going to happen? They'll make the mistakes that most people make when they first move into this field, like make traces too small to be manufacturable, just for ease of routing. They're going to not design for the assembly or fab processes or set boards up to be testable. They will not balance copper in and on the board. They'll use the wrong dielectrics in the board or place plane and trace layers in a manner that doesn't create balanced stack-ups or good containment of fields. They're going to do all the things that new engineers classically want to do that experienced board designers instinctively know not to do. I’ve known engineers who want to put an odd number of layers in the board. Every experienced designer knows better and knows why.
Warner: It's just unfortunate that it keeps happening over and over again, one person at a time. I have a lot of compassion for them!
Hartley: Wouldn't it be nice if there was an essential training course somewhere to be able to teach people who want to do printed circuit design correctly, how to?
Warner: You think they would integrate it as part of the EE coursework by now.
Hartley: You would think so. Another sad issue, a lot of EEs don't realize the value of PCB West.
Warner: Oh gosh yes, the courses you guys teach are like gold--not to mention the level of expertise. You're there, Dr. Eric Bogatin, Gary Ferrari, Susy Webb and many others. All veterans who've suffered all the mistakes and become experts that are eager to teach others how to take the shortcuts and not make the same mistakes.
Hartley: The value of the classes people teach at PCB West is massive! There are the folks you mentioned and others like Dan Beeker, Mark Finstad, Phil Zarrow, Keven Coates, Doug Brooks, Mike Creeden and Doug Smith, just to name a few. There are instructors from fab and assembly houses, material suppliers, CAD companies, and numerous other industry experts.
Warner: Which leads me to my next question--you've become a signal integrity and high-speed guru. How did that happen?
Hartley: Well, it started in the late seventies when I was doing RF board layout. Because I was an EE also, the RF engineer I was working with put a lot of trust in what I did because he knew I understood his goals and what he needed. So, we worked well together. Through him I started to grasp the concept of a waveguide. The fact is, a lot of people think only RF circuits have waveguides, because that's what engineers refer to them as in that domain. The truth is, any transmission line that's operating above low frequencies is a waveguide. A signal line and its return path, steer the fields through the dielectric… that's a waveguide. Yes, RF circuits are waveguides but so are most circuits. So, it started in the seventies with this RF engineer. For a while after that I went back to digital design, up through the mid-eighties. But it was in the later eighties that IC rise times became fast enough that we were having noise and EMI problems in our circuits. Frankly, none of us understood why. We were all scratching our heads thinking, "Why is this happening?” What we didn't know is that it wasn't related to the clock frequency of the circuit, it was related to rise times. Over a period of a few years, from the late eighties into the early nineties, I figured this out by looking at circuits in the lab, measuring the fields that were emanating from them using near-field probes and a spectrum analyzer and looking at the signal rise times. Finally, I realized there's a direct relationship between signal rise time and the fields coming from these circuits. Once all that hit me, it started to fall into place. Then I went to PCB West for the first time in 1993 and sat through Lee Ritchey's class. I realized, listening to what the man was teaching, this guy has a lot of knowledge. I learned a lot from him that I'd never thought about before, but it all made sense. The pieces fit with everything I'd learned up to that point in my life. So, I set about trying to put all these pieces together and over a period of two or three years, it all became a plan of how to control signal integrity in circuits and how to control and contain fields so that we didn't have noise and interference problems. All of that just came together through a combination of experience and training. That's when I also started buying books. From 1990 until now, I have purchased over 100 different books.
You can't name an author who's written about signal integrity, high-speed design, EMI or noise control, whose books I don't own. Henry Ott, Dr. Howard Johnson, Dr. Eric Bogatin, Dr. Bruce Archambeault. All of these guys--I own all of their books and all the other people like them. Kimmel and Gerke, Terrell and Keenan, Ralph Morrison-- just every book you can think of. It was a combination of what I already knew coupled with reading, then putting two and two together.
Over about a ten year period from '85 to '95, I moved into a position of deeper understanding of these issues. 1996 is when I started teaching, as I thought, "You know, I know enough about this, I can probably pass some of my advice and knowledge on to others as well." And that was my goal, to help others.
Warner: That's great and I know so many people have benefitted from your willingness to teach. Besides, you're a natural teacher, too.
Hartley: Thank you, that's kind of you to say.
Warner: Well, you really are. People enjoy learning from you because you teach in such a plain-spoken way that even someone like me, without a formal technical education, can understand.
Hartley: That's the idea. Many people who attend these classes are printed circuit designers who don't have an engineering background. Yet, they understand that circuit board design has become an engineering discipline. They need to understand this stuff and my goal is to convey it to them in a manner that's not confusing.
Warner: The first time we met, Rick, we were carrying on about how we both wished more designers would visit their board suppliers and learn more about how boards are manufactured. Would you be willing to discuss that further and share a few pieces of advice for today’s Engineers and PCB Designers in our next newsletter?
Hartley: I’d be happy to do that.
Warner: Great, thank you very much, Rick. I look forward to continuing this conversation and passing on some of your hard-won wisdom.
Hartley: Looking forward to it and thank you, Judy.