The validity, efficiency and stability of two different nonlinearity correction methods proposed byHeydemann and Wu respectively and applied for the sine-approximation method (SAM) of homodyne version in primary vibration calibration is fully investigated through various comparison tests on the basis of worst-case simulated digital photodetector signals with limited precision. A novel simplified algorithm for phase unwrapping of the SAM is proposed and used in all tests of this paper. Experimental results demonstrate that a correction is absolutely necessary to compensate systematic errors at high frequency for primary vibration calibration, and that Heydemann correction depends heavily on the effective resolution of original signals and has a more predictable performance while Wu correction is generally more efficient. It is also shown that a minimum effective resolution of >=10 bits of waveform recorder is necessary for a successful application of Heydemann correction at frequency up to 10 kHz. The theoretical analysis of the correction effect is also described in the paper.
展开▼