A procedure is described for the measurement of chromatic dispersion of single-mode optical fibers over the 1.0 to 1.7 micrometer wavelength range. The dispersion coefficient at a particular wavelength is determined from the differential group delay between two closely-spaced wavelengths.

In this procedure, a modulated light source is coupled into the fiber under test, and the phase of modulated signal exiting the fiber at a first wavelength is compared with the phase at a second wavelength. Average chromatic dispersion over the interval between the two wavelengths is determined from the differential phase shift, wavelength interval, and fiber length.

The chromatic dispersion coefficient at a wavelength midway between the two test wavelengths is assumed to be equal to the average chromatic dispersion over the interval between the two wavelengths. This assumption is valid provided that the two wavelengths are sufficiently close.

Parameters such as zero-dispersion wavelength lambda-omicron and dispersion slope S-omicron at lambda-omicron may be obtained by curve fitting to dispersion data. See Appendix A for Class IVa fibers, and Annex B for Class IVb fibers. (The classes are defined in EIA/TIA-4920000-A.)

Typical optical sources suitable for this measurement include laser diodes or filtered light-emitting diodes.

This method can be applied to laboratory, factory and field measurements of chromatic dispersion and the wavelength range can be tailored as required.

This test method is suitable for fiber or cable lengths greater than L = 1 km (0.6 mile). (Generally, the length of a cable is shorter than the length of its fibers.) This method may be applied to shorter lengths, although for short lengths the accuracy or repeatability may be