Two measurements of optical power required

Compared to Category 5e or Category 6, fiber optics wire is easy to test. A “flashlight test” can make sure the fiber has continuity and is correctly identified. A fiber optic power meter and test source are used to measure the light loss and the loss is compared to the expected loss to insure the installation was done correctly. In short cables, most of the loss is from connectors, not the fiber itself.

The test equipment required can be a simple, inexpensive power meter and LED source or it can be an expensive, automated optical loss test set (OLTS) that prompts you through all tests and stores data for printing “certification reports.”Both provide equally accurate results, so, if necessary, the choice can be made based more on economics than the quality of measurements.

Now lets focus on some new issues, like the differences in test procedures affecting the accuracy of the measurements and using other test instruments like OTDRs.

Test procedures

Testing loss requires making two measurements of optical power: First, the zero-loss reference power (the output of the test source) and second, the lower power level once the cable being tested is attached (See Figure 1). The difference between these two power levels is the loss of the cable tested. The confusing issue is how the reference power is measured.

The number of different designs in fiber optic connectors as well as some differences in test philosophies has led to three different methods of setting the reference power for loss, using one, two or three reference cables. The EIA/TIA 568 standard for premises cabling calls for a one-cable reference using the cable marked “1” in the figure between the meter and source.

That method works fine for most fiber optic connectors like the ST or SC, but is not possible on many duplex connectors like the small form factor MT-RJ. Since most power meters or OLTSs cannot accept both male and female MT-RJ versions, a two-cable reference (using cables 1 and 2) or three-cable reference is required.

The differences in measurements obtained using these reference methods includes both the total loss measured and the precision (variability) of the measurement. A single-cable reference includes the loss of the connectors on both ends of the cable under test. A two-cable reference includes one connector in the reference measurement, so the total loss measured will include only one of the two connectors on the ends of the tested cable. On a three-cable reference, the loss is lower by the loss of two connections included when making the reference measurement, so effectively no connectors on the tested cable are included.

Besides changing the loss values measured, referencing with two or three cables reduces the precision of the measurement or makes it less consistent. The loss of the connections included in the reference are unknown, making the final measurements have a higher variability, so the final measurement is more uncertain.

How can you deal with this? It’s necessary to use the technique that allows making the measurements, so you have little choice. But you should review this with your customers so they know how the measurements were made and are aware of the implications on the acceptance criteria. Compared to EIA/TIA 568, the two-cable reference should have 0.3-0.5 dB lower loss, while the three-cable reference method will give losses 0.6-1.0 dB less.

Can you test loss with an OTDR?

The OTDR (optical time domain reflectometer) is often described as “radar” for fiber. It measures length, finds the location of connectors and the end of the fiber and can estimate the loss of the cable. But it measures loss in a different way than the meter and source which mimics the transmitter and receiver in an actual datalink.

The OTDR measures light scattered in the fiber to create a picture of the losses in the fiber. OTDR-measured loss is always significantly less than the loss measured by the meter and source and may lead to situations where the cable plant will pass loss tests but not work with the system.

A better use for the OTDR is to verify that the installation did not cause problems with the cable itself. Bends in the cable that cause stress loss in the fibers can be located with the OTDR and fixed if necessary. This is the big use of the OTDR for telco applications, in addition to finding faults for restoration, an extremely rare use in premises cabling.

A new testing standard

The EIA/TIA 568 committee is adding a new document to the 568 B.3 fiber standard to better explain fiber optic testing. TSB-140 will cover visual fiber tracing and fault location, meter and source loss testing and OTDR testing. As the representative of The Fiber Optic Association, we worked with committee editor Bob Jensen on this document and are quite impressed with the result. When published, we suggest you read it to add to your understanding of fiber optic testing. EC

HAYES is a VDV writer and trainer as well as president of the Fiber Optics Association. Find him at