In this procedure, a modulated light source at a given wavelength is coupled into the component under test, and the phase shift of the modulated signal exiting the fiber at that wavelength is recorded with respect to the original modulation signal. The phase shift can be converted directly to RGD in the device. By collecting RGD data at various wavelengths, the chromatic dispersion of the component can be determined at these and other wavelengths from the derivative with respect to wavelength of the relative group delay.

This FOTP does not apply to components that have in excess of 10 dB of polarization dependent loss (PDL). The main intent of this FOTP is to determine the average group delay and chromatic dispersion of the component and moreover to isolate the RGD from all polarization sensitivity in the device.

Note that in the case of fiber pigtailed components, the pigtails may give a large absolute (standing) delay but also some relative delay, over the component test wavelength range. In most cases of component measurements, only the relative group delay is of interest, so it is important to account for any RGD in the leads, and to ensure the standing delay is stable over the measurement time. Any changes in the standing delay and any RGD in the leads will be evident in the test results.