Testing the loss of an installed fiber optic cable plant involves checking the loss of the fiber itself, plus any splices and terminations added during installation. Insertion loss—expressed in decibels (db)—is measured by coupling a test source to the fiber and recording how much light is lost when transmitted through the fiber.

It’s easy to see that measuring the loss in transmitted light will include the fiber attenuation and losses at any splices or connections along the way. But what about measuring the loss of connectors on each end of the fiber?

It’s important to know that a single fiber optic connector cannot be said to have loss; loss only occurs when two fibers are mated with connectors. So to be correct, we should say “connection loss.” For most connectors that have a round ferrule design—such as the SC, ST and LC—the actual connection loss will be determined by three pieces: the quality of both of the connectors and the mating adapter that aligns and holds them in place.

Therefore, to include the loss of the connectors on each end of the fiber we want to test, we need to attach them to another connector and a reference connector and include the loss at that point. For this connection, we use a reference test cable. Reference test cables must be made from cable with the same fiber size that is used in the cable to be tested, e.g., 50/125, 62.5/125 or single-mode.

I often hear from installers who are getting consistently bad results. That usually occurs from mismatched fibers. Mating different multi-mode fibers causes about 2–4 dB higher loss in one direction (62.5 to 50/125 fiber) and practically no loss in the other direction. Mixing multimode and single--mode fibers causes about 20 dB loss.

The connectors on a reference cable must be compatible with the connectors on the cable being checked as well as those on the test equipment. SC and ST connectors have the same size ferrule, so a hybrid-mating adapter can be used to mix those two types.

Now, the increasing usage of LC connectors with a smaller ferrule causes a problem because one cannot get hybrid mating adapters to properly align the different size ferrules of ST/SC and LC connectors. The solution requires using hybrid reference cables, with a SC or ST to mate to the test equipment and a LC on the other end to mate to the cable under test. Such hybrid cables make setting a 0 dB loss reference difficult, a topic we’ll cover in more detail next month.

The most important issue with reference cables is that the connectors need to be high quality, since the losses you measure will be dependent on the quality of these reference connectors. Over the years, standards-making groups have tried to specify reference cables more tightly, but nothing beats simply having low loss connectors. If the patchcords being used for reference cables have losses of less than 0.5 dB when the reference connectors are mated to each other, they are generally considered acceptable. If the losses are lower, say 0.2 to 0.3 dB, that’s much better because these better connectors should have lower loss when mated to the connectors on the cable being tested.

Even with the best reference connectors, you must keep them clean. Dirty connections are high loss connections, since dirt is large compared to the small size of an optical fiber’s core. In addition, dirt may scratch the connectors, causing permanent damage. Every test kit must contain a cleaning kit, either wet or dry style, and it should be used before every test is conducted.

Finally, reference connectors and mating adapters wear out. The process of mating can cause scuffing on the fiber ends and the alignment sleeve in the mating adapters (which will affect the connection loss). Ensure you use the best mating adapters. Inexpensive ones with plastic alignment sleeves wear out in a few uses and deposit a black plastic dust on the connectors. Mating adapters qualified for single-mode applications use either metal or ceramic alignment sleeves that are good for hundreds of matings.Most connectors are good for about 500 matings before they need replacing.

Check the ends of the connector ferrules with an inspection microscope after cleaning to determine if the surface is scuffed or scratched. It can be repolished, but it must be done on diamond polishing film on a soft rubber surface to ensure the end profile of the ferrule is not ruined. Usually it’s easier to replace the cable.

HAYES is a VDV writer and educator and the president of The Fiber Optic Association. Find him at www.jimhayes.com.