Back to Connector Basics: A Q&A with APEX
APEX Electrical Connector Consultants LLC is an engineering consulting firm specializing in connector and interconnection challenges, ranging from basic core technologies to problem solving, specific connector selection, and applications. Based in Harrisburg, Penn., APEX is a seasoned group of connector engineers and scientists with many years of experience in this industry. Throughout 2014, APEX will be contributing a series of articles to Connector Supplier called “Connector Basics.”
We thought it would fun to launch the series with a Q&A, led by Bishop & Associates’ own John MacWilliams (a former colleague of many in the APEX crew). We’d like to get their take on some of the major challenges facing the industry today and what they envision looking forward.
The Industry’s Challenges
John MacWilliams: Looking at the core competencies in the connector industry, would you say your main focus is in the area of connector technology and its applications in electronic systems?
APEX: Yes, that’s an excellent start, but there are a lot of supporting areas that APEX serves. At last count we have 22 expert consultants as part of the APEX group. APEX is the broadest service firm that we know of that serves two industry segments – both OEM users of connectors, and designers and manufacturers of connectors.
JM: In working with your customer base, what are some of the biggest challenges you are faced with? What issues most frequently need to be solved?
APEX: Often it is failure analysis when a company is in production with a qualified solution and all of a sudden there’s no system. Many times they can’t determine what caused the problem on their own, so we will be asked to investigate and make recommendations.
Instead of thinking in terms of big, broad technology changes or problems in the industry, the fact of the matter is, in the real world there are a lot of very specific problems that people want answers to, and they can be almost anything across the board. For example, it could be a connector failure in the field or a sequence of erratic behavior with a connector and they’re not sure why. It could be a production issue, such as with plating or quality control in general; and clients come to APEX for specific advice in a given technology area. The other side of that is thorough design at the head end, so that issues don’t crop up later as failures – often times requiring extensive analytical analysis to determine and eradicate the cause. Also, clients benefit from simulation analysis on the front end, through both mechanical and electrical modeling, to predict how well a design is going to perform before it gets too far along. This saves a lot of money and effort in the long run.
APEX also gets involved with helping OEMs with connector selection. We’ll review a number of manufacturers’ designs for a particular application and evaluate to what extent the connectors will satisfy their needs. It could be that only one manufacturer’s connector will work; it could be that every one they look at will work (although that’s usually the exception). They don’t want to get into production on their equipment and then run into a connector issue, particularly with smaller-sized, high-density connectors.
JM: So an end user/OEM might come to APEX because they want an independent and objective viewpoint on a particular issue, whereas if they went to one manufacturer at a time to get the information, it would be much more time-consuming?
APEX: Independent is the key word there, because not all connector companies are unbiased and may not have a broad industry viewpoint.
JM: If you look at the connector industry as a whole, what do you think are some of the biggest technology challenges it faces now or five to 10 years from now?
APEX: One challenge that continues is whether to choose gold plating or alternative finishes due to the cost of gold. Another challenge is the space limitations — everything is getting smaller, smaller, smaller.
Copper Versus Fiber
JM: So the key issues are miniaturization, high-speed signaling, and multi-Gigabit range — I guess we’re talking about 40GbE right now and then 100GbE. I don’t know how you feel about the changeover from copper to fiber, but it seems that copper just keeps improving. We’re already seeing 25Gb/s backplane connectors and that probably can go up to 40Gb/s, for a relatively short distance. Are we reaching the end point of where copper can go or do you think there’s another five to 10 years left before they have to switch over to fiber?
APEX: The bulk of the applications can still be handled more than adequately with copper. That, plus the fact that we’re now starting to add chips into the connectors to do signal processing and signal conditioning. I think there’s going to be a lot of years before the high-speed backplanes go away from a copper-to-copper interface.
JM: When the Thunderbolt interface was introduced, they were originally planning to do it in fiber. However, the cost was too high for a mainstream commercial product, so they went with copper and put the chips in the connector. With USB, the 3.0 is 4.8G, but there will be a 4.0 at some point. That particular interface seems ripe for a switchover to fiber optics. What do you think?
APEX: Once they get the cost of the opto-electronics down, then yes; price is holding things back right now. Fiber is still the preferred method if you have any distance to cover.
JM: Going back at least a decade, following the dot-com bust in 2000 and then the recession in 2001 that devastated the telecom industry, there was a lot of fiber technology that had been developed up to that point but was lost because of company setbacks. For a while, fiber optics didn’t go down, but it also didn’t go up the way people thought it was going to go. Now we’re reaching a point where it’s just starting to fit inside the box and I think in the next five to seven years we’ll see a significant amount of new activity in fiber optics.
APEX: Isn’t there a big overbuild as far as the major trunk lines and such? There’s a lot of overbuild and dark fiber out there that’s over capacity for today’s needs, particularly in the cabling area.
JM: The question is: Will fiber optics find its way into the system itself? The roadmaps seem to indicate that over the next five to seven years we’re going to see a significant amount of new activity inside the box, where typically fiber hasn’t been found.
APEX: Because of the difficulty in termination, particularly in high-density applications, it is a problem in the box. There is some work afoot, particularly in Japan, in optical fiber backplanes, primarily flexible circuitry-type things, but they are still terminated with conventional types of connectors. Here the labor costs for those interconnections are higher for fiber than for copper.
JM: Related to that, looking forward, do you see any activity in superconductivity that might have some impact on the connector industry?
APEX: There’s been work done by some supercomputer companies and others over the years in an attempt to do that. One company spun off a nitrogen-cooled system that helped the speeds considerably, but obviously, with the core processors, one of the major problems is the amount of heat that’s generated. So the possibility of liquid cooling some of these next-generation systems adds to those equations but it’s a cost driver. If you go back to your comment about the fiber, you’re probably still going to end up in a serial mode rather than a parallel mode to get chips that are fast enough to deal with it. I can process a lot of data in parallel mode, but if I have chips that must run serially, it’s going to limit my processing capabilities.
JM: Aren’t there also some pretty significant improvements in copper alloy technology right now, in terms of conductivity?
APEX: Companies have made progress in the last few years with the alloys, and at least one supplier brought out an 80% conductivity alloy for the automotive industry. There has been a move afoot there. The push is coming from the automotive industry because they need so much power now with all these microprocessors and conductivity becomes critical.
JM: What you are saying is that the major challenges facing the industry now are things like high-speed signaling and very small subminiature contact technologies—but what if we look out five to seven years? The semiconductor industry is reaching the end of Moore’s Law scaling, reaching the endpoint of these smaller and smaller interconnects on semiconductors.
APEX: You’ve got to push these walls. They are approaching the limits of the dielectrics today in semiconductors. We don’t know how much thinner they can get. The other aspect of it is, every time you make that wall thinner, you’ve got to drop the voltage so that you don’t punch through it, so you get down to a situation where right now a lot of these chips function at 1V of working voltage, but you’re going to end up at half of a volt once you push these walls much thinner.
JM: The real question there is, with the level of integration that we’ve already achieved, with System-in-Package and System-on-Chip, do you think that in five to seven years we’ll see a much higher level of integration so that the number of outboard components, outside of the chips themselves, might be reduced?
APEX: The thing that is really going to make or break SoC is how do you get rid of the heat. If you can’t get rid of the heat, it becomes an academic exercise. Obviously you have to be able to bring enough power into it, too, but that’s doable. There’s an optimal position that will come out of this System-On-Chip that will allow for it. The other driver, when you start to look out that way, is wireless. We’re going to end up with a lot of wireless interconnects — antenna technology, etc. — that add to what a computer is going to look like five years from now.
JM: I’d like to focus on the core competencies of the connector industry, things like contact technology, plating, molding, materials, simulation, and so on. I’d like to hear your thoughts on what is really required today.
APEX: When we were working for OEMs, we had customers such as large semiconductor companies and automotive companies; their emphasis was on being involved in the development process, including the validation of qualification testing and the manufacturing process, especially for connectors. We’re not sure that OEMs are interested in how the manufacturing technology actually works. What we’ve run into, when people have trouble developing connectors, is the lack of knowledge of just the basics – understanding the basics of what to look for when you’re specifying components and choosing a supplier; that’s probably what more engineers are interested in, rather than the real nitty-gritty of the technologies. There seem to be fewer component engineers than there were about 10 to 20 years ago. This is a lost competency that served the connector industry very well.
In the connector industry overall, in the last few years, the level of expertise has generally declined; in other words, through mergers, retirements, staff reductions and the like, it seems like connector expertise has been lost — certainly with OEMS, and to a large degree, too, the connector manufacturers themselves. APEX finds itself in the position of providing a lot of fundamental foundational training to the industry overall.
We call it “losing the recipe.” Over the years, the industry used to keep pretty tight wraps on its technology base and maintained a strong internal knowledge base that has eroded over time; they’ve lost the recipe and need to relearn it. APEX is in the strong position of retaining a large part of that knowledge base (expertise) and making it available on a consulting basis.
Senior Consultant and Analyst, Bishop & Associates Inc.John has enjoyed a long and diverse career in the electronics industry, including management positions with IRC, TRW, AMP, and his own company, US Competitors LLC. He is the author of many industry articles, including past and current iNEMI.org connector industry roadmaps, US government competitiveness initiatives, and numerous Bishop Reports on the computer and consumer electronics industries. He is an outspoken supporter of the future of US manufacturing in a global marketplace.John is a graduate of the William Penn Charter School in Philadelphia and Lehigh University in Bethlehem, Penn. He and his wife Louise reside in Newark, Del., and Delray Beach, Fla.
Latest posts by John MacWilliams (see all)
- The Return of Manufacturing to the US: Fantasy or Reality? - November 17, 2015
- Convergence of Smart Computing Device Markets - October 20, 2015
- Opinion: US Oversight of the Internet - August 19, 2015