For most people working in the fiber optic communications field, their experience with optical fiber involves using jacketed fiber optic cables. This includes network and data center infrastructure cabling, which may contain hundreds of fibers in each cable, as well as patch cables with fewer fibers used for connecting devices. Due to the multi-fiber nature of jacketed cables and network devices such as optical transceivers, which require a Tx/Rx fiber pair cabling connection, it's commonplace for many people to think of spooled bare optical fibers for network simulation testing and optical delay applications in a similar format. Additionally, since the fiber manufacturers provide maximum signal attenuation/loss specifications for a fiber, there is sometimes a misconception that the resulting performance will always be within that defined specification when using spooled and terminated bare optical fibers.
Due to this, when sourcing bare optical fibers in a spooled format for testing and latency-driven applications, it's crucial to understand how bare fibers are manufactured, along with the impacts of spooling and terminating a bare fiber, which will add additional signal attenuation/loss. Otherwise, the resulting solution and/or fiber performance may not align with user needs or expectations.
Spooled Bare Optical Fiber - Technical Insight #1: Simplex vs Duplex / Multi-Fiber
One of the most important things to remember about bare optical fiber in its most basic, unjacketed form is that it is always manufactured as a single (simplex) strand, a result of the fiber drawing process. A brief overview of how optical fiber is made can be viewed in this video created by Corning®:
Click the image to view this YouTube video
Coming off the draw tower as a single strand, it is then spooled onto the manufacturer's shipping reel. Afterward, most bare fiber (99%+) is then delivered to cabling entities that add the necessary extra protective jacketing materials to turn it into a finished cable product.
However, when sourcing network simulators and optical time delay products that include bare optical fibers, it is essential to remember that they will be spooled or coiled in a simplex format. Therefore, if a person is replicating a network span to certify a new optical transceiver that requires a Tx/Rx fiber pair or duplex connection, it will require two individual strands of bare fiber, each on a separate spool or coil.
As a real-world example, when planning to acquire a network simulator to test an optical transceiver with an 80km transmission range, since a Tx/Rx fiber pair is required, a total of two each 80km bare fiber lengths (ie, 160km total) will be necessary to deliver the necessary duplex span. Otherwise, if a duplex configuration is not requested, the user will receive a simplex 80km strand, which will be insufficient. While a reputable manufacturer of high-quality network simulator products should always confirm this detail during the consultation process, specifying whether a simplex, duplex, or multi-fiber bare ribbon setup is needed will help ensure the solution configuration matches the project needs.
After mentioning bare ribbon fiber, it is worth noting that an exception to the above is that a specialized network simulation and optical time delay solution provider, such as M2 Optics, can also build solutions that utilize spooled "bare ribbon fiber" terminated with MPO connectors. Instead of a single fiber, two or more bare fibers are joined together using an additional layer of coating and delivered in a ribbon format (ex: 12-fiber bare ribbon). However, although it is called bare ribbon, it has a much larger physical diameter due to the multiple fibers and extra coating, so the total length of ribbon that can be wound onto a spool or coiled is significantly less than the length of a single strand of bare fiber.
Spooled Bare Fiber - Technical Insight #2: The Impact of Tension and Terminations on Span Attenuation/Loss
One of the most important performance specifications of an optical fiber is the signal attenuation (or "loss") value of the fiber. This is most often expressed as a db/km value on the data sheet for the fiber and often provided for several common signal wavelengths. For example, the attenuation specifications for most standard G.652.D single mode fibers are provided for 1310nm, 1550nm, 1625nm, and sometimes others like 1490nm, etc. For example, a manufacturer's data sheet may show an attenuation value of ≤ 0.18 db/km at 1550nm.
What many people do not realize is that the published attenuation values for an optical fiber are measured by the manufacturer in a "tension-free" state, typically in a cabled format similar to its intended use case in a network. However, winding or spooling bare optical fiber onto a reel requires applying enough tension to not only ensure the fiber stays on the reel, but also so it consistently lays down straight and tight enough to avoid any bending or looseness that would result in significant attenuation events. This winding tension naturally adds a slight amount of additional signal attenuation. Additionally, using spooled bare optical fiber for network simulation or delay line applications, unless directly spliced to another fiber, requires connectors, which also add some insertion loss. Therefore, when working with spooled bare optical fibers, it is not uncommon for the final span attenuation value to be slightly over the published data sheet attenuation specification, resulting from the added signal loss from the spooling tension and connectors.
As a practical example, an entity requires lengths of a G.654 single-mode bare fiber in M2's Fiber Lab network simulators to precisely replicate several known submarine fiber link distances for device certification testing procedures. The attenuation specification of the fiber on the manufacturer's data sheet is 0.158 db/km at 1550nm. However, after the fiber is precisely spooled, terminated, and packaged, OTDR test trace results show overall span attenuation values averaging around 0.162 db/km. An initial thought might be that the fiber attenuation performance is "out of spec" because the loss values are slightly above the stated maximum value on the manufacturer's data sheet for the fiber. However, a final attenuation value just 0.005 db/km over the tension-free fiber specification is an exceptional result, especially when accounting for these two minor attenuation-inducing factors.
Key Takeaways When Sourcing and Using Bare Optical Fibers
Aside from selecting the most appropriate fiber types when procuring spooled and packaged bare optical fibers for network simulation and latency-driven networking applications, two important things to remember are:
- Bare optical fiber is always manufactured and spooled in a single strand (simplex) format. If requiring two or more bare fibers in the span, it will require specifying two or more individually spooled/coiled strands.
- Published optical fiber attenuation specifications, along with other optical characteristics, are measured by fiber manufacturers in a tension-free cabled format. A bare optical fiber on a spool/reel has additional and necessary tension placed on it to achieve effective winding. This extra tension, along with the addition of connectors, can introduce a small amount of additional attenuation to the bare fiber span.
M2 Optics - Your Bare Optical Fiber Solutions Partner
For over two decades, M2 Optics has specialized in manufacturing customized network simulation and optical time delay products. Combining unparalleled access to all types and brands of bare optical fibers from leading manufacturers, including Corning®, OFS®, Prysmian®, and Sumitomo®, with the most efficient packaging techniques and advanced testing capabilities in the industry, M2 is the trusted go-to solution partner for leading communications engineering teams.
Contact M2 today to source precise length bare optical fibers and specify a customized solution for your project needs.
