For many of us, our knowledge of Denmark surrounds the picturesque capital city of Copenhagen and, within that city, the famous Tivoli Gardens. But Denmark also has a robust tech industry. On several occasions, Speeding Edge partnered with Rolf Ostergaard, founder and owner of EE-Training, to present our two and three-day training courses in Copenhagen and Aarhus, Denmark as well as in Stockholm and Malmo, Sweden and Oslo, Norway. Like us, through conducting these courses, Rolf has had the opportunity to gain a unique perspective on the PCB product development process and the challenges and methodologies deployed by these engineers within their companies for their customer base. This article focuses on the history of technology within Denmark, the types of solutions and services provided by EE-Training, Rolf’s experience and expertise, current trends, future challenges and the impact COVID has had on his business.
EE-Training is based in Copenhagen Denmark and was founded by Rolf in 2004 with a focus on delivering high-quality training courses with a special emphasis on signal integrity and high-speed hardware design, for serious electrical engineers who want and need to stay at the top of the curve.
Rolf notes, “Back in 1999, I was working for 3Com in Santa Clara, California. I was facing a tough problem of getting a fairly advanced cable modem product to pass EMI testing on a 4-layer board. This turned into a hunt for a solution and a deeper understanding of signal integrity issues. Luckily enough, Lee Ritchey happened to be working in the same building. That turned into a fascination for signal integrity and a drive to help fellow engineers understand the dynamics of this aspect of PCB product development.”
He continues “In essence, the humble but significant vision of EE-training is to keep electronic design as an important part of the local economy by providing world-class education to electrical engineers based in Scandinavia as well as other European countries.”
In addition to the courses he teaches through the auspices of his own company, Rolf has a partnership with Dulos, a training company based in Hampshire England, where he serves as a certified instructor. This has led to him being recognized as a worldwide respected trainer in signal integrity.
Upon graduating from the Technical University of Denmark, Rolf began his career in telecom with satellite transceivers and long-range communication systems at Thrane and Thrane (now Cobham). He states, “One of the projects that we did there was the sat phone that they used in the movie ‘Everest’. The movie came out a few years back but is actually based on a true story that happened way earlier with that sat phone.”
Rolf adds, “In 1998, I moved to Santa Clara, California, where I began working for 3Com. It was there that I attended Lee’s internal course and I thought, ‘some of the stuff that I learned back in University actually makes sense now.’”
Rolf moved back to Denmark and in 2004 founded a consulting company where he did advanced electronics and embedded software design projects. He explains, “We did things like high-performance computers and laser systems with the European Space Agency. We also worked on the development of doppler radar equipment for golf players. And we developed ball-tracking technology for a company called Trackman that is used by major league baseball in the U.S. today.”
Rolf then decided to turn his focus on developing training courses and launched EE-Training as a business entity. As the organizer of Nordic SI Week, he demonstrated that pulling in 60+ participants for a topic this unique and “niche-y” was truly doable even as a full-week program.
As Rolf notes, “The reality is that when you select electronics as a career, you accept a life-long struggle to keep up with all the technology changes. There’s no way around this. This can be done through a combination of formal training or on-the-job learning where you either fail or learn something valuable.”
While there is no specific national requirement for on-going training, there may be company rules that mandate it. He explains, “It varies from company to company as to what the rules are or what you should do. I am aware that many companies have that as part of their employment packages where you have a number of days each year that you have to do education.”
“Making the transition from university to making real working hardware is difficult,” he continues. “I know I got a bit unpopular with the Technical University of Denmark because I pointed out that dropping the key course in high-speed digital design was a bad idea. Just because the professor retired, I didn’t think that subject became less important. I voiced my opinion, and then it took a few years before I was called in to help set up something, so we actually have a really popular three-week, full-time summer course on SI today. It’s been running for some years now. When I do training and talk to graduates from other universities, I know the high speed digital and SI topics are generally lacking. There’s still a gap there that is quite significant. But I also understand that these topics are super hard for academia, especially around here, because it requires expensive equipment and it requires preparation of expensive systems to fully comprehend the challenges. So, I totally understand it. It’s a difficult challenge to bridge the two (real world and academia). That’s still a reality.”
Rolf describes the current design environment as follows. “To make it easier, so much of the design work now is really modules and not actual design. It’s not low-level design but more module based.” We talked about how designers are faced with trying to squeeze the last ounce of capability out of existing board and component technology. And how each time something is tweaked, it has an impact on the design process.
Rolf states, “That’s the thing. The designers have to continually keep up with this and try to learn what is required. One of the problems I see is a lot of designers doing design by reference design. They take a reference design and then maybe combine it with another reference design, and then merge that into their product, and that’s their final product. The problem is that they don’t understand the details. So, when they do this, they end up making some kind of mistakes and then when something doesn’t work, or they have to qualify it, they’re a little bit in the dark, and it often creates problems. This is a really scary thing to see. I don’t believe that this is the right way to do things, but I also understand there’s a push to get projects completed. This is probably the natural approach, but it is not the best idea in terms of optimizing a design.”
He continues, “As a designer, you really have to understand the details in your design to a level where you can make modifications. It also comes back to the documentation that is available. You need to have the correct documentation so that you can actually make the right design choices. But, then again, we’re constantly battling an environment where things are moving so fast that maybe the documentation isn’t keeping up with the component technology. New chips are introduced, and designers are challenged to build something with the latest and greatest. It’s a difficult situation, but I think we just have to keep fighting. We have to keep insisting that we understand what we design.”
The foregoing echoes what we experience at Speeding Edge. The least knowledgeable people often write application notes within an organization. Designers and product developers are supposed to follow this information without knowing if what is put in the notes is valid and if it is not valid, what the real truth is.
In terms of current and future trends, Rolf states, “I see a lot of activity in all kinds of health-related technologies. Then, I see lots of IoT stuff with sensors and systems often related to smart city or energy optimization. And, then obviously, there’s a lot of stuff around electric vehicles in different shapes and forms. The automotive industry has this idea of having automotive-qualified components, and I see that technology being specified by some nonautomotive companies requesting that from there sub-suppliers.”
I asked him about the problems product developers are having when they make a new product. He noted, “I have noticed that engineers routinely build 10 Gb/S stuff using reference designs. But, there are some gaps in fully understanding what is important. When they try to push something to the limits, maybe with more connectors or longer links there isn’t the full understanding of what is actually important. Should we focus on skew? Should we focus on loss? Should we focus on impedance mismatches? When do we switch from one type of PCB material to the next or does that even really matter? When I talk about these factors in my classes, I basically see lots of blank stares.” He continues, “But these guys still manage to build stuff that works. But, I think you end up on thin ice when that happens, and I don’t like that. I like it when you are on more solid ground. It’s better long term.”
As noted above, Rolf started his training based on SI and high-speed hardware design. But, he notes, “It’s been a while since I have done a lot of consulting on SI issues. There’ve been some issues around faster memories. There are two factors-things get too squeezed in a way, and that’s combined with trying to reduce cost. Product developers are not using enough layers. With the lower motives of memory systems today, this is apparently a problem.”
In terms of other issues, Rolf notes, that in terms of EMI, things are getting better. He comments, “Lots of engineers are a lot smarter today than they were ten years ago. I see a huge improvement with how good the engineers are and how they are able to understand a lot of issues.”
The improvement is not necessarily the result of formal education. He cites, “I think in general engineers are moving up the curve, and a lot more of them are exposed to doing really fast stuff and they have to read about it. Maybe they learn a bit about what they need, and they are a little more scared. They are doing a lot of learning on the Internet. And, obviously, lots of engineers are getting formal training, and that’s really good.”
When it comes to PDS design which, as noted in previous articles, is the most difficult element in the design process. Rolf declares, “I have been chasing this stuff for many years. When we go through some of this power delivery or bypass information, and one of the engineers comes up after the session and explains that he finally understands how this whole thing works I find it very satisfying.”
He continues, “The challenge is that they have been doing it wrong for such a long time. It’s quite satisfying for me when they finally get it. But it’s also kind of sad that I keep having the same experience year after year after year.”
“There’s usually one or two guys in the room who say, ‘Wow we never thought about that.’ It’s not about anything that is actually really complicated or is really leading edge. It’s actually something that most people should have a good understanding of coming from university. That training just isn’t there in university.”
We also talked about the issues surround EMI problems and how, if they exist, they’re due to issues relative to PDS design. Rolf states, “There’s a very strong correlation there, and these guys learn that one way or the other. Sometimes, they never realize that there is a correlation. But, once they understand it, things are a lot better. But, overall, PDS design is still an issue.”
In terms of challenges that we will encounter in the next 2-5 years, Rolf offers the following: “I see a lot of pressure on all this stuff around UIs. We have these generations, or all of us really, who are used to having these super nice UIs on smartphones. I see pressure to have that on equipment that doesn’t usually have that type of super sleek UI. You can’t do that for something that’s supposed to have a long life. In addition, there’s all this massive trade we have had with Asia and China and the amount of pressure that’s being put on price. As a result, sometimes, I see that turning into quite unrealistic cost expectations. I think we will continue to see even more of this and there are challenges that will be difficult to deal with because it’s difficult to get good quality and also be able to do it very inexpensively. When we have a lot of inexpensive stuff coming in, how do you charge enough for the product to have engineers in higher-wage areas to develop these products? I don’t see this trend going away actually.”
He continues, “I hope we will see more emphasis on low power consumption that is basically driven by the climate agenda. We still require a whole education system around software. It requires rethinking about how we do things. Until now, the whole focus has been on how to lower the development cost of software while achieving higher levels of abstraction. A lot of this needs to change relative to power consumption because you need more efficient software, not higher abstraction. The power consumption of everything on the Internet is way higher than what it needs to be. This can be addressed if we put more resources into developing power-efficient software.”
“Maybe hardware has been getting faster a little bit too fast because the software wasn’t optimized for performance or power consumption” Rolf adds, “software is getting really, really complex without focusing on power consumption. I hope that changes.”
COVID-19 has had a definite impact on the training business as we have readily seen at Speeding Edge. Rolf notes, “We had to switch everything to online, but that’s kind of good because all that traveling does not work well with the climate agenda anyway. Hopefully, we can keep it that way or keep the best of it that way going forward. The funny thing I have noticed is that those people who were previously super busy going from meeting to meeting and were very hard to contact have suddenly been available. That’s been beautiful, and it’s been a complete change of how things work.”
Have more questions? Call an expert at Altium.