Toward a nonlinearity model for a heterodyne interferometer: not based on double-frequency mixing.

Residual periodic errors detected in picometer-level heterodyne interferometers cannot be explained by the model based on double-frequency mixing. A new model is established and proposed in this paper for analysis of these errors. The multi-order Doppler frequency shift ghost beams from measurement beam itself are involved in final interference leading to multi-order periodic errors, whether or not frequency-mixing originating from the two incident beams occurs. For model validation, a novel setup free from double-frequency mixing is constructed. The analyzed measurement signal shows that phase mixing of measurement beam itself can lead to multi-order periodic errors ranging from tens of picometers to one nanometer.