A novel accurate calibration-based phase-shifting projected fringe profilometry (Calibration-based PSPFP) for finding the absolute shape of objects is proposed. In addition to a tremendous savings in time, the benefits of using Calibration-based PSPFP also include greatly reduced environmental vulnerability. Since Calibration-based PSPFP employs a sinusoidal fringe pattern to perform the phase-shifting algorithm, the quality and accuracy of the sinusoidal fringe pattern becomes critical. To evaluate the performance of this measurement scheme and reduce the phase error caused by projected fringes, fabrication of various digital sinusoidal fringe patterns is necessary. Thus, we propose a method to fabricate various digital patterns. Application of a 2-D fringe pattern for Calibration-based PSPFP is proposed as well.

In this paper, we propose tag antennas and reader antennas, which are suitable for radio frequency identification system. Antennas operating in the ISM and the UHF bands are designed. The antenna dimensions are minimized by employing the meander-line technology. Adopting the concepts of Yagi-Uda antennas, We use parasites to increase the antenna gain and to create an end-fire radiation pattern. The fact that the design comes without a ground plane can achieve cost down of the antenna. Higher antenna gains allow a greater effective read-zone distance. Also, the directional radiation pattern eliminates the degradation of performance due to multi-path loss. The design of tag antennas shows good performances in terms of both cost and antenna characteristics for radio frequency identification systems. The reader antennas are designed. Etching a slit in a square patch can achieve the requirement of circular polarization. It can effectively reduce the polarization mismatch when the tags are point to different orientation. Therefore the reliability of system operation will be improved. The antenna performance is measured by the experiments. Moreover, it will be compared with the simulation. In the latter part of paper, the simulation of the power coupling will be carried out. According to the result of power coupling simulation, we analyze and discuss which factor will affect the reliability of the identification. Identification failure will be effectively avoided and a robust identification system can be built if the analyses are properly employed.

Silicon and silicon nitride thin films were growth on Si and glass substrates at room temperature by ratio-frequency (r-f) magnetron sputtering. The electrical characteristics of the silicon nitride films were characterized using I-V and C-V measurement under different growth condition, including r-f power, nitrogen partial pressure, and hydrogen partial pressure. Minimum current leakages for MIS structure as low as 2¡Ñ10 A/cm were obtained at 1 MV/cm electrical field with hysteresis voltage about 2V. The root-mean square surface roughness of the silicon nitride film is less then 1nm. In addition, silicon nitride capacitors with indium-tin-oxide as electrodes were fabricated. Silicon thin films prepared by R.F. magnetron sputtering at room temperature are amorphous. The measurements on the variation of the photo-conductivity were used to characterize the characteristics of the Si film.

We transport the standard time signal by using the single-mode fiber (SMF) media through the synchronous optical network (SONET)/ synchronous digital hierarchy (SDH) data signal, in which the time signal generated from the atomic clock was inserted into the unused bytes in the overhead of the SONET/SDH OC-3 155 Mb/s frame. Based on the NIST MODEM architecture, which was developed by the National Institute of Standard and Technology in the States, the time transfer system is composed of the control unit of the OC-3 transceiver module and the time-signal transmiting/receiving unit. We design and fabricate the required time-signal transmitting/receiving circuits and the control and monitoring circuits. The personal computer (PC) can read/write and monitor the operation status of the optical transceiver module through the control circuit, and the 1-PPS (pulse per second) time signal can therefore be read/written from/into the transceiver module. The frequency stability of about 2.6E-16 within one day can be achieved for this 1310-nm optical-fiber-based time transfer system over a 35-km SMF link without using electronic regererator. Such stability performance is comparable with that of today¡¦s GPS-based time transfer system.

In this proposal, an advanced local oscillator with high resolution, low phase noise and fast switching characteristics is designed for wireless communication applications. The circuit is based on fractional-N frequency synthesis technique in which the use of delta-sigma modulator can remove the fractional spurs effectively. The mechanism in regard to fractional spurs and phase noise for a fractional-N frequency synthesizer will be studied and simulated by developing proper mathematical models. In the implementation of the local oscillator, the analog circuit includes a 1000-1033 MHz VCO, crystal oscillator and loop filter. The digital circuit includes a phase frequency detector, dual modulus divider and 3rd order delta-sigma modulator. At first a FPGA will be used to prototype the digital circuit. The final digital circuit will be implemented in a CMOS process and require 3V operation with low current consumption. The design specifications include that under 1 KHz resolution the phase noise levels are less than -90 dBc/Hz at frequency offets within a loop bandwidth more than 100 KHz. Spurious components are less than -90 dBc/Hz and switching time is less than 1 ms over a 30 MHz tuning range.

A complete methodology has been proposed to model and evaluate IC packages in the high-speed digital and radio-frequency applications. The package types that are studied in this dissertation include BGAs, TSSOPs and BCCs. In characterization of BGAs, both frequency-domain and time-domain techniques have been applied and compared to each other. It was found that the best strategy was to find a rough coupled transmission-line model in the time domain and refine it through the optimization scheme in the frequency domain. Equivalent lumped model has been further derived from the coupled transmission-line model using the concept of distributed parameters. For RFIC applications, the electrical model of BCC, one type of lead-frame CSP, has been established based on the frequency-domain technique. To evaluate the package performance, an on-chip 50-ohm microstrip line housed in the package has been investigated. The insertion and return losses were analyzed and measured. Excellent agreement has been observed up to Ku band. The package acts as a low-pass filter to cause a cut off for the line above a certain frequency, which was predicted successfully from the established package model. The simulation results also show that BCC exhibits higher cut-off frequency and lower insertion loss in the passband when compared to TSSOP, one of the currently most popular RFIC packages.

­^¤åºK­n¡G In this thesis, a two-step de-embedded techniques was applied to measure the important parameters, ft and fmax , of the heterojunction bipolar transistors(HBTs). The same technique was also used to measure the wide-band S parameters for modeling and evaluating the bump chip carrier(BCC) packages. In the simulation, the Ansoft HFSS simulator was used to calculate the insertion and return losses for some bare and packaged test chips. Comparison between simulated and measured results has been discussed in detail to illustrate the applicability of the HFSS simulator.

An ISM-band frequency synthesizer is introduced in this thesis. The technique allows digital phase/frequency modulation to be achieved in a closed phase locked loop (PLL) without mixers and D/As. According to the simulation results using ADS, quantization noise will be filtered by the PLL bandwidth. But the data rate is also bounded by the PLL bandwidth. Two key components of this closed-loop architecture, Gaussian filter and delta-sigma modulator have been implemented by FPGA together with the Qualcomm Q3236 synthesizer IC.

Abstract Ultrasonic technique is widely used in medical application and food industry; however, much work has focused on harmful biological effects of tissues and cells by ultrasound exposure; only little information is mentioned about the beneficial biological effects of ultrasound. Thus, the objective of this thesis is to observe the beneficial biological activities of Paramecium induced by ultrasound exposure. Since the structure or biophysical will stimulated into the interaction between ultrasound and living matter. When multi-cell creature is exposed by ultrasound, this reaction will lead the biological effect becomes complex. Therefore, a single cell creature is chosen to understand the beneficial biological effects induced by ultrasound exposure. The oscillation of the monad in response to the ultrasound radiation is simulated using Rayleigh-Plesset¡¦s bubble activation theory. The resonance frequency of the unicellular creature is then calculated. The diffuse field theory of Sabine is used to create a uniform sound field for the radiation experiment. The images obtained from a microscope can be analyzed and recorded by a personal computer. The number of cells was counted in the haemacytometer after irradiation. The calculated resonant frequency range of the Paramecium shape is 0.2~0.27 MHz. The relative growth rate of the Paramecium suspensions exposed to ultrasound was about 20% slower than that of unexposed sample. It was found that the phenomenon of inhibition and destruction appeared during irradiation. Also, the growth curve is retarded during the period. The resonance frequency of the Paramecium vacuole is 0.5~1.09 MHz. The maximum relative growth rate was increasing 18% with 1MHz irradiation.

Abstract In this thesis, the variation of the natural frequencies of a drill during the high speed drilling processes has been investigated. The Pro-E and MARC finite element packages were used to formulate the twisted drills. Two numerical methods,i.e. Lanczos and Inverse Power Sweep, have been employed to solve the corresponding eigen value problems.The effects of following parameters,e.g. drilling speed,axial load and drilling conditions on the natural frequencies of a drill was simulated numerically and measured experimentally. Four springs with different spring stiffness attached on the drill tip is used to simulate the drill as it penetrated into the workpiece. The variation of the drilling responses under different drilling speeds have been measured. Results indicate the measured results agree very well with the measured data.Frequency spectrum distributions indicate the drilling frequency and the twice drilling frequency are the key response frequency of the axial thrust force, and the drill of frequencyisthe only key frequency which dominates the torque response.Experimental results also show the thrust forces estimated from the empirical equation have good correlation with these measured data.