Permanent Magnet Brushless DC (PMBLDC) motor is a new generation of converter fed machines (CFMs) becoming more and more popular. PMBLDC motors are extensively used as industrial motors due to its fast dynamic response, high power density, large torque to inertia ratio, high efficiency with increased reliability, less noise, long life, silent operation, compact form, low maintenance and better controllability. For innumerous applications, PMBLDC motors are used as replacement for AC motors. PMBLDC motor is a trapezoidal shaped back EMF permanent magnet synchronous motor with solid state commutation system. The absence of a commutator and brushes or slip rings in the PMBLDC motors reduces maintenance needs and raises mechanical reliability. In solid state commutation system, rotor position information must be needed for proper commutation sequence with proper control algorithms. Motor performance degrades due to improper control action. In this study, an adaptive PI speed controller based field oriented vector controlled current fed delta modulated PMBLDC motor drive is designed. An adaptive PI controller is proposed based on motor speed error. Scalar control of this PMBLDC motor is also performed at direct axis current component is equal to 1.0 and at direct axis current component equal to quadrature axis current component. The performance of these control drives is compared under different conditions. In this thesis, a novel approach to enhance the torque handling capacity of a PMBLDC motor is invented. The novelty is that, torque handling capacity of a PMBLDC motor can be increased up to 25% from the conventional 1200 conduction square wave current fed drive by only changing the pattern of reference current of a field oriented vector controlled PMBLDC motor drive. Toque handling capacity has to be enhanced without exceeding the maximum current rating of the PMBLDC motor. The performances of trapezoidal, square and sinusoidal current fed field oriented vector controlled PMBLDC motor drives are compared on the basis of response time, load torque handling capacity, dynamic speed and load torque changing condition, settling time of the system and torque pulsation. In conventional PMBLDC motor, for proper commutation of the phase currents rotor position information must be obtained from the position sensors. But there are some vital disadvantages of position sensor including high cost, installation difficulty of mechanics, and poor reliability. Moreover, misalignments in position sensors, running in extreme ambient conditions, or electromagnetic interference introduce error in the position information. Instead of rotor position or speed sensor, two novel algorithms are proposed in this thesis to estimate the rotor position and speed to perform the operation of position sensorless field oriented vector control of PMBLDC motor. For algorithm 1, both rotor position and speed of the motor are determined from the estimated flux linkage. For algorithm 2, only rotor position of the PMBLDC motor is determined from the estimated flux and the speed of the motor is detected by the estimated developed electromagnetic torque and power equation. The performance of flux estimation algorithm, rotor position estimator, speed estimator is depicted by the comparative study between the actual and estimated flux, rotor position and speed respectively. The performance of the proposed algorithms for sensorless operation is also justified through the sensorless trapezoidal current fed field oriented PMBLDC motor drive. In this drive, both the advantages of position sensorless operation and enhanced torque handling capacity are incorporated due to establish a novel high performance control drive for PMBLDC motor. A novel position sensorless two phase conduction direct torque controlled PMBLDC motor drive is also proposed in this study. Without considering flux control, two-phase conduction direct torque control (DTC) of a PMBLDC motor on the basis of electromagnetic torque, rotor position and speed estimation is pictured in the proposed drive system. The performance of the proposed position sensorless direct torque controlled PMBLDC motor drive is compared with the performance of conventional position sensored two phase conduction direct torque controlled PMBLDC motor drive in terms of starting, dynamic speed and load torque changing characteristics.

Spectroscopic measurements are widely used for detecting many physical and chemical environment as they can reach to extremely low level of the target analytes with high selectivity and accuracy. The recent generation of smart sensing applications are increasingly relying on the multiple channel measurements and ubiquitous operation of spectrometers. However, the traditional spectrometers with multi-channel capacity use extra optical elements including source spiting hardware and detectors that incur additional cost and increased size. In this thesis, smartphone-based ultra-low-cost, field-portable novel multichannel spectrometers are demonstrated that utilize multiple diffraction orders of an ordinary dispersive element. This allows high throughput spectral measurements with small form factor utilizing the in-built sensors and advanced processing power of smartphone. As a proof-of-concept, a dual channel spectrometer is reported by exploiting the first diffraction orders (m = +1 and -1) which have the highest intensity and identical in nature. The first cannel is used to measure light absorption by a dedicated sample placed on its path whereas the second channel is used for real-time referencing. In the conventional double-beam spectrometers, reference measurement is typically achieved with beam splitting optics or a second monochromator. This proposed instrument utilizes smartphone’s CMOS camera to capture the diffracted flash LED light of smartphone into multiple orders simultaneously, producing a 2D image. Each of these diffraction orders can be utilized as a single optical channel. A customized application (app) is developed to process each diffraction image into spectra and 3D printing is used to fabricate the entire instrument prototype. For the first time, nonlinear wavelength distribution along the diffraction direction is appropriately addressed in this instrument to minimize the errors in spectral measurements. It produces a tunable spectral resolution of δλ ~ 1.6 to 3.0 nm/pixel over the bandwidth of Δλ = 300 nm in the visible region. Absorption spectra for a laser dye (Rhodamine B) and a pH-responsive buffer (bromothymol blue) have been measured with comparable accuracy with respect to the standard spectrometers. This offers a low cost (< 30 USD) portable instrument comparable to conventional double beam benchtop instruments in performance but more rugged for field use. One of the shortcomings of the dual channel spectrometer is that it does not consider the refractive index of the sample and reference solutions as well as for the containers (glass or quartz or plastic cuvettes). It is often challenging for the conventional spectrometers as it can affect the absorption due to the change in optical path length for refractive index variation. To address the limitations, an accurate wavelength calibration was performed by considering the non-linear pixel increment over the spatial region and dispersion associated with the effective refractive index variation of a sample and its holder. The latter option offers refractive index sensitive spectral measurement by the instrument – an added capacity where it can be used as differential refractometer. In the presence of a significant variation of refractive index, the spectrometer measures different vegetable oils’ (olive and soybean) spectra at different adulterated level with accuracy 97.3%. Despite this successful approach, two key challenges remain in further increasing the number of channels: (1) size and light intensity variation between the successive diffraction orders and (2) limited field-of-view (FOV) of the smartphone camera for the increased number of channels where extension in the 1D plane is necessary. To overcome the limitations, further advancement of the instrument has been made where a novel concept of diffraction grating holographic imaging is applied to converge and capture the multiple diffraction orders of two-dimensional (2D) grating within the limited field-of-view of the smartphone’s camera detector. This also eliminates the requirement of collection, collimating and convergence optics as used in conventional spectrometer systems. As proof-of-concept, a customized 2D grating composed of two identical thin-film 1D gratings orthogonally stacked together is utilized to diffract the light rays emanating from a broadband visible source and produce four identical 1st orders dispersive diffraction in two orthogonal axes. The smartphone camera captures the diffraction orders within its solid angle – a quad-channel spectrometer exploiting the 2D properties of a CMOS detector is demonstrated. Mindful of losses, further increases in channel numbers is possible by simply increasing the number of gratings (N) within the stack spatially separated by an angle θ_N=(180°)⁄N between successive grating axes. The instrument offers an ultra-low-cost (< 14 USD), lightweight (~ 175 gm), high throughput, rugged and small form factor multichannel spectrometer ideal for field use as well as a laboratory benchtop alternative. To evaluate the performance of the instrument, three chemically and biologically significant parameters are measured simultaneously: - (a) iron concentration [Fe], (b) chlorine concentration [Cl], and (c) pH of the drinking water. The smartphone data of [Fe], [Cl], and pH measurements are in good agreement with the standard meter readings and have an average accuracy of 97.5%, 96.7% and 98.3% respectively. Moreover, spectral detection range of the smartphone spectrometer is extended from visible band to near infrared band (∆λ = 400 – 799 nm) to detect chlorophyll-a in olive oil samples by tailoring the camera module of smartphone. Thus, smartphone-based multichannel spectrometers have been instrumented for ultra-low-cost concurrent spectral measurements of multiple samples (analytes) in the field. The entirely new concepts such as utilization of multiple diffraction orders, 2D diffraction grating imaging technique, nonlinear wavelength distribution along the diffraction direction, and refractive index sensitive spectral detection have been proposed to design, fabricate and calibrate these instruments. The proposed novel concepts especially 2D diffraction grating imaging technique not only contribute to the fundamental knowledge of smartphone-based multichannel instrumentation but also it will break up the traditional trend of commercial spectroscopic instrumentation.