Everything You Always Wanted to Know about Fiber Cable Assemblies (But Were Afraid to Ask)

New fiber to the home (FTTH) networks are being built at a record pace. An integral part of any fiber network is the fiber jumper assemblies that help interconnect the systems, its branching devices, and fiber drops. Fiber network owners put a lot of emphasis on quality components to ensure a futureproof network with minimal maintenance and power costs that will last at least 15 years, and jumpers are one of those components.

There are cable assemblies, and then there are GR-326 approved cable assemblies. Additionally, some cable assemblies are referred to as the “designed to meet or exceed” style of cable assembly. However, in this scenario there is no mention if the assembly is Telcordia GR compliant.

Tier 1 and Tier 2 telecom provider networks are reaching farther than ever before – which is key in closing the digital divide. Reaching new and existing markets, it is vital to consider the elements to which each network is exposed. Customers across the globe require high-quality and high-performance cable assemblies to not only survive temperature extremes, but to maintain their performance through harsh conditions with minimal change in optical performance over a long period of time.

Steps in the rigorous design of a GR-326 certified cable assembly

Source quality components

Connectors: Many sources exist in the U.S. and dozens more offshore. When sourcing, look for well-known connector vendors that have a good reputation and have accreditation for independent testing of their connectors. Ask to see the certified test report. This will reveal who performed the tests, provide the actual insertion loss and return loss (IL/RL) results over temperature, and expose the baseline pre-test data vs. the post-test data.

Regarding the material itself, concentrate on the ferrule specifications. For example, a 125.5-μm inner diameter (ID) ferrule is recommended for high-performance single-mode (SM) connectors.

The next specification on the check list is fiber hole concentricity. Use a better than 0.5-μm concentricity. These specifications hold true for all SM connectors with a 2.5-mm or 1.25-mm ferrule. The second component to examine is the connector strain relief boot.

Cable: For cables with jackets ranging from 1.6-mm diameter to 3.0-mm diameter, use a jacket material that offers a low shrinkage rate—say, less than 5%. All indoor cables will have a certain amount of shrinkage, so to select the best fit a “round robin” can identify the right cable for the job. For 900-μm cable, be less concerned about shrink rates. That said, a tight-buffered 900 μm is best for this size of cable assembly.

Epoxy: This is likely the single most important component of the entire cable assembly. Without a superior adherence between the glass and ferrule, temperature cycling IL results would be very erratic. Make sure that the epoxy has an operating temperature that exceeds the GR-326 temperature requirements. Also, make sure the epoxy is rated for a glass transition temperature testing (Tg) of 100 degrees Celsius or better. If a differential scanning calorimetry (DSC) analyzer is not available, many regional labs offer Tg testing.

Lapping film/Rubber pads: Instrumental in determining the desired result of the end-face geometry are the accuracy and consistency of pad durometer/pad height and pad life. These specifications can usually be recommended by the machine manufacturer along with the lapping film. Diamond lapping film is required for efficient ferrule end-face shaping, and a silica dioxide final film is recommended for controlling the fiber height of the polish. Attributes such as grit size and density need to be checked constantly.

Preparation

Ferrules: They often are machined, which tends to leave some residue in the inside chamfer of the ferrule hole from the machining and the ferrule hole itself. Many vendors do clean their ferrules before shipment, but an additional cleaning step at the factory provides better results. This step will help the adherence of the glass to the ferrule. An ultrasonic cleaner with an isopropyl mix or perhaps a detergent mixed into the water is best to use.

Cable: It is important to use the connector manufacturer’s template for jacket and fiber strip lengths as well as Kevlar lengths. Be sure not to strip the 900-μm coating from the glass until the connector is ready to be terminated. This will protect the fiber from excess exposure to air, which has been shown to leave fibers brittle. Once the fibers are stripped of their coating, a cleaning of the fiber with a lab wipe and acetone has been shown to have the best performance when observing connector fiber heights over temperature. Once cleaned, the fiber should be immediately terminated to the connector.

Epoxy: Most epoxies used for GR-326 compliant use are two-part epoxies (with previously mentioned specifications). Removing air bubbles from the two-part mix is vital for a fully cured epoxy. This can be done using a centrifuge or a bell jar style vacuum. Epoxy with air bubbles in it can cause attenuation of the fiber over temperature or, even worse, cause a catastrophic failure. The other important factor is the mix ratio. If willing to go with a higher-priced epoxy, a pre-mixed and frozen (PMF) product is a great option because the two parts of the epoxy will be measured exactly to the correct ratio and the need to remove excessive air bubbles will be avoided.

Polishing machine: Select a machine vendor that has very tight tolerances on their machines and polishing fixtures. The machine should have controls for downward pressure as well as platen speed. Use of a high-quality diamond lapping film is required for consistent geometric shaping of the ferrule end-face. Attributes like grit size and density need to be checked constantly.

Fiber buffer stripper: The fiber stripper and epoxy are probably the two main root causes for latent continuity issues. The fiber buff stripper is used to strip the 900-μm tight buffer from the 125-μm glass. This is basically a buffer removal tool that uses heat as part of the process to ensure pristine glass, void of nicks or scratches. Lastly, the blades only penetrate a percentage of the buffer, which ensures the blade will never touch the glass.

Years ago (circa 1988), the mechanical strip tools were much cheaper and easy to replace. But these tools needed to be inspected on a weekly basis or even daily. However, a high-power microscope is all that is needed to inspect the whole length of glass fiber for nicks and scratches. If any of these defects show up in the microscope, then the fiber should be cut even with the buffer. Mechanical fiber buffer strippers that are calibrated to around 120 μm often cause these defects, which could lead a nick or scratch to propagate into the cladding and/or fiber core. This occurrence would be classified as a catastrophic failure – one that could be avoided if using the automated heated buffer stripper.

Connector curing oven: Use of a high-capacity oven that allows the heating contact/element to have direct contact with the ferrule is preferred. More “economical” ovens are available on the market that position the ferrules/connectors in a loose-fitting fixture, but this type of oven will not likely achieve a full cure of the epoxy as recommended by the epoxy vendor. A Tg measurement is recommended for determining if the cure time is too short due to the lack of heat transfer to the ferrule fiber hole.

Laser cleaver: A laser fiber cleaver is recommended for its ability to make consistent cuts to the glass and for reducing the epoxy bead removal step in the polishing process. The laser cleaver requires cleaning during the day but has virtually no consumables to replace. (Note: Mechanical-style cleavers have become more accurate and automated lately, so we should consider them as an alternative to laser cleaving.)

Interferometer: This tool is crucial for both the process design and as part of the manufacturing process for connector polishing. Most of today’s interferometers will do the job, but using a well-known brand, as well as one offering regional customer support, is ideal.

IL/RL test equipment: It’s common knowledge that laser and power meter (LPM) testing is the most accurate way to measure mated-pair connector loss/reflectance. The equipment selected to perform IL and RL measurements should have good dynamic range, accuracy, and repeatability. Choose a well-known test equipment vendor whose equipment offers a database and ability to serialize and store IL/RL results. This is very important as GR-326 requires test results and component traceability.

Process

This paper has outlined materials and equipment recommendations but has not yet shared any of the process details/tips. Let’s start with connector curing. If the epoxy vendor specifies a certain cure schedule, pick one and add either five degrees or another five minutes as a buffer for the cure schedule.

Polishing: Just as important as the quality of the lapping film is the quality of the rubber pads that the lapping film is placed on. Determining which rubber pad is needed requires a bit of process development, with ferrule pre-dome and ferrule hardness as the main variables in the equation. Empirical testing can dial in the process as needed along with a quality rubber pad. Care in the maintenance of the rubber pad in manufacturing is of utmost importance. Ultraviolet (UV) light from ambient sunlight can cause the rubber to lose the durometer specified by the vendor.

Geometry: The current specification for ferrule radius is from 10 μm to 25μm. Because there is a temporary deformation of the ferrule end-face when mated, and certain interferometers may have different tolerances, it is recommended to start with a geometry limit that is well within the GR-326 limit. For example, use an acceptable range of 12 μm to 22 μm, as opposed to accepting results that may be a few tenths of a micron inside of the limit. Further, the geometries can be altered to reflect optimal performance, such as using different durometer rubber pads and sometimes the pad height, depending on the manufacturer of the polish machine. Much like the radius requirements, the fiber height should have an internal control that is well inside of GR-326 limit. Because the post-cycling geometry limit for fiber height is +50 nm to -50 nm, pre-test fiber heights should be in the +10-nm to +40-nm range.

Testing: For per-test IL and RL measurements, consider using an internal limit of better than 0.2 dB (objective) and 0.4 dB (requirement) to allow for some fluctuation that will certainly happen to some extent due to temperature cycle/aging. It is usually the company’s choice whether it uses a random mating process for testing or a master test Jumper. This choice will need to be determined before environmental testing.

A good amount of empirical testing is required to fine tune the GR-326 process. Consequently, this is precisely the reason why there are only a few Telcordia GR-326 approved cable assemblies. Many Tier 1 providers require Telcordia-approved assemblies, and it follows that this is the reason why GR-326 approved/accredited jumpers fetch a higher price - quality is worth paying for, and some companies demand it.

Jim Henschel is director of compliance engineering at Clearfield.