etd AT Indian Institute of Science
https://etd.iisc.ac.in:443
The etd@IISc digital repository system captures, stores, indexes, preserves, and distributes digital research material.2021-02-28T07:17:33Z0.5V Subthreshold Region Operated Ultra Low Power Passive Sigma Delta ADC in 180 NM CMOS Technology
http://etd.iisc.ac.in/handle/2005/4318
0.5V Subthreshold Region Operated Ultra Low Power Passive Sigma Delta ADC in 180 NM CMOS Technology
Satyadev, Singh Kamlesh
With increasing demand of IoT devices, medical devices, remote sensors; the design of low power analog interface is becoming focus. Generally, for low frequency applications the Sigma Delta ADCs are used due to their very good resolution capability for such interfaces. Hence extensive work is being done to design ultra-low power Sigma Delta ADC. Most of the work has been done on optimizing loop filter design both in terms of architecture and its basic building element, op-amps. Recently, one of the prime focus of such research is Passive Sigma Delta ADC, where the loop filter is implemented with passive elements instead of active elements like op-amp.
In this work, a subthreshold region operated Passive Sigma Delta design has been explored. The thesis discusses a different analytical approach to analyze passive SDM ADC than the usual circuit level analysis used traditionally. The Simulink modeling of a passive SDM ADC was addressed to study block level performance. The circuit level implementation was carried out in Cadence environment. Both pre-and-post layout level simulations were conducted.
The passive SDM ADC designed in this work has a Sampling frequency of 10MHz, with a signal BW of 10KHz. An ENOB of 10.4 bits is achieved at power dissipation of only 4μW. The proposed ADC has very competitive FOM (Figure of Merit) in comparison with published literature.
1-D And 3-D Analysis Of Multi-Port Muffler Configurations With Emphasis On Elliptical Cylindrical Chamber
http://etd.iisc.ac.in/handle/2005/2218
1-D And 3-D Analysis Of Multi-Port Muffler Configurations With Emphasis On Elliptical Cylindrical Chamber
Mimani, Akhilesh
The flow-reversal elliptical cylindrical end chamber mufflers of short length are used often in the modern day automotive exhaust systems. The conventional 1-D axial plane wave theory is not able to predict their acoustical attenuation performance in view of the fact that the chamber length is not enough for the evanescent 3-D modes generated at the junctions to decay sufficiently for frequencies below the cut-off frequency. Also, due to the large area expansion ratio at the inlet, the first few higher order modes get cut on even in the low frequency regime. This necessitates a 3-D FEM or 3-D BEM analysis, which is cumbersome and time consuming. Therefore, an ingenious 1-D transverse plane wave theory is developed by considering plane wave propagation along the major-axis of the elliptical section, whereby a 2-port axially short elliptical and circular chamber muffler is characterized by means of the transfer matrix [T] or impedance matrix [Z]. Two different approaches are followed: (1) a numerical scheme such as the Matrizant approach, and (2) an analytical approach based upon the Frobenius series solution of the Webster’s equation governing the transverse plane wave propagation. The convective effects of mean flow are neglected; however the dissipative effects at the ports are taken into account. The TL predicted by this 1-D transverse plane wave analysis is compared with that obtained by means of the 3-D analytical approach and numerical (FEM/BEM) methods. An excellent agreement is observed between this simplified 1-D approach and the 3-D approaches at least up to the cut-on frequency of the (1, 1) even mode in the case of elliptical cylindrical chambers, or the (1, 0) mode in the case of circular cylindrical chambers, thereby validating this 1-D transverse plane wave theory. The acoustical attenuation characteristics of such short chamber mufflers for various configurations are discussed, qualitatively as well as quantitatively. Moreover, the Frobenius series solution enables one to obtain non-dimensional frequencies for determining the resonance peak and trough in the TL graph. The use of this theory is, however, limited to configurations in which both the ports are located along the major axis in the case of elliptical chambers and along the same diameter for circular chambers. The method of cascading the [T] matrices of the 2-port elements cannot be used to analyze a network arrangement of 2-port elements owing to the non-unique direction of wave propagation in such a network of acoustic elements. Although, a few papers are found in the literature reporting the analysis of a network of 2-port acoustic elements, no work is seen on the analysis of a network of multi-port elements having more than two external ports. Therefore, a generalized algorithm is proposed for analyzing a general network arrangement of linear multi-port acoustic elements having N inlet ports and M outlet ports. Each of these multi-port elements constituting the network may be interconnected to each other in an arbitrary manner. By appropriate book-keeping of the equations obtained by the [Z] matrix characterizing each of the multi-port and 2-port elements along with the junction laws (which imply the equality of acoustic pressure and conservativeness of mass velocity at a multi-port junction), an overall connectivity matrix is obtained, whereupon a global [Z] matrix is obtained which characterizes the entire network. Generalized expressions are derived for the evaluation of acoustic performance evaluation parameters such as transmission loss (TL) and insertion loss (IL) for a multiple inlet and multiple outlet (MIMO) system. Some of the characteristic properties of a general multi-port element are also studied in this chapter. The 1-D axial and transverse plane wave analysis is used to characterize axially long and short chambers, respectively, in terms of the [Z] matrix. Different network arrangements of multi-port elements are constructed, wherein the TL performance of such MIMO networks obtained on the basis of either the 1-D axial or 1-D transverse plane wave theory are compared with 3-D FEA carried on a commercial software. The versatility of this algorithm is that it can deal with more than two external or terminal ports, i.e., one can have multiple inlets and outlets in a complicated acoustic network. A generalized approach/algorithm is presented to characterize rigid wall reactive multi-port chamber mufflers of different geometries by means of a 3-D analytical formulation based upon the modal expansion and the uniform piston-driven model. The geometries analyzed here are rectangular plenum chambers, circular cylindrical chamber mufflers with and without a pass tube, elliptical cylindrical chamber mufflers, spherical and hemispherical chambers, conical chamber mufflers with and without a co-axial pass tube and sectoral cylindrical chamber mufflers of circular and elliptical cross-section as well as sectoral conical chamber mufflers. Computer codes or subroutines have been developed wherein by choosing appropriate mode functions in the generalized pressure response function, one can characterize a multi-port chamber muffler of any of the aforementioned separable geometrical shapes in terms of the [Z] matrix, subsequent to which the TL performance of these chambers is evaluated in terms of the scattering matrix [S] parameters by making use of the relations between [Z] and [S] matrices derived earlier. Interestingly, the [Z] matrix approach combined with the uniform piston-driven model is indeed ideally suited for the 3-D analytical formulation inasmuch as regardless of the number of ports, one deals with only one area discontinuity at a time, thereby making the analysis convenient for a multi-port muffler configuration with arbitrary location of ports. The TL characteristics of SISO chambers corresponding to each of the aforementioned geometries (especially the elliptical cylindrical chamber) are analyzed in detail with respect to the effect of chamber dimensions (chamber length and transverse dimensions), and relative angular and axial location of ports. Furthermore, the analysis of SIDO (i.e., single inlet and double outlet) chamber mufflers is given special consideration. In particular, we examine (1) the effect of additional outlet port (second outlet port), (2) variation in the relative angular or axial location of the additional or second outlet port (keeping the location of the inlet port and the outlet ports of the original SISO chamber to be constant) and (3) the effect of interchanging the location of the inlet and outlet ports on the TL performance of these mufflers. Thus, design guidelines are developed for the optimal location of the inlet and outlet ports keeping in mind the broadband attenuation characteristics for a single inlet and multiple outlet (SIMO) system. The non-dimensional limits up to which a flow-reversal elliptical (or circular) cylindrical end chamber having an end-inlet and end-outlet configuration is acoustically short (so that the 1-D transverse plane wave theory is applicable) and the limits beyond which it is acoustically long (so that the 1-D axial plane wave theory is applicable) is determined in terms of the ratio or equivalently, in terms of the ratio. Towards this end, two different configurations of the elliptical cylindrical chamber are considered, namely, (1) End-Offset Inlet (located along the major-axis of the ellipse) and End-Centered Outlet (2) End-Offset Inlet and End-Offset Outlet (both the ports located on the major-axis of the ellipse and at equal offset distance from the center). The former configuration is analyzed using 3-D FEA simulations (on SYSNOISE) while the 3-D analytical uniform piston-driven model is used to analyze the latter configuration. The existence of the higher order evanescent modes in the axially long reversal chamber at low frequency (before the cut-on frequency of the (1, 1) even mode or (1, 0) mode) causes a shift in the resonance peak predicted by the 1-D axial plane wave theory and 3-D analytical approach. Thus, the 1-D axial plane wave analysis is corrected by introducing appropriate end correction due to the modified or effective length of the elliptical cylindrical chamber. An empirical formulae has been developed to obtain the average non-dimensional end correction for the aforementioned configurations as functions of the expansion ratio, (i.e., ), minor-axis to major-axis ratio, (i.e., ) and the center-offset distance ratio, (i.e., ). The intermediate limits between which the chamber is neither short nor long (acoustically) has also been obtained. Furthermore, an ingenious method (Quasi 1-D approach) of combining the 1-D transverse plane wave model with the 1-D axial plane wave model using the [Z] matrix is also proposed for the end-offset inlet and end-centered outlet configuration. A 3-D analytical procedure has also been developed which also enables one to determine the end-correction in axially long 2-port flow-reversal end chamber mufflers for different geometries such as rectangular, circular and elliptical cylindrical as well as conical chambers, a priori to the computation of TL. Using this novel analytical technique, we determine the end correction for arbitrary locations on the two end ports on the end face of an axially long flow-reversal end chamber. The applicability of this method is also demonstrated for determination of the end corrections for the 2-port circular cylindrical chamber configuration without and with a pass tube, elliptical cylindrical chambers as well as rectangular and conical chambers.
2013-08-28T00:00:00Z1H-NMR Study of Proton Glasses - Nuclear Spin Lattice Relaxation in BPxBPI(1-x) and BPxGPI(1-x) - Effects of Disorder on the Proton Group Dynamics
http://etd.iisc.ac.in/handle/2005/4693
1H-NMR Study of Proton Glasses - Nuclear Spin Lattice Relaxation in BPxBPI(1-x) and BPxGPI(1-x) - Effects of Disorder on the Proton Group Dynamics
Ramanuja, M N
Mixed systems made from a combination of ferroelectric (FE) and antiferroelectric (AFE) compounds, exhibit various effects of disorder in different temperature regions. The kind of effects observed, depend on the technique and the window of observation employed. Model systems, like Potassium Ammonium dihydrogen phosphate (KADP), Rubidium Ammonium dihydrogen phosphate (RADP) and BPxBPI(1-x) , with H-bonding networks, have been well studied by dielectric techniques. These investigations have revealed disorder effects like deviations from Curie Weiss law, progressive broadening of dielectric loss curves and dispersion of dielectric constant, at sufficiently low temperatures. NMR studies in such systems are meager and mainly members of the KDP family, like Rubidium ammonium dihydrogen phosphate (RADP) and arsenate (RADA) have been investigated using mainly 2H and 87Rb NMR. On the other hand, proton NMR has been much less used, and our focus is to exploit its power/potential to study 1H group dynamics in the presence (and absence) of disorder in condensed matter systems.
This thesis describes the results of proton NMR investigations in two mixed systems of ferroelectric and antiferroelectric compounds namely, (i) Betaine phosphate (BP, AFE) and Betaine phosphite (BPI, FE) and (ii) Betaine phosphate and Glycine phosphite (GPI, FE). The aim of the study is to obtain information on 1H group dynamics (activation energies and pre-exponential factors) and the effects of micro-spatial disorder. The former system is shown to exhibit orientational glass behavior by extensive dielectric investigations. BP-GPI system is synthesized for the first time and our proton NMR investigation has exhibited interesting effects of disorder like deviation from expected BPP behavior. Further, both systems have exhibited quantum tunneling effects, revealing a gradual transition from classical regime to quantum regime. Biexponential magnetization recovery at low temperatures has also been observed indicating the existence of disorder.
A combination of AFE and FE compounds of this type form a mixed system, over a broad range of compositions, in which the long-range electric order is suppressed owing to frustration effects. Such systems have been treated as dipolar analogues of spin glasses and are known as ‘orientation glasses’ (OG), ‘proton glasses’ (PG) or ‘pseudo-spin glasses. Although the frustrated condensed matter system is crystalline in nature, there is an underlying microstructural randomness due to local fluctuations of the composition which usually results in static lattice strains, which are called random fields. It has been shown that these random fields can also have a pronounced effect on the spin lattice relaxation time as observed in NMR experiments. Depending on the relative concentration and temperature, the mixed system exhibits a range of states (x-T phase diagram) like FE, OG, coexisting OG and AFE, and AFE.
These mixed systems exhibit various kinds of effects of disorder in different temperature regimes which depend upon the technique and window of observation. For e.g., using dielectric spectroscopy we can study the behavior of the electric dipoles during various phases and the effects of frustration seen as dispersion of dielectric constants and broadening of loss curves etc. Through quadrupole perturbed NMR study of systems containing nuclei like 87Rb or 2H, we learn about site-specific inhomogeneities and distribution of EFG in the system. Proton NMR study in the mixed systems, though not much used so far, is a powerful technique to shed light on the dynamics, disorder and Quantum tunneling effects.
Our proton SLR time measurements have been carried out at two Larmor frequencies of 23.3 MHz and 11.4 MHz, in the temperature range of 300 K to 4 K and the results are presented in this thesis, which is divided into four chapters
2D Compressible Viscous Flow Computations Using Acoustic Flux Vector Splitting (AFVS) Scheme
http://etd.iisc.ac.in/handle/2005/277
2D Compressible Viscous Flow Computations Using Acoustic Flux Vector Splitting (AFVS) Scheme
Ravikumar, Devaki
The present work deals with the extension of Acoustic Flux Vector Splitting (AFVS) scheme for the Compressible Viscous flow computations. Accurate viscous flow computations require much finer grids with adequate clustering of grid points in certain regions. Viscous flow computations are performed on unstructured triangulated grids. Solving Navier-Stokes equations involves the inviscid Euler part and the viscous part. The inviscid part of the fluxes are computed using the Acoustic Flux Vector Splitting scheme and the viscous part which is diffusive in nature does not require upwinding and is taken care using a central difference type of scheme. For these computations both the cell centered and the cell vertex finite volume methods are used. Higher order accuracy on unstructured meshes is achieved using the reconstruction procedure. Test cases are chosen in such a way that the performance of the scheme can be evaluated for different range of mach numbers. We demonstrate that higher order AFVS scheme in conjunction with a suitable grid adaptation strategy produce results that compare well with other well known schemes and the experimental data. An assessment of the relative performance of the AFVS scheme with the Roe scheme is also presented.
2007-05-07T09:56:12Z