Fiber Optic Connectors Make a Slow But Steady Comeback
By John MacWilliams, Bishop & Associates Inc.

During their long evolution, fiber optic (FO) connectors have gone through significant changes in the marketplace, due to changing technology and business conditions. First made available in the 1980s, fiber optics was thought by many to be the future of electronic interconnect, and hundreds of millions were invested in research and development in the following decades to pave the way for that anticipated nirvana. Cost was a problem, as was the serial, cable-intensive nature of FO designs. They didn’t fit the typical architecture of an electronic system. Then the 2000 telecom bust occurred.

In many ways, the FO industry is still recovering. Fortunately, the point-to-point characteristic of the FO system was put to good use in the telecom long-haul network prior to the 2000 meltdown, and therein we find the foundation of the technology’s renaissance.

FO replaced almost all copper in the long-haul network because of its compelling economic advantage in that cabling environment. This resulted in a multi-billion-dollar, 20-year build out, primarily with single-mode (SM) FO designs, which can reach > 60km @ 10Gbps. These designs have few conventional FO connectors and a lot of fusion welding. FO later expanded into the metropolitan and local loop, using both SM and multimode (MM) fiber, with more connectors, organizers, and switches.

After years of delay caused by regulatory issues and the 2000 bust, FO is now beginning to connect homes to high-speed broadband networks, but it’s still well behind hybrid fiber-coax CATV networks. In this implementation (shown below) fiber is dropped from overhead, or underground cables to the house, where a network interface device (NID) converts the signal back to copper Cat. 5-7 twisted-pair, or (as shown) coaxial cable. This means that in many metropolitan areas, two to four competing broadband systems will be deployed: CATV, satellite, fiber-to-the-home (FTTH), and WiMax.

FO’s eventual “take-over” of electronics applications hasn’t happened, and probably won’t for at least two more decades, (maybe never, as Si integration eats up more and more real estate), but there are an increasing number of equipment applications. Some are specific niches, such as turnpike toll readers. Others have more significant volume potential, ranging from gigabit and above Ethernet LANs, to enterprise storage systems, to plastic optical fiber (POF) use in S/PDIF and other consumer/audio applications, to hybrid optical/electronic interfaces in high-speed backplanes—to the aforementioned FTTH.

Now and Beyond
Fiber optics has been primarily used intra-system where speed, bandwidth, and cable, leverage the strength of fiber optics. Cost, loss budgets, latency, and other issues have limited fiber’s use inside systems, which are still primarily electronic, from the silicon chip to the I/O panel.

Where will fiber optic circuitry be employed?

Is an all optical system possible?
Not at this time. Roadblocks include lack of low-cost (Si) optical ICs or FO printed wiring boards. Most importantly, copper circuits meet 98 percent of all existing applications—and are continually improving performance to levels not thought possible 10 years ago.

What elements of a photonic system are available?
GaAs, InPh lasers, WDMs, transceivers, connectors, and other components—enough to address high-speed backplane, I/O, and inter-system requirements up to 5 Gbps—some say as high as 20 Gbps or more. There are also multiple connector standards, including ST, LC, SC, and LG.

What breakthroughs may be coming?
Intel has developed Si Haman lasers, waveguides, and hybrid Si Laser sources. This means future systems could be mass-produced with low-cost Si optical chips. PWBs, with flexible fiber interconnects or waveguides, already exist, as do lower-cost transceivers and connectors. The most likely implementations of a hybrid or all-optical system will be multiple systems-in-package, interconnected in an OEM-controlled environment.   

Over the years FO has migrated from the telecom infrastructure to networks and some equipment. This inward trend will continue because FO has some compelling performance advantages, including immunity to electrical interference. As data rates rise above 1.0 gigabit/sec toward >10 Gb/s, the crossover will narrow. Breakthroughs are necessary in optical ICs, waveguides, and printed circuits (see above) to enable new high-performance electro-optic equipment. When this happens, photonics-based equipment will need packaging and interconnect in the optical realm—most likely in a hybrid FO-Cu circuitry environment.

Barriers to FO proliferation:

• FO cost vs. continual, surprising improvements in copper circuit performance and integration.

• EO conversion issues: propagation delay, lack of optical ICs limit EO circuitry

• Wireless and Internet protocol alternatives to a wired telecom infrastructure.

• Telecom meltdown 2000-2004 was a significant setback to FO progress.

• FO is primarily a cable-based technology vs. copper printed circuit board platforms.

• Globalization is both a geographic and competitive issue.

• China and India will be the next telecom frontier, but mostly wireless.

Roadmap Issues:

• No known materials or process issues.

• End face preparation, alignment, etc. are being addressed (see iNEMI OE TIG).

• Cost differential vs. copper limits scope of FO growth, except where copper barriers exist.

• Infancy of commercial optical Si ICs limits all-photonics circuitry and equipment.

• Industry-wide inertia exists to evolve and adapt copper-based technologies.

• There are ROI constraints on FO research needed to leapfrog copper, particularly since most organizations today are focusing on nearer-term opportunities.

John MacWilliams
Senior Consultant and Analyst, Bishop & Associates Inc.
John MacWiIliams has been in the electronics industry for over 40 years. His main areas of experience have included: U.S. competitiveness programs, market research studies, authored articles, field sales and management, product marketing management, strategic marketing, new product planning, venture development, advertising and media relations, direct sales, manufacturers representative, distribution sales management, and international marketing. MacWilliams has worked with AMP, Diceon Electronics, TRW, and IRC in marketing management positions. Prior to joining Bishop & Associates, MacWilliams served as the group director of marketing and new product planning for AMP.

MacWilliams graduated from Lehigh University with degrees in business management and engineering.

Bishop & Associates, Inc. © 2007