Working Prim and Proper: Testing and measuring fiber optic power and loss budgets

When you design and build a fiber optic network, the goal is to have it work properly with the communications networks you need. Communications networks generally specify only two parameters for the cable plant: the type of fiber (single-mode or multimode, and maybe a special grade of fiber) and the maximum length or decibel (dB) loss of the cable plant.

For the cable plant designer and installer, this is very important information that requires consideration from the beginning to the end of a fiber optic project. The specification of the decibel range of the communications equipment is called the power budget. It’s a simple function of the difference between the output of the transmitter and the power required at the receiver, expressed in dB.

The designer plans the route of the cable plant to determine the actual fiber length required, looks at splice points to join lengths of cable or add in drops, determines the number of splices and locates any connections at patch panels and each end. Factoring in the loss of the fiber over that distance and the loss of splices and connections creates an estimated loss for the cable plant, which is what we call the loss budget.

The designer then compares the power budget to the loss budget, and hopefully the power budget is bigger, with the difference being the system margin. If the margin is adequate—usually around 3 dB for outside plant (OSP) networks—the design is good. If not, one needs to either find ways to lower the loss of the cable plant, which can be difficult, or find communications electronics with a larger power budget.

Very high-speed or long-distance OSP networks and high-speed premises networks on multimode fiber may have issues with fiber bandwidth that also factors into the loss budget. This may necessitate choosing a special grade of fiber that must be designed into the cable plant before construction begins.

When the design is completed and project documentation turned over to the construction crew, it becomes their responsibility to build the network to specifications.

Once the cable plant is installed, it must be tested and documented. The loss of each fiber must be tested end-to-end and the results compared to the loss budget. OSP cable plants are usually tested with an optical time-domain reflectometer (OTDR) that can test loss and verify each splice in the link. Insertion loss with a test source and power meter is also recommended, but for single-mode fiber it may not be necessary if bidirectional OTDR testing is done. Premises cabling is shorter in length and rarely has splices, so only insertion loss testing is necessary.

Before the techs test the link, they need to know the loss budget as calculated by the designer. When they compare their results to the design specifications, they will know if their installation was done properly. If loss is too high, now is the time to troubleshoot the problems and fix them.

Use good judgment

A question I often get from techs and cable plant owners is how you decide whether to pass or fail a fiber that is close to the loss budget. I have been told by installers that customers wanted to fail fibers that were only 0.1 dB more than the loss budget.

Comparing loss budgets to test results requires understanding the uncertainty of both. Loss budgets are simply estimates of loss and they depend on the estimates you use for component loss. Likewise, test results have an uncertainty of several tenths of a decibel or more, depending on the length of the link being tested and the test method. Even the power budget for the equipment has variation, since not all transmitters and receivers are the same.

When comparing test results to loss budgets, use some judgment. If the loss budget is 2.15 dB and the test result is 2.17 dB, don’t immediately consider the fiber tested to be a failure. The loss budget could be uncertain by several tenths of a decibel, as could the measurement.

If the measured loss exceeds the loss budget by a significant amount on a number of fibers, troubleshoot starting with visual inspection and cleaning of the connectors, then retesting. Most installation and testing problems are traceable to dirty connectors, but bad connectors or splices will require replacement.

About the Author

Jim Hayes

Fiber Optics Columnist and Contributing Editor

Jim Hayes is a VDV writer and trainer and the president of The Fiber Optic Association. Find him at www.JimHayes.com.

Stay Informed Join our Newsletter

Having trouble finding time to sit down with the latest issue of
ELECTRICAL CONTRACTOR? Don't worry, we'll come to you.