| Summary: | 博士 === 國立臺灣大學 === 機械工程學研究所 === 90 === Displacement measurement capabilities of high sensitivity and high ac-curacy have become one of the most important techniques to pursue research in sub-micrometer and nanometer fields. To engage in micro-machining research of VLSI and the precision machining study of optical components, the need to have high resolution and high accuracy displacement sensors or sensing systems is becoming even more indispensable. As even a very small load can easily induce micrometer scale displacement in a typical specimen, non-contact measurement is almost a must for sub-micrometer level of measurements.
The laser encoders studied in this dissertation are one type of the grating interferometers with the linear grating scale or the radial grating scale. By virtue of the non-contact optical approach, laser encoders cleverly convert the grating pitch as their measuring scale from the laser wavelength. This feature has thus made this type of systems is immune to the environmental disturbances. The newly developed Diffractive Laser Encoder System (ab-breviated as DiLENS) was developed. It is shown that DiLENS is com-posed of three sub-systems, which includes a 1-x telescope design, a diffrac-tion efficiency optimized grating, and a circularly polarization interferometer. The adaptation of the 1-x telescope assured that a twice diffracted beam could trace along an optical path that is parallel to the original one, which has made the head-to-scale tolerance of DiLENS about 6 to 20 times better than all of the leading encoders of the same class. It was also shown that the DiLENS optical head design possess the capabilities that it is immune to the effect of differences in polarization diffraction efficiency between the in-cident TE and TM waves. To optimize the performance of DiLENS, the grating scale geometry adopted was optimized to achieve maximum diffrac-tion efficiency. It was found that the optimized geometry for the grating is a sinusoidal surface relief grating with 190 nm deep and this optimized grat-ing scale can raise the intensity of the interference signals 6 to 20 times bet-ter when compared to the Canon encoder of the same class.
The circular polarization interferometer can transfer the sig-nal-processing domain from the high frequency laser source to the low fre-quency Doppler frequency shift, which represents the displacement informa-tion. The intrinsic reduction factor for the measuring accuracy was found to come from the inclined elliptical signals of the circular polarization inter-ferometer adopted. From the analysis, the causes are the polarization transmission axis misalignment of the polarizers, and the misalignment be-tween the photodetectors with respect to the interference signal was found to be the primary factors. Using alignment approach to minimize the align-ment error has thus been proposed within the dissertation to improve the overall system performance.
It was also found that conical diffraction phenomenon has to be used to analyzing the diffraction behaviors of DiLENS. It was identified that this behavior is not equivalent to the well-known two-dimensional grating equation prediction. Combining the conical diffraction and the ray-tracing program, a useful tool to analyze the head-to-scale alignment tolerance was developed.
As radial grating can introduce significant amount of wavefront aberra-tion to the diffracted beam, which is mostly astigmatism, an analysis model for analyzing the diffracted wavefront of the radial grating is also proposed. This model can estimate the variation of the diffracted wavefront. The re-sults obtained from this model match well with the Fourier optics calculation and the experimental observations.
In summary, to verify the performance of the newly developed configu-ration for DiLENS, the HP5529A interferometer was adopted as the calibra-tion tool. The obtained average measurement accuracy at regular labora-tory environment with a general disturbance is around 35.4 nm, and the av-erage of the standard deviation is 15.3 nm. By slightly controlling the temperate drift, the experimental accuracy was measured to be 33.71 nm and the average measurement repeatability is 4.48 nm.
|
|---|
