Standards-based factory testing of fiber-optic cable

The final installed performance of a structured cabling system is arguably the only physical measure of cable characteristics that is important to the end-user. If the cable meets the requirements after it is manufactured, shipped to the work site, installed, terminated, and tested, why should the results of any other testing along the way be important? File this rationale with myths and legends.

Mar 1st, 1999
Th Cim64024 29

Standards-based factory testing of fiber-optic cable

Users of fiber-optic cable should know what tests are performed, and why.

Andrew K. Straw

The final installed performance of a structured cabling system is arguably the only physical measure of cable characteristics that is important to the end-user. If the cable meets the requirements after it is manufactured, shipped to the work site, installed, terminated, and tested, why should the results of any other testing along the way be important? File this rationale with myths and legends.

Production and verification tests along the way ensure that the structured cabling system is going to work before any investment is made in installation labor, termination, and project time. Testing protects the interests of the manufacturer, installer, and end-user alike. Manufacturer testing on fiber-optic cable falls into two general categories: production testing and characterization, or type, testing. These two kinds of tests are quite different, but each is useful in its own way.

Production testing

Production testing is performed on each cable that rolls off the manufacturing line. Some production tests, in fact, are actually conducted online as the cable is being fabricated.

These test results are useful primarily for internal quality control and process feedback by the cable manufacturer and are of little use to the purchaser. Accordingly, they are not usually furnished to customers. The production test that has the most meaning for the purchaser is the final attenuation or optical-loss measurement. Every fiber in every cable should be measured for both optical loss and point discontinuities, provided that the finished cable is long enough to obtain meaningful, repeatable measurements. Sometimes, short production cable lengths (500 meters or less for multimode cables and 1000 meters or less for singlemode cables) cannot be accurately measured in factories equipped with long-length optical-measurement equipment. Many purchasers require that tabulated optical-loss measurements be packaged and shipped with the finished cable, which is a good idea. Thus, data is useful for comparison with post-delivery or final-acceptance measurements to ensure that no cable damage has occurred during shipping or installation. To be of value, though, the data must reflect post-manufacturing measurements. Values obtained on the raw fiber before it is processed into cable are of little use because they do not reflect the influence of the cable design and cabling process.

Optical loss can be a sign of a poorly designed, poorly manufactured, or otherwise defective cable. If a fiber is subjected to any undue mechanical stress during manufacture, it will be manifested as an increase in optical loss. A twisted, crushed, or pinched fiber will show a point loss or "step" at the location of the defect. It is therefore important that factory measurements include both end-to-end loss measurement and inspection for point discontinuities on every fiber. These measurements should be made and recorded at the wavelength(s) at which the installed system will operate. Because the operating wavelength might not be known when the cable is ordered, most manufacturers measure fibers at the two most commonly used wavelengths: 850 and 1300 nanometers for multimode fibers and 1310 and 1550 nm for singlemode fibers.

For multimode fibers, bandwidth is another important optical parameter. Because the bandwidth of a fiber is a function of the graded-index profile of the glass core, it is measured and specified by the fiber manufacturer rather than the cable manufacturer. In providing bandwidth values to a purchaser, cable manufacturers typically cite the values provided by the fiber manufacturer, rather than re-measure this characteristic. Because of the complexity of the equipment involved, field measurement of multimode bandwidth is rarely recommended or performed. Bandwidth values from the fiber factory have served the industry satisfactorily, and further testing is not required.

Type testing

The purpose of type testing, which is quite different from production testing, is to ensure that the cable will function properly under a range of adverse mechanical and environmental conditions, not just when the final system tests are performed. In many cases, type testing is destructive, so it is not conducted on production cables. It is called "type testing" because selected cables of a particular type may be tested as representative of a whole family of cables of that type.

Type testing should be performed when a product line is introduced and when it is changed or expanded. As part of a thorough quality-assurance program, a manufacturer may repeat type tests periodically just to be sure that there has been no unanticipated change or "drift" in the product, process, or raw material supply.

Common mechanical tests performed on optical cable include tensile strength, compressive loading (crush), repeated impact loading, torsion loading, flexing, and bending. Temperature cycling is the key environmental type test and is sometimes supplemented with high- and low-temperature bending, water penetration, and freezing tests for outside-plant and indoor/outdoor cables. These tests are conducted in accordance with standard industry evaluation procedures using failure criteria from other industry standards or individual customer specifications. The key to understanding type testing is to understand the various standards and specifications used in designing these tests and how they are applied.

Test methods

Three general categories of standards are used for testing fiber-optic cable. The first of these categories encompasses test methods, which are usually Fiber Optic Test Procedures (fotps) approved by the Telecommunications Industry Association and the Electronic Industries Alliance (tia/eia--Arlington, VA). Test methods provide a standard procedure and apparatus for conducting a certain measurement, but often leave to the user`s discretion the magnitude of the test and the failure criteria. A reference to a test method is, therefore, often meaningless without a companion reference to a loading value of some sort and a definition of "pass" or "fail." The iso/iec series of standards contains documents that are the international parallels to the ansi/eia/tia-455 series used in the United States.

fotps are part of ansi/eia/tia-455, "Standard Test Procedure for Fiber Optic Fibers, Cables Transducers, Sensors, Connecting and Terminating Devices, and other Fiber Optic Components." Each fotp is numbered and can be referenced as a subsection of the parent document. For example, fotp-41, entitled "Compressive Loading Resistance of Fiber Optic Cables," can be referenced as "ansi/tia/eia-455-41."

fotp-41 is a method for evaluating compressive force durability or crush resistance of a fiber-optic cable. It provides detailed drawings of the test apparatus and tells how the load should be applied. But it does not specify the load to be applied to the plates or define the failure level. To be executable, the reference to fotp-41 must include a load value (such as 89 newtons/centimeter or 50 pounds per inch) and a failure criterion, such as "no increase in attenuation greater than 0.5 decibel per kilometer at 1300 nm for multimode fibers while the cable is under load." Another, less-stringent, criterion for the same test might be "no broken fibers after the load is removed." It is clear that defining type testing and failure criteria can become quite complicated, and that is where the second general category of standards plays a role.

Product specifications

The second category of standards for fiber-optic cable consists of documents that detail the specific type tests to be performed, the test methods to be used, and the failure criteria to be applied. These documents can be called detailed product specifications, which have been written to provide the user one-source freedom from the quagmire of test methods, definitions, and failure criteria.

Two sister documents published by the Insulated Cable Engineers Association (icea--South Yarmouth, MA) are very useful as detailed product specifications. ansi/icea s-87-640, "Standard for Outside Plant Communications Cable," and ansi/icea s-83-596, "Standard for Fiber Optic Premises Distribution Cable," cover outside- and inside-plant cables, respectively. In addition to detailed references for test methods, loading, and failure criteria for finished cable, these documents also include similar details for the optical fiber. ansi/icea s-83-596 also includes a summary of the flammability listing requirements from the National Electrical Code (nec). By referencing these documents as appropriate for indoor or outdoor cable, purchasers can ensure that they are specifying a full battery of environmental and mechanical type testing and failure criteria. For the indoor/outdoor cables that have recently become available, a reference to the applicable portions of both ansi/icea standards would be appropriate.

One caution about the ansi/icea documents is that they also contain optical-transmission specifications. The values in the ansi/icea document are less stringent than those required by many applications, and it is recommended that a source other than ansi/icea be referenced for optical performance to meet transmission equipment requirements.

In the United States, the federal government is another source for a detailed product specification. The Department of Agriculture`s Rural Utilities Service (rus) has published 7 cfr 1755.900, "Specification for Filled Fiber Optic Cables." This document provides detailed product specifications for singlemode and multimode fiber and for outside-plant loose-tube cable. The rus specification even outlines requirements for production and type testing as well as data reporting and manufacturers` recordkeeping. Upon request, the rus conducts technical reviews of cable manufacturers` products and programs and places those it finds compliant on a list of accepted products. By specifying that an outside-plant cable must be rus-listed and accepted, an end-user can take advantage of the fruits of this effort at no cost.

A tip on detailed product specifications is to avoid highly specialized documents and requirements such as those developed for the military or particular uses such as shipboard or aircraft. Some telephone industry specifications are also very stringent and include requirements that add unnecessary testing and product cost to cables intended for commercial use. A "mil-spec" or "telco-spec" cable may appear to be better, but in some cases the difference is wasted in commercial applications and can even add unnecessary installation and product costs.

Optical transmission standards

Documents in the third category of standards contain only optical-transmission levels for the cable and do not deal with specifications for mechanical and environmental type testing. An optical transmission standard, when cited in conjunction with a detailed product specification, can be used to fully define both type testing and production testing requirements. The most popular optical transmission standard for customer premises communications in the United States comes from ansi/ tia/eia-568a, "Commercial Building Telecommunications Cabling Standard," which specifies attenuation and bandwidth for multimode fiber and maximum attenuation for singlemode fiber. Other documents that contain optical transmission standards are ansi x3.166, "Fiber Distributed Data Interface (fddi) Physical Medium Dependent (pmd)," iso/iec-11801, "Generic Cabling for Customer Premises," and a forthcoming document called ieee 802.3z, "Physical Medium Dependent Sublayer and Baseband Medium, Type 1000Base-LX (Long-wavelength Laser) and 1000Base-SX (Short-wavelength Laser)."

With the exception of cables listed and accepted by the rus, all manufacturer testing is conducted on the honor system. If a manufacturer says the product has been tested and meets a specification, the buyer can accept the statement at face value or insist on a factory visit to verify that the tests are, in fact, performed. Since this expense is usually beyond the means and desire of all but the largest end-users, the remaining choices are blind trust or third-party verification. Certification to iso-9001, "Quality Systems--model for quality assurance in design, development, production, installation, and servicing," can provide a purchaser some assurance that the cable manufacturer has had to demonstrate to an independent third party that its product meets advertised claims and that customer specifications are subject to credible technical review. rus acceptance, though only pertinent to outside-plant cable, is another form of third-party verification of a manufacturer`s claims. Because the rus specification and list of accepted suppliers is in the public domain, it costs nothing to access and use.

Manufacturer testing should not be overlooked, but from the perspective of the distributor, contractor, and end-user, it can be simplified by using references to two or three appropriate documents: an indoor or outdoor product specification as appropriate (or both for indoor/ outdoor cable) and an optical-transmission standard. For indoor cables, the required listing from the nec completes the package. An end-user can thus write a thorough and comprehensive testing requirements package for optical cable in less than half a page.

Click here to enlarge image

A technician performs in-process optical tests on cable cores with an optical time-domain reflectometer.

Click here to enlarge image

Andrew K. Straw is an applications engineer at Siecor (Hickory, NC).

More in Home
AV over Twisted-Pair