Copper and Fiber Vie for Position
By Bob Hult, Bishop & Associates Inc.

For nearly 40 years, fiber optic communication technology has experienced a unique combination of success and failure. Fiber optic cables offered the ability to send thousands of voice channels many kilometers without amplification, and they were quickly adopted by the telecommunications industry for long-haul applications. These tiny glass strands replaced bundles of copper wire that were prone to crosstalk, signal loss, and corrosion.

As the Internet expanded, devotees of fiber saw it as the ultimate communication media, and predicted immense growth of fiber optics in all types of equipment. Huge investments in the development of fiber, connectors, filters, splitters, and transceivers were made in expectation of the eminent demise of copper.

But these grandiose forecasts proved wildly optimistic as engineers continued to find ways to extend the useful bandwidth of conventional copper conductors in both I/O and PCB applications. Low voltage differential signaling reduced distortion due to external EMI. Improved dielectrics and signal/ground pin assignment patterns improved crosstalk characteristics in connectors. Major advances in chip technology introduced signal conditioning features, including pre-emphasis and equalization that enabled circuits to distinguish and reconstruct data signals buried in noise and distortion. The science of signal integrity developed guidelines for successful design of multi-gigabit circuits. Designers found that copper remained the most cost-effective choice while minimizing risks associated with the introduction of any new technology.

Vast improvements in copper signal transmission, together with the lingering effects of multiple economic recessions, caused many promising optical research projects to be shelved. Fiber continued to offer almost unlimited bandwidth to next-generation equipment in a future that continued to be pushed out. When asked to estimate when fiber optics will achieve economic parity with copper, engineers came up with a range that stretched from two years to never. On one recent survey, very few respondents were even willing to offer a guess.

Today, copper conductors remain the primary media in the vast majority of consumer, commercial, and military circuits. Copper is widely expected to remain the dominant technology for many years, but evolutionary advances in technology and component production costs are influencing the balance between copper and fiber. It may be time to look at fiber again for the future. The laws of physics may ultimately impose limitations on copper that, over time, will tip the advantage to fiber.

As data rates increase into the gigabit range, high-speed signals tend to propagate on the surface of a copper conductor. The much-reduced area available for signal conduction results in greater attenuation. One solution to this problem is to increase the gauge of the conductor to provide more surface area. This can reduce the attenuation problem, but adds to cable bulk and weight, both undesirable characteristics in high-density I/O panels and cable trays that are filled to capacity. Electronic systems in commercial and military aircraft must fit whatever cramped spaces are available. Large square card cages may be broken into multiple subsystems located throughout the aircraft and cabled together. Minimizing weight and bulk are critical issues for high-speed cables used to link these remote boxes.

Another issue is crosstalk. As signal speeds increase, crosstalk between parallel conductors goes up. Crosstalk distorts the signal and eventually makes it difficult or impossible to detect at the receiver.

Designers can:

1. Increase the distance between conductors, reducing signal density

2. Add shielding structures, consuming more space and adding cost and weight

3. Add active signal-conditioning features on the PCB or to the cable assembly. But these increase cost, consume power, and potentially cause interoperability issues between multiple vendors

A third issue is related to the practical length of a cable assembly, also known as reach. As signal speeds increase, the effective length of a copper cable becomes shorter. The combined effects of attenuation and crosstalk increase with length, effectively capping the reach of copper high-speed cables.

Additional issues associated with high-speed copper conductors include susceptibility to external EMI, signal skew, latency, inter-symbol interference, security, voltage isolation, and equipment damage resulting from conducted electrostatic discharge. Continuing increases in the cost of copper on the world commodity market has also been cause for concern.

Fiber optic links are not perfect, but they offer almost unlimited bandwidth, can propagate high-speed signals many miles without amplification, are immune to EMI and ESD, weigh a fraction of what copper cables do with increased bandwidth, and are nearly impossible to tap, making fiber ideal where security is a concern.

One of the factors that has delayed wide adoption of fiber has been the added cost of components that convert electrical signals to optic at the transmit end, and optic signals back to electrical at the receiver. The added cost of electro-optic conversion, as well as reluctance to espouse a new media, has slowed consideration of fiber. Introducing a new manufacturing process to replace known and thoroughly documented copper conductors can result in understandable reluctance to adopt a new technology. Reliable solder joints have existed since the introduction of electronics, and crimp technology has been proven for more than 70 years, since its introduction. Relatively untrained personnel can quickly create highly reliable connections using simple tools readily available in the field. The laborious, time consuming, and challenging termination process required by some fiber optic connectors contributed to this concern.

Two major economic recessions over the past 10 years also tended to discourage designing a new product dependent on new technology. One engineer commented that fiber would become the preferred media when copper cannot provide the required bandwidth and fiber becomes the only cost-effective alternative.

Optical fiber and connector manufacturers have been busy addressing user concerns about fiber. Terminating a fiber optic connector has been one of the largest objections in applications that require field termination. Standard fiber connectors often required special stripping and polishing tools and the use of disposables, including polishing fixtures, lapping paper, and epoxy. Specialized inspection scopes determine if the termination process was successful, or not. More importantly, successful termination required highly trained personnel with practiced expertise to produce consistently good terminations.

The introduction of newer connectors that feature crimp or self-actuated mechanical attachment eliminates the requirement for adhesives. Connectors that feature pre-polished mating faces has made field termination of fiber connectors faster and much more reliable.

In addition to standard connector styles, including LC, FC, SC, ST, MT-RJ, and MPO, connector suppliers have introduced new fiber optic connectors with features that address the needs of a wide range of applications.

Some of these new styles are sealed against dust and moisture, and/or utilize expanded beam technology that is inherently less sensitive to contamination at the interface. Many of these connectors are focused on military communications and medical equipment applications where absolute reliability in harsh environments is required.

The state of technology today indicates that the majority of cable assemblies in large server, storage, and network installations longer than 10 meters are often best served in fiber. Those that are in the three- to 10-meter range may be either copper or fiber, and represent the most active battleground between copper and fiber today.






If high-speed fidelity or longer reach is critical, copper cables that incorporate active signal conditioning components can satisfy these requirements. Designers wanting bandwidth headroom to support next-generation equipment may choose to go with fiber.




Cables that are less than three meters in length are typically copper. Duplex or simplex LC, FC, SC, or ST fiber jumper assemblies used in telecom and datacom applications are an exception, with high-volume offshore manufactured assemblies dominating this market segment.







Small Form Factor Pluggable modules in standardized formats, including XFP, SFP+, and QSFP+, enable designers to offer both copper and fiber I/O options without modifying the equipment. Optical modules that offer up to 10 Gb/s channels provide high panel signal density, as well as the building blocks to 40 and 100 Gb/s Ethernet links.

The introduction of active optical cables provides another option in the transition to fiber optic media.

These cables offer the advantage of mating to the standard copper interface, performing electro-optic conversion within the strain relief of the cable. A small diameter optical fiber links the two ends. By using the common electrical connector on the equipment, an active optical cable can greatly extend the length of an I/O cable with a link that consumes much less space and weighs much less. Cables that mate with standard Infiniband and QSFP connectors are available from multiple suppliers now, but volume markets are expected to develop for assemblies based on HDMI and USB interfaces in the future.


Manufacturers of fiber optic cable have continued to address concerns about the durability and performance of these tiny glass fibers. Cables that incorporate multiple discrete fibers, as well as in a ribbon format, are available in a wide variety of configurations that may include strength members, buffers, and multiple external jackets that can withstand the worst abuse. Bend-tolerant fibers have reduced concerns about failure due to micro cracking or fracture due to small radius kinks in the fiber. Fibers optimized for laser sources reduce spectral distortion and attenuation.

Exciting advances at the chip level promise to introduce the advantages of fiber into many more applications.

Intel has demonstrated their LightPeak optical I/O chip that has initial bandwidth of 10 Gb/s, but could be scaled up as required by new applications.

It is unclear at this time when commercial products incorporating LightPeak will be introduced or how LightPeak may impact USB 3.0 in high-end consumer computing applications.

IBM, Intel, Luxtera, and others are making significant progress in the evolving field of silicon photonics. The objective is to integrate all of the elements necessary to communicate optically on the microprocessor chip. The result would be the ability to couple high-speed optical I/O signals directly to a discrete fiber or to optical waveguides integrated into a PCB. Real progress has been made in achieving this using low-cost CMOS manufacturing processes.



The material and fabrication process building blocks necessary to achieve terabit links are the focus of intense R&D efforts today. Initial applications will likely be in supercomputing equipment, but will quickly migrate down to high-speed data center and network applications.

There is little doubt that copper will remain the primary media well into the future, but we are starting to see how evolving fiber optic technology will begin to assume a greater role in high-speed applications in both external and internal communication links.



Bishop & Associates Comments: 

1. Copper conductors will continue to reign supreme for many years, but basic laws of physics are pointing to limitations beyond which fiber can provide more cost-effective solutions.

2. Costs of advanced high-speed copper cables are increasing while optical component prices are coming down.

3. Fiber offers key advantages of EMI/ESD and crosstalk immunity, along with reduced cable bulk and weight.

4. New fiber optic connectors entering the market simplify the termination process and are being adapted to survive harsh environments.

5. Silicon photonics research may open the door to true optic backplanes as well as direct chip-to-chip optical links, effectively eliminating the I/O performance bottleneck.

6. Bishop & Associates estimates the global market value of fiber optic connectors will top $1 billion in 2010, a 20+% increase over 2009.

For more information about fiber optic products, visit the manufacturers featured in this article: Molex, Tyco Electronics (TE), Siemon Interconnect Solutions, and Quellan Inc.

Visit Connector Supplier’s online Buyers Guide for FO connectors.

Robert Hult
Director of Product Technology, Bishop & Associates Inc.
Robert Hult has been in the connector industry for more than 36 years. Hult began his career as a sales engineer for Amphenol. He joined AMP in 1972 and served in several management positions through 1996. In 1997, Hult joined Foxconn as group marketing manager for Intel in Chandler, Arizona, U.S. Prior to joining Bishop & Associates, Hult was the regional application engineering manager for Tyco Electronics.

Hult graduated in 1968 from Bradley University with a bachelor of science degree in electronics technology and a minor in business.