The extension of optical links deeper into homes and buildings is putting an emphasis on coupling both plastic and glass multimode indoors fibers in a low-cost low-loss manner. This task checked the statistics of
connecting a variety of such fibers to get an idea of what can be expected in a real-world environment.
The rapidly rising use of sophisticated electronic devices such as laptop PCs, tablet computers, smart 3G phones, and 3D electronic games is resulting in an ever-increasing bandwidth demand on indoor
communication links. Many studies have examined the problem of how to deliver high-quality digital services to all such devices that typically are scattered throughout a building. Among the candidate communication media are wireless links, glass single-mode optical fibers, glass multimode optical fibers (MMF), and multimode plastic optical fibers. Owing to their relatively low cost and ease of installation compared to single-mode fibers, multimode plastic optical fibers (POF) have become an attractive indoor medium. If such POF links are implemented properly, they have the following advantages: (1) the fibers are easy to
interconnect because of the larger core areas, (2) the use of POF links will reduce network costs because connectors are easier to attach to plastic fibers versus glass fibers, (3) and, compared to single-mode links, less expensive light sources can be used. In addition, the standard core sizes of POF include 50- and 62.5-μm diameters, which are compatible with the core diameters of standard multimode glass telecom fibers. Whereas multimode glass fibers used in early indoor installations had 62.5-μm core sizes to enhance the coupling performance of fiber-to-fiber interconnections, currently many premises installations use 50-μm core diameter fibers. This is done to accommodate Gigabit and 10-Gigabit Ethernet applications for which laser-optimized fibers, such as OM3 and OM4, can be run over distances of several hundred meters. In ISO 11801 the International Standards Organization has classified multimode fibers by the designations OM1, OM2, OM3, etc., which indicates the modal bandwidth of the MMF. OM3 and OM4 fibers provide sufficient bandwidth to support 10 Gigabit Ethernet over distances up to 300 m and 550 m, respectively.
This study task examined the practical implementation issues and pitfalls that can arise when attempting to interconnect POF and MMF to form the distribution network inside a building. A major concern is the random interconnection of glass and polymer optical fibers, which will occur in an actual application
environment inside a building. Of particular concern is the loss performance at joints between different types of fibers that are terminated with factory-installed connectors. These fibers could be either newly installed or they could already be in place in a building. In either case, the terminations on these fibers may not
necessarily prescribe to a strict termination-loss criterion.
The experimental setup for measuring the losses in individual POF and OM3 segments is shown in Fig.
10. As shown in Fig. 10a for the POF segments, a PC-to-SC adaptor was used at the transmitter end to couple the VCSEL flylead to the POF. An optical fiber patch cord (1-m in length) was used to adapt the SC connector on the POF fiber segments to the PC connector on the power meter. For measuring the power losses in
individual OM3 segments, the OM3 fiber output end was attached directly to the power meter, as shown in Fig. 10b.
Figure 10. Configuration for characterizing link losses of individual and combinations of fibers; (a) Setup for measurement of POF segment losses, (b) Setup for measurement of OM3 segment losses, (c) Setup for measurement of paired fiber segment losses.
To fully characterize the connection losses all the POF and OM3 segments, loss measurements were made in both directions. In order to keep track of the direction in which the losses were measured, one end of the fiber segment was labeled 1 and the other end was labeled 2. For the loss testing, first a fiber segment was inserted into the setup (with end 1 being the input) and its loss was measured (with end 2 being the output).
This particular configuration was called “1 in, 2 out.” The segment was then disconnected and reinserted with the original input and output ends switched around. Because now the light is inserted into end 2 and the loss measured at end 1, this configuration was called “2 in, 1 out.”
Following measurements of individual POF and OM3 fibers, paired combinations of POF segment D and the five OM3 segments were characterized. As shown in Fig. 10c, the setup consisted of POF segment D followed by a selected OM3 segment. A PC-to-SC adaptor was used to interconnect the POF and OM3 fibers.
The OM3 fiber output end was attached directly to the power meter. A bar graph comparison chart of the measurement results is shown in Fig. 11. As can be seen, the losses are quite consistent at a specific operating wavelength.
Figure 11. Comparison of losses for various fiber segment combinations between POF(D) and OM3(X).
Similar loss results were obtained at an 850-nm wavelength when the POF(D) segment was replaced with POF(A), as shown in Fig. 12. However, in this case the loss variation at 1310 nm was much larger with loss values ranging from 14.116 dB for the POF(A)-to-OM3(A) link to 16.288 dB for the POF(A)-to-OM3(E) link.
This 2.172-dB variation means the power loss for POF(A) coupled to the OM3(E) segment was 1.65 times higher than when POF(A) was coupled to the OM3(A) segment.
Figure 12. Comparison of losses for various fiber segment combinations between POF(A) and OM3(X).
Conclusion: Emerging high-capacity in-building telecom networks are expected to use an interconnected mix of multimode glass and plastic fibers. Preliminary loss measurement results for individual and combinations of POF and OM3 fiber segments indicate that careful connection loss assessments must be made in a real environment in order to have transmission links with consistent losses.