You're reading an older article from ELECTRICAL CONTRACTOR. Some content, such as code-related information, may be outdated. Visit our homepage to view the most up-to-date articles.
Laser bandwidth—information-carrying capacity—has become critical for premises networks as these systems ramp up to gigabit speeds using inexpensive lasers and multimode optical fiber. Electrical contractors will benefit from understanding bandwidth and laser-fiber interaction related to to network design and installation.
Multimode fiber (MMF) bandwidth is dependent on modal dispersion, which is the spreading of light pulses as modes travel different paths of varying lengths within the core of the multimode fiber. The modes arrive at a receiver at different times, discrete pulses merge, and bit errors occur.
To help compensate for modal dispersion, multimode fiber (MMF) has a graded refractive index. Light travels faster toward the core-clad interface and slower at the center of the core. If this graded index profile were perfect, very little modal dispersion would occur because the modes traveling greater distances would move faster than those covering shorter distances, and all modes would arrive simultaneously. In practice, however, some modal dispersion is inherent to MMF and is its primary bandwidth-limiting factor.
The bandwidth of MMF has been specified according to a minimum over-filled launch bandwidth (OFL BW) standard, TIA FOTP 204, at 850nm and 1,300nm wavelengths. This measurement procedure, which distributes light uniformly throughout a fiber’s core, has long been a reliable method of specifying information-carrying capacity in MMF when used with light-emitting diodes (LEDs).
LEDs distribute optical power throughout the fiber’s core and couple power in many hundreds of the fiber’s modes. But fiber specified according to elevated OFL BW alone may not meet expected performance requirements when used with lasers in high-speed networks.
With the OFL, most of the optical power is focused outside the center of the fiber. Lasers, on the other hand, distribute optical power in less than 5 percent of the fiber modes, resulting in low modal dispersion and the potential for much higher bandwidth. However, even small errors in the index profile at or near the fiber’s center can dramatically reduce laser performance, although these same small errors do not necessarily degrade OFL BW.
The result: MMF specified according to elevated OFL BW alone may not meet expected performance requirements when used with lasers. In fact, many fibers demonstrate very high OFL BW and decreased performance.
One solution is to use a mode-conditioning patch cord to direct the laser light source to enter the core of the fiber several microns from the center. These special connectors, which Gigabit Ethernet (IEEE 802.3z) requires for use at 1,300nm, launch the light from a single-mode fiber into a MMF through controlled offset, thus avoiding the defects present in some MMFs. However, the patch cords are cumbersome and expensive.
A better solution is an MMF manufactured to eliminate errors in the index profile. New laser-optimized MMFs, which are designed for optimum performance with low-cost GbE laser sources, now are available. Yet, problems remain.
Studies have consistently shown that vertical cavity surface emitting lasers (VCSELs) and other short-wavelength sources that create a small spot of light—known as a restricted mode launch—produce increased bandwidth in MMF. The trouble is that short-wavelength sources have exhibited a variety of launch conditions. To guarantee consistent link performance, the laser launch condition must be specified using a reliable measure of restricted mode launch.
Manufacturers and end-users are used to specifying fiber performance based on OFL BW, which is not a good indicator of functional performance in laser-based systems. Therefore, for optimum performance in laser-based systems using MMF:
• Laser launch conditions must be restricted such that fewer modes are used, because using fewer modes increases bandwidth.
• Fiber must be specified for laser bandwidth, rather than OFL BW, under these restricted launch conditions.
The Telecommunications Industry Association Task Group on Modal Dependence of Bandwidth (TIA FO-2.2.1) is completing work on launch criteria for laser sources and on new fiber bandwidth measurements: FOTP-203, Launched Power Distribution Measurement Procedure for Graded-Index Multimode Fiber Transmitters, FOTP-204, and Measurement of Bandwidth on Multimode Fiber.
FOTP-203 provides a standard procedure for measuring launched power distribution for MMF transmitters. It involves shining light through a short length of MMF and measuring the percentage of power within a given radius at the receiver end. This measurement is used to characterize the launch spot size carried by an MMF, specify transmitter launch conditions, and ensure a restricted mode launch (RML), which is necessary to limit the number of modes stimulated.
FOTP-204 describes methods for determining and measuring information-carrying capacity of MMF when using a restricted launch. In the test, restricted launch is created by filtering an over-filled launch condition with a mode-conditioning patch cord consisting of a special RML fiber: graded-index MMF with a 23.5µ core. After testing a variety of fibers for restricted launch, the Task Group determined that 23.5µ fiber was the best choice because it produces a launch condition most like that of a VCSEL.
RML bandwidth measurement test procedures for 62.5µ MMF are completed and the work is well along to establish comparable measurements for next-generation 50µ MMF for 10 Gigabit Ethernet (IEEE 802.3ae). These are important developments in the evolution of optical fiber local area network (LAN) designs from LED-based to laser-based systems, which soon will operate at multi-gigabit-per-second speeds.
PONDILLO is senior market development engineer at Corning Incorporated. He can be reached at either (607) 974-7311 or [email protected].