Some years, it seems That little changes in fiber optics, but 2010 certainly was not one of those years. Last year brought new components, applications and standards, all of which can affect this work in the short term.
Let’s start with what may be the biggest news: bend-insensitive fiber. Optical fiber has always been sensitive to bending losses. As you bend the fiber, you change the effective optical structure of the core/cladding materials, and some of the light in the fiber will be lost. This has always been an issue in tight spaces, such as crowded racks, and when installing fiber in existing structures.
The solution was to simply surround the core of the fiber with another layer of glass that would reflect the lost light back into the core. This works with both single-mode and multimode fiber, creating fibers that can be installed in places that regular fiber would not tolerate. One manufacturer likes to demonstrate this fiber at trade shows by stapling the company’s cable to a wood board, a practice that is discouraged, as it is very bad if one is not working with this special fiber type.
While bend-insensitive fiber makes fiber easier to design into new applications and easier to install, it has some drawbacks. Its compatibility with other types of fiber is not yet fully proven, and since it is not color-coded differently, it would be possible to mix fiber types. Testing multimode systems using this fiber may require special handling, since the fiber’s design affects the modal distribution in the fiber. If you use a mandrel-wrap mode filter as required by some standards, much smaller mandrel sizes are required, since the fiber is affected less by bending, and the mechanism that reduces bend sensitivity may add more light into higher order modes. (I will cover this confusing issue in the future.)
In addition to bend-insensitive fiber, manufacturers began offering OM4 multimode fiber, a higher bandwidth 50/125 micron fiber optimized for 40 and 100 gigabit data. OM4 fiber is more expensive but a good choice for backbones and data centers where data rates continue to increase.
On the applications side, fiber optic local area networks (LANs) using fiber to the home passive optical network (FTTH PON) technology (Electrical Contractor, February 2010) began to gain traction in the government and private enterprise network sectors. The growth in wireless data and video on smart phones created a large market for expanding fiber optic backbones for cellular phone towers. On the towers themselves, the increased number of new antennas has created a problem in the bulk of all the coax cable needed to connect them, so fiber is now being used on many cell towers for that application.
Data centers continue to grow, and new Ethernet standards at 40 and 100 gigabits per second have been adopted for future use. These higher speeds limit copper to about an arm’s length, so fiber is the preferred medium. However, both 40 and 100 gigabit links use multiple 10 gigabit channels to implement the higher bit rate. Since a link requires two fibers transmitting in opposite directions for full duplex operation, 40 gigabits per second needs 8 fibers, and 100 gigabits per second needs 20 fibers.
Both of these higher bit rate links are implemented using multifiber MTP type connectors, even at the transceivers. Since these connectors are not readily field-terminated, data center links are migrating to prefabricated cabling systems, which only require pulling cables and testing them. Since few test sets accommodate these multifiber connectors, cabling manufacturers have already expressed concerns over testing them, and new test standards may complicate the issues.
A new version of OFSTP-14 now covers testing multimode cable plants. The version I helped write more than 20 years ago has been replaced by a new international standard that has several substantial changes. First, the new standard changes the names of all the reference methods. Forget “Method A, B or C,” and just remember one, two or three reference cables.
A potential problem is the addition of allowing optical time-domain relectometers (OTDR) for cable plant testing as long as you use both launch and receive cables when testing. This can prove to be an issue if either end-users or contractors do not understand the techniques and limitations of OTDR testing, which I will cover in the next few months.
The biggest confusion factor in the new OFSTP-14 is a new method of specifying modal launch conditions for test sources called “encircled flux.” It’s hard to understand and harder to test but potentially more accurate than the older methods. However, its implementation is still poorly defined and will also be a subject of a future column.
HAYES is a VDV writer and educator and the president of The Fiber Optic Association. Find him at www.jimhayes.com.