Many distance measuring interferometers use orthogonally polarized beams in their reference and measurement arms. In practice, instrumental imperfections corrupt the polarized beam in each arm with a small amplitude signal from the other arm having the perpendicular polarization. The impact of this polarization mixing on the absolute accuracy of a single frequency laser interferometer is shown to be significantly smaller than on the accuracy of a two-frequency interferometer. Certain signals obtained from a single-frequency interferometer have no phase error due to polarization mixing. Other signals are affected in proportion to the ratio between the small mixing amplitudes and the primary polarization amplitudes. Roughly, those errors in the single frequency device depend on the square of the small mixing ratios while the error in the two frequency interferometer depends on a larger quantity, the sum of these ratios. For a mixing ratio of 0.1, the error in the single-frequency interferometer is about 20 times smaller than in the heterodyne interferometer.
Available at: http://works.bepress.com/paul_davis/1/