This isn't necessarily because "fiber" demands stricter adherence to specifications, but rather because the quality of workmanship is extremely important. Creating a single bend and having to redo the entire run is unacceptable. Even applying a quick burst of air instead of performing a thorough cleaning will become apparent during troubleshooting.
The best way to avoid unnecessary troubleshooting is through formal certification of fiber networks. Both the ISO/IEC and ANSI/TIA standards require two types of testing to certify a fiber network, known as basic/extended and Tier 1/Tier 2, respectively.
Basic certification is performed using an optical loss tester (OLTS) and includes line length and line loss measurements, as well as a polarity test. Enhanced certification requires the use of an optical time-domain reflectometer (OTDR) to analyze the performance of each component of the fiber link. An OLTS test only determines whether the link in question has passed the test. An OTDR test assures the engineer that each component of the fiber link meets or exceeds the limits defined by the standards and, in the event of a failure, pinpoints the exact location of the problem. Event analysis software can be used to automatically define parameters for the engineer to highlight where a loss occurs in a link, while the software identifies and qualifies the event.
Starting with enhanced testing, an OTDR works by sending light pulses down a fiber and then tracking and recording the reflections of the light pulse through a highly sensitive light-detecting unit. The reflections are then mapped to a trace that shows the consistency of the test over a predetermined period of time.
A modern OTDR will measure the length of a given fiber and characterize other link components, such as splices, connectors, cables, and other loss events. This allows an engineer to form a picture of the cable's attenuation uniformity, as well as highlight insertion losses of individual connectors, insertion losses of individual splices, or any other deficiencies that could affect performance and reliability.
Figure 1 shows a series of OTDR traces during advanced troubleshooting for dual-wavelength OTDR measurements at 850/1300 nm or 1310/1550 nm. During troubleshooting, loss events are measured, and these events are then summarized in a table for later report generation.
Know Your Risks.
Although recent research (New Scientist, June 15, 2010) suggests that “dark impulses” may be the future of long-haul fiber optics, there is currently no magic involved in establishing reliable fiber connections. Fiber installations may require stricter adherence to specifications to pass certification than copper installations, and they may seem to require more work at first. However, fiber's characteristics have the advantage that, in most cases, they behave in a binary fashion—either working or not. Having a certificate confirming that testing has been completed guarantees that the link will work, but keeping it operational while the rest of the system is fine-tuned or daily adjustments are made requires disciplined effort.
To receive Level 1 (TIA) or basic (ISO) certification, link losses must meet specifications, typically measured with an Optical Loss Testing System (OLTS) consisting of a light source and a power meter—hence the alternative name LSPM. Standards prescribe the use of an OLTS.
Having a light source at one end and a light meter at the other provides an absolute measurement of loss within a link and is much more accurate and consistent across tests than using an OTDR that uses reflected light energy. Figure 2 shows a sample result of such a measurement.
Loss/length certification provides the fastest possible certification solution for testing two fibers—a transmit/receive pair—at two wavelengths. It measures the fiber length using time-of-flight (TOF) measurement, measures the insertion loss of both fibers at both wavelengths, calculates the loss budget, compares the results to a user-defined test limit, and displays the link's pass/fail status.
The LSPM allows engineers to perform higher-level troubleshooting, such as verifying the basic polarity of a link, which helps reduce the time-consuming task of identifying cables between patch panels. The OTDR is more powerful, as it allows for the identification of marginal components, splices, or connections anywhere along a link. It saves engineers time by accurately identifying faults such as breaks in a fiber optic cable and excessively sharp bends in a cable that could cause signal loss or lead to breaks.
Reference Definition
Before performing the test, it is essential to establish a reference for the loss measurement. A fixed reference point for loss is defined by connecting a test lead (TRC) of one to three meters between the light source and the power meter.
It is important to ensure that each end surface of the CRT is clean and undamaged, as contamination at either end will result in inflated measurements and erroneous losses. An engineer can use a fiber optic or video inspection microscope to inspect and clean the CRT, allowing for accurate measurement and reference definition. For operator safety, never look directly at an active fiber link. Data communication light is invisible to the human eye but can cause damage.
Defining a reference should be viewed as a defined process with its own specific rules. When using a CRT, use only high-quality fiber optic cable with high-performance connectors at each end to obtain an accurate loss measurement reading. The correct type of CRT for defining the reference should be supplied by the manufacturer with specially treated surface-mount connectors that are scratch-resistant. This allows the CRT to be used repeatedly without reducing performance measurement levels.
The process for establishing the loss measurement of a fiber optic link is based on two power measurements. The test reference measurement is necessary to determine the "lossless" power level against which the engineer can compare the power across the link under test. One or more CRTs connect the light source to the power meter, and a reference power measurement is taken to compensate for inconsistencies in the link that could cause measurement errors or inaccuracies.
Industry standards require that multimode fiber measurements be performed using an LED source to create a "full" light on the fiber. For single-mode fiber, a laser source must be used. There are three key elements to the test. First, establish the appropriate test pass/fail limits for the link in question. The next step is to define the test reference against which all subsequent loss calculations are performed. With a reference established and pass/fail limits stored in the test unit, the current link certification check can proceed.
The test will measure attenuation and therefore quantify losses in links, but it will not strictly verify the quality of the entire fiber link. Verification and certification must be extended to the entire link and include all patch cords, connectors, and splices. Furthermore, it must consider all moves, additions, and changes that occur within the network.
Unlike copper, even tightening cables to achieve a neat bundle can cause problems, as overtightening can create new sources of signal loss. One troublesome loss fault, for example, was induced by movement and ultimately resolved by locating an incorrectly mounted connector that was letting in ambient light and interfering with the signal.
Risks
When using an OTDR, there are some things to consider that in the past caused some to view OTDRs with skepticism. Previously, “dead zones,” “ghost events,” and “gain events” were negative terms, but today they are minor problems easily overcome by an engineer equipped with a high-end OTDR. Since the dead zone is defined as the shortest length of fiber that an OTDR can physically detect and measure, the engineer only needs to use a sufficiently long launch fiber when performing certification with an OTDR. This ensures that the device can measure the initial link connection.
Similarly, phantom events, which are echoes of highly reflective events occurring within the link, and gain events are apparent, but actually false, losses within a link that typically occur because two spliced fibers have different refractive indexes. Today's high-end OTDR units are designed to enable and remedy such occasional mishaps when certifying an installation and, as such, help the engineer create a cable installation that is as reliable, efficient, and high-quality as possible.
Even fiber optic networks installed with the utmost care to meet industry standards can experience troublesome failures during deployment. The most common issues in a fiber optic installation are related to foreign objects. Troubleshooting before handover involves checking for tiny dust particles that may enter a connection (not always unavoidable in an industrial or commercial environment) and also checking for smudges or fingerprints on the surface of a connector. Oily deposits are particularly difficult to locate but reduce performance. Also check every port on the equipment, including routers, switches, and network cards, as these also attract dust, dirt particles, and oily deposits.
The checklist for troubleshooting a fiber optic cable link also includes identifying misaligned fibers and excessive connections within a channel. While most copper cable can be wrapped around structures and flexed around network corners, fiber is highly flexible but severely damaged when bent. Be sure to check for pressure on the fiber cable or connectors, both when cabinet or rack doors are open and closed.
Author: Robert Luijten, EMEA Marketing Manager, Fluke Networks.
