abstract: Power supply management is important for MEMS (Micro-Electro-Mechanical-Systems) bio-sensing and chemical sensing applications. The dissertation focuses on discussion of accessibility to different power sources and supply tuning in sensing applications. First, the dissertation presents a high efficiency DC-DC converter for a miniaturized Microbial Fuel Cell (MFC). The miniaturized MFC produces up to approximately 10µW with an output voltage of 0.4-0.7V. Such a low voltage, which is also load dependent, prevents the MFC to directly drive low power electronics. A PFM (Pulse Frequency Modulation) type DC-DC converter in DCM (Discontinuous Conduction Mode) is developed to address the challenges and provides a load independent output voltage with high conversion efficiency. The DC-DC converter, implemented in UMC 0.18µm technology, has been thoroughly characterized, coupled with the MFC. At 0.9V output, the converter has a peak efficiency of 85% with 9µW load, highest efficiency over prior publication. Energy could be harvested wirelessly and often has profound impacts on system performance. The dissertation reports a side-by-side comparison of two wireless and passive sensing systems: inductive and electromagnetic (EM) couplings for an application of in-situ and real-time monitoring of wafer cleanliness in semiconductor facilities. The wireless system, containing the MEMS sensor works with battery-free operations. Two wireless systems based on inductive and EM couplings have been implemented. The working distance of the inductive coupling system is limited by signal-to-noise-ratio (SNR) while that of the EM coupling is limited by the coupled power. The implemented on-wafer transponders achieve a working distance of 6 cm and 25 cm with a concentration resolution of less than 2% (4 ppb for a 200 ppb solution) for inductive and EM couplings, respectively. Finally, the supply tuning is presented in bio-sensing application to mitigate temperature sensitivity. The FBAR (film bulk acoustic resonator) based oscillator is an attractive method in label-free sensing application. Molecular interactions on FBAR surface induce mass change, which results in resonant frequency shift of FBAR. While FBAR has a high-Q to be sensitive to the molecular interactions, FBAR has finite temperature sensitivity. A temperature compensation technique is presented that improves the temperature coefficient of a 1.625 GHz FBAR-based oscillator from -118 ppm/K to less than 1 ppm/K by tuning the supply voltage of the oscillator. The tuning technique adds no additional component and has a large frequency tunability of -4305 ppm/V. / Dissertation/Thesis / Ph.D. Electrical Engineering 2012

This thesis reports the development and analysis of high performance CMOS readout electronics for increasing the performance of MEMS gyroscopes developed at Middle East Technical University (METU). These readout electronics are based on unity gain buffers implemented with source followers. High impedance node biasing problem present in capacitive interfaces is solved with the implementation of a transistor operating in the subthreshold region. A generalized fully differential gyroscope model with force feedback electrodes has been developed in order to simulate the capacitive interfaces with the model of the gyroscope. This model is simplified for the single ended gyroscopes fabricated at METU, and simulations of resonance characteristics are done. Three gyroscope interfaces are designed by considering the problems faced in previous interface architectures. The first design is implemented using a single ended source follower biased with a subthreshold transistor. From the simulations, it is observed that biasing impedances up to several gigaohms can be achieved. The second design is the fully differential version of the first design with the addition of a self biasing scheme. In another interface, the second design is modified with an instrumentation amplifier which is used for fully differential to single ended conversion. All of these interfaces are fabricated in a standard 0.6 &micro / m CMOS process. Fabricated interfaces are characterized by measuring their ac responses, noise response and transient characteristics for a sinusoidal input. It is observed that, biasing impedances up to 60 gigaohms can be obtained with subthreshold transistors. Self biasing architecture eliminates the need for biasing the source of the subthreshold transistor to set the output dc point to 0 V. Single ended SOG gyroscopes are characterized with the single ended capacitive interfaces, and a 45 dB gain improvement is observed with the addition of capacitive interface to the drive mode. Minimum resolvable capacitance change and displacement that can be measured are found to be 58.31 zF and 38.87 Fermi, respectively. The scale factor of the gyroscope is found to be 1.97 mV/(&deg / /sec) with a nonlinearity of only 0.001% in &plusmn / 100 &deg / /sec measurement range. The bias instability and angle random walk of the gyroscope are determined using Allan variance method as 2.158 &deg / /&amp / #8730 / hr and 124.7 &deg / /hr, respectively.

This thesis reports the development of high-performance symmetric and decoupled micromachined gyroscopes for tactical-grade inertial measurement applications. The symmetric structure allows easy matching of the resonance frequencies of the drive and sense modes of the gyroscopes for achieving high angular rate sensitivity / while the decoupled drive and sense modes minimizes mechanical cross-coupling for low-noise and stable operation. Three different and new symmetric and decoupled gyroscope structures with unique features are presented. These structures are fabricated in four different micromachining processes: nickel electroforming (NE), dissolved-wafer silicon micromachining (DWSM), silicon-on-insulator (SOI) micromachining, and silicon-on-glass (SOG) micromachining. The fabricated gyroscopes have capacitive gaps from 1.5&micro / m to 5.5&micro / m and structural layer thicknesses from 12&micro / m to 100&micro / m, yielding aspect ratios up to 20 depending on the fabrication process. The size of fabricated gyroscope chips varies from 1x1mm2 up to 4.2x4.6mm2. Fabricated gyroscopes are hybrid-connected to a designed capacitive interface circuit, fabricated in a standard 0.6&micro / m CMOS process. They have resonance frequencies as small as 2kHz and as large as 40kHz / sense-mode resonance frequencies can be electrostatically tuned to the drive-mode frequency by DC voltages less than 16V. The quality factors reach to 500 at atmospheric pressure and exceed 10,000 for the silicon gyroscopes at vacuum. The parasitic capacitance of the gyroscopes on glass substrates is measured to be as small as 120fF. The gyroscope and interface assemblies are then combined with electronic control and feedback circuits constructed with off-the-shelf IC components to perform angular rate measurements. Measured angular rate sensitivities are in the range from 12&micro / V/(deg/sec) to 180&micro / V/(deg/sec), at atmospheric pressure. The SOI gyroscope demonstrates the best performance at atmospheric pressure, with noise equivalent rate (NER) of 0.025(deg/sec)/Hz1/2, whereas the remaining gyroscopes has an NER better than 0.1(deg/sec)/Hz1/2, limited by either the small sensor size or by small quality factors. Gyroscopes have scale-factor nonlinearities better than 1.1% with the best value of 0.06%, and their bias drifts are dominated by the phase errors in the demodulation electronics and are over 1deg/sec. The characterization of the SOI and SOG gyroscopes at below 50mTorr vacuum ambient yield angular rate sensitivities as high as 1.6mV/(deg/sec) and 0.9mV/(deg/sec), respectively. The NER values of these gyroscopes at vacuum are smaller than 50(deg/hr)/Hz1/2 and 36(deg/hr)/Hz1/2, respectively, being close to the tactical-grade application limits. Gyroscope structures are expected to provide a performance better than 10 deg/hr in a practical measurement bandwidth such as 50Hz, provided that capacitive gaps are minimized while preserving the aspect ratio, and the demodulation electronics are improved.

A piezoelectric energy harvesting (PEH) system can harvest electrical energy from ambient vibration energy. In a PEH system, the interface rectifier circuit is critical because it converts AC from the output of piezoelectric harvester to DC that can power the load. Hence, improving the efficiency of the interface circuit can directly increase the efficiency of the entire PEH system; consequently, more power can be harvested. Commonly used interface circuits in PEH systems, such as full-bridge and voltage- doubler rectifiers,lead to relatively simple circuit implementations but they show serious limitations in energy-harvesting efficiency. Several innovative solutions have been reported to improve the efficiency of the interface rectifiers, such as ‘switch-only’ and ‘bias-flip’ techniques [7]. Such solutions utilize additional switches or switched inductors to speed up and even quickly reverse (flip) the voltage on the rectifier input to the desired voltage-level and condition for energy transfer, ultimately improving the overall efficiency of the energy harvesting. However, such techniques rely on accurate timing and synchronization of the pulsed switches every time the current produced by the piezoelectric harvester changes polarity. This thesis studies and investigates the impact of the non-ideal switching effects on the energy efficiency of the switch-only and bias-flip interface rectifiers in a PEH system, by theoretical derivation and experimental simulation.

La récupération d'énergie mécanique de vibration à l’aide de transducteurs à capacité variable mène à l’étude de systèmes non linéaires complexes, mais présente des perspectives applicatives très prometteuses. Notre travail a porté sur l’étude d’une nouvelle famille de circuits d'interface pour transducteurs capacitifs. Entre autres avantages, ces circuits sont réalisables avec des rendements élevés à très basse puissance, typiquement dès quelques dizaines de nano-watts de puissance moyenne, ce qui les distingue des solutions présentées dans de l’état de l’art. De plus, Les circuits étudiés dans cette thèse ne contiennent aucun composant magnétique, ce qui constitue un atout considérable en termes de miniaturisation et d’intégration et permet eu outre la compatibilité avec l’imagerie par résonance magnétique. Les différentes structures qui constituent la famille de circuits proposés permettent de répondre à différentes contraintes imposées par le transducteur capacitif, en particulier le rapport des capacités maximale et minimale Cmax/Cmin. A partir d’une tension de sortie donnée, la tension appliquée sur le transducteur capacitif peut être modifiée en utilisant différents circuits ou en utilisant un circuit unique dont la topologie est modifiée à l’aide d’un interrupteur électronique. Les modèles théoriques développés prennent en compte le couplage électromécanique du transducteur de manière à décrire le comportement global des systèmes étudiés. Les circuits étudiés ont été validés expérimentalement avec deux transducteurs capacitifs de structure différente. En pratique, le rendement de ces circuits est proche de 80% pour des puissances converties aussi basses que la centaine de nano watts. / The mechanic vibration energy harvesting using variable capacitance transducers leads to the study of complex nonlinear systems but has very promising application perspectives. Our work focused on the study of a new family of interface circuits for capacitive transducers. Among all the advantages, these circuits are achievable with high efficiencies at very low power, typically a few tens of nanowatts average power, which distinguishes them from the solutions presented in the state of the art. Moreover, the circuits studied in this thesis do not contain any magnetic components, which is a considerable asset in terms of miniaturization and integration and also allows compatibility with magnetic resonance imaging. The various structures which constitute the family of circuits proposed make it possible to satisfy various constraints imposed by the capacitive transducer, in particular, the ratio of the maximum and minimum capacities Cmax / Cmin. For a given output voltage, the voltage applied to the capacitive transducer can be varied by using different circuits or by using a single circuit whose topology is modified by the operation of an electronic switch. In order to describe the overall behavior of the studied systems, the electromechanical coupling of the transducer is taken into account in the developed theoretical models. The studied circuits have been validated experimentally with two capacitive transducers of different structure. In practice, the output of these circuits is close to 80% for converted powers as low as the hundred nanowatts.

This thesis reports the development of advanced readout and control electronics for MEMS gyroscopes developed at METU. These gyroscope electronics are separated into three main groups: high sensitive interface circuits, drive mode amplitude controlled self oscillation loops, and sense mode phase sensitive amplitude demodulators. The proposed circuits are first implemented with discrete components, and then integrated on CMOS chips. A self oscillation loop enabling constant amplitude drive mode vibrations independent of sensor parameters and ambient conditions is developed. A fully functional angular rate system, which is constructed by employing this advanced control electronics together with the transresistance amplifier type interfaces and sense mode electronics, is implemented on a dedicated PCB having 5.4x2.4 cm2 area. This system demonstrates an impressive performance far better than the best performance achieved by any angular rate system developed at METU. Bias instability and angle random walk values are measured as 14.3 &ordm / /hr and 0.126 &ordm / /&amp / #8730 / hr, respectively. The scale factor of the system is found as 22.2 mV/(&ordm / /sec) with a nonlinearity of 0.01%, and a zero rate output of 0.1 &ordm / /sec, in &plusmn / 50 &ordm / /sec measurement range. CMOS unity gain buffer (UGB) and transimpedance amplifier (TIA) type resistive and capacitive interfaces are characterized through AC, transient, and noise tests. It is observed that on chip biasing mechanisms properly DC-bias the high impedance nodes to 0 V potential. UGB type capacitive interfaces demonstrate superior performance than TIA counterparts due to stability problems associated with TIA interfaces. CMOS differential drive mode control and sense mode demodulation electronics give promising results for the future performance tests.

This thesis reports the development of various high performance readout and control electronics for implementing angular rate sensing systems using MEMS gyroscopes developed at METU. First, three systems with open loop sensing mechanisms are implemented, where each system has a different drive-mode automatic gain controlled (AGC) self-oscillation loop approach, including (i) square wave driving signal with DC off-set named as OLS_SquD, (ii) sinusoidal driving signal with DC off-set named as OLS_SineD, and iii) off-resonance driving signal named as OLS_OffD. A forth system is also constructed with a closed loop sensing mechanism where the drive mode automatic gain controlled (AGC) self-oscillation loop approach with square wave driving signal with DC off-set named as CLS_SquD. Sense and drive mode electronics employ transimpedance and transresistance amplifiers as readout electronics, respectively. Each of the systems is implemented with commercial discrete components on a dedicated PCB. Then, the angular rate sensing systems are tested with SOG (Silicon-on-Glass) gyroscopes that are adjusted to have two different mechanical bandwidths, more specially 100 Hz and 30 Hz. Test results of all of these cases verify the high performance of the systems. For the 100 Hz bandwidth, the OLS_SquD system shows a bias instability of 4.67 &amp / #730 / /hr, an angle random walk (ARW) 0.080 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 22.6 mV/(&amp / #730 / /sec). For the 30 Hz bandwidth, the OLS_SquD system shows a bias instability of 5.12 &amp / #730 / /hr, an ARW better than 0.017 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 49.8 mV/(&amp / #730 / /sec). For the 100 Hz bandwidth, the OLS_SineD system shows a bias instability of 6.92 &amp / #730 / /hr, an ARW of 0.049 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 17.97 mV/(&amp / #730 / /sec). For the 30 Hz bandwidth, the OLS_SineD system shows a bias instability of 4.51 &amp / #730 / /hr, an ARW of 0.030 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 43.24 mV/(&amp / #730 / /sec). For the 100 Hz bandwidth, the OLS_OffD system shows a bias instability of 8.43 &amp / #730 / /hr, an ARW of 0.086 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 20.97 mV/(&amp / #730 / /sec). For the 30 Hz bandwidth, the OLS_OffD system shows a bias instability of 5.72 &amp / #730 / /hr, an ARW of 0.046 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 47.26 mV/(&amp / #730 / /sec). For the 100 Hz bandwidth, the CLS_SquD system shows a bias instability of 6.32 &amp / #730 / /hr, an ARW of 0.055 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 1.79 mV/(&amp / #730 / /sec). For the 30 Hz bandwidth, the CLS_SquD system shows a bias instability of 5.42 &amp / #730 / /hr, an ARW of 0.057 &amp / #730 / /&amp / #8730 / hr, and a scale factor of 1.98 mV/(&amp / #730 / /sec). For the 100 Hz bandwidth, the R2 nonlinearities of the measured scale factors of all systems are between 0.0001% and 0.0003% in the &plusmn / 100 &amp / #730 / /sec measurement range, while for the 30 Hz bandwidth the R2 nonlinearities are between 0.0002% and 0.0062% in the &plusmn / 80&amp / #730 / /sec measurement range. These performance results are the best results obtained at METU, satisfying the tactical-grade performances, and the measured bias instabilities and ARWs are comparable to the best results in the literature for a silicon micromachined vibratory gyroscope.

This research explores a novel in-situ technique for monitoring film thickness in the reactive etching process that incorporates a micromachined sensor. The sensor correlates film thickness with changes in resonant frequency that occurs in the micromachined platform during etching. The sensor consists of a platform that is suspended over drive and sense electrodes on the surface of the substrate. As material is etched from the platform, its resonant vibrational frequency shifts by an amount that is proportional to the amount of material etched, allowing etch rate to be inferred. This RIE monitoring methodology exploits the accuracy of resonant micromechanical structures, whereby shifts in the fundamental resonant frequency measure a physical parameter. A majority of these systems require free-standing mechanical movement and utilize a sacrificial layer process as the key technique to develop and release the structure on a substrate. A sacrificial layer technique that incorporates a low temperature sacrificial polymer was utilized to develop and release the suspended RIE sensor with excellent performance and is capable of fabricating other low cost, high performance and reliable suspended MEMS devices. The integration of sensors and electronic circuitry is a dominant trend in the semiconductor industry, and much work and research has been devoted to this effort. The RIE sensor relies on capacitive transduction to detect small capacitance changes and the resulting change in resonant frequency during the RIE process. The RIE sensor's overall performance is limited by the interface circuit, and integration with the proper circuit allows the RIE sensor to function as a highly sensitive measure of etch rate during the RIE process. A capacitive feedback charge amplifier interface circuit, when configured with the RIE senor at the input achieves very low noise sensing of capacitance changes and offers the potential for wide dynamic range and high sensitivity. As an application vehicle, process control was demonstrated in the PlasmaTherm SLR series RIE system located in the Georgia Tech Microelectronics Research Center.

With the proliferation of industries world-wide, there is a growing need and interest in sensing and monitoring environmental pollutants and monitoring the concentration of chemicals/gases in industrial process control. There is also an increasing demand for chemical sensors in other applications such as home security, breath analysis and food processing. Design and development of metal-oxide based gas sensor system is reported in this thesis. The system consists of three components viz. micro heater(which aids inheating the sensor film to required temperatures), CMOS ASIC (the sensor interface circuit) and the thin film transducer(a semiconducting metal oxide thin film whose resistance changes with the concentration of the target gas). Microheaters were realized through PolyMUMPs process. Thermal characterization of surface-micromachined microheaters is carried out from their dynamic response to electrothermal excitations. An electrical equivalent circuit model is developed for the thermo-mechanical system. The mechanical parameters are extracted from the frequency response obtained using a Laser Doppler Vibrometer. The resonant frequencies of the microheaters are measured and compared with FEM simulations. The thermal time constants are obtained from the electrical equivalent model by fitting the model response to the measured frequency response. Microheaters with an active area of140m × 140m have been realized on two different layers(poly-1 andpoly-2) with two different air-gaps (2m and 2.75m). The effective time constants, combining thermal and mechanical responses, are intherangeof0.13msto0.22msforheatersonpoly-1,and1.9s to0.15ms for microheaters on poly-2 layer. The thermal time constants of the best microheaters are in the range of a few s, thus making them suitable for sensor applications that need faster thermal response. The mechanical deformation of the microheaters subjected to an electrothermal excitation, due to thermal stress, is also analyzed using lensless in-line digital holographic microscopy (DHM). The numerically reconstructed holographic images of the micro-heaters clearly indicate the regions under high stress. Double exposure method has been used to obtain the quantitative measurements of the deformations, from the phase analysis of the hologram fringes. The measured deformations correlate well with the theoretical values predicted by a thermo-mechanical analytical model. The results show that lensless in-line DHM with Fourier analysis is an effective method for evaluating the thermo-mechanical characteristics of MEMS components. A sensor interface circuit comprising a resistance-to-time period converter as the front-end circuit and a proportional temperature controller to control the microheater temperature is designed and realized in 130nm UMC CMOS technology. The impact of biasing the transistors in subthreshold versus saturation conditions on analog circuit performance is systematically analyzed. A cascode current mirror, designed in 130nm CMOS technology, is biased in subthreshold and saturation regions and its performance has been analyzed through rigorous analytical modeling. The analytical results have been validated with SPICE simulations. It is demonstrated that the subthreshold operation provides better performance in terms of linearity, power, area, output impedance and tolerance to temperature variation, making it a preferable option for applications such as signal conditioning circuitry for environmental sensors. On the other hand, biasing the circuit in saturation is preferable for applications like transceivers and data converters where high bandwidth, SNR and low sensitivity to process variations are the key requirements. Based on this analysis, a sensor interface circuit has been prototyped for resistance measurement on 130nm CMOS technology, using subthreshold cascode current mirrors as the key building blocks. This current mirror results in 14X lower power compared to above-threshold operation. The interface circuit spans 5 orders of magnitude of resistance, and consumes an ultra low power of 326W. A proportional temperature controller with an integrated on-chip power MOSFET is also realized on the same chip for heating and temperature control of microheaters. The microheater is reused as temperature sensor. The entire circuit works with 1.2V supply, except the power MOSFET and the heater driver circuit, which operate with 3.3V supply. ZnO, a semiconducting metal-oxide, is used as the sensing material. Thin films of ZnO are spin-coated over insulating substrates using sol-gel processing technique. Gold pads deposited over the sensing film act as electrodes. The sensor film is characterized at different temperatures for its sensitivity to ethanol. A peak response of 14% change in resistance is observed for 5ppm ethanol, at a working temperature of 275◦C.

Indiana University-Purdue University Indianapolis (IUPUI) / The link between volatile organic compounds (VOCs) from breath and various diseases and specific conditions has been identified since long by the researchers. Canine studies and breath sample analysis on Gas chromatography/ Mass Spectroscopy has proven that there are VOCs in the breath that can detect and potentially predict hypoglycemia. This project aims at developing a smart nanosensor system to detect hypoglycemia from human breath. The sensor system comprises of 1-Mercapto-(triethylene glycol) methyl ether functionalized goldnanoparticle (EGNPs) sensors coated with polyetherimide (PEI) and poly(vinylidene fluoride -hexafluoropropylene) (PVDF-HFP) and polymer composite sensor made from PVDF-HFP-Carbon Black (PVDF-HFP/CB), an interface circuit that performs signal conditioning and amplification, and a microcontroller with Bluetooth Low Energy (BLE) to control the interface circuit and communicate with an external personal digital assistant. The sensors were fabricated and tested with 5 VOCs in dry air and simulated breath (a mixture of air, small portion of acetone, ethanol at high humidity) to investigate sensitivity and selectivity. The name of the VOCs is not disclosed herein but these VOCs have been identified in-breath and are identified as potential biomarkers for other diseases as well. The sensor hydrophobicity has been studied using contact angle measurement. The GNPs size was verified using Ultra-Violent-Visible (UV-VIS) Spectroscopy. Field Emission Scanning Electron Microscope (FESEM) image is used to show GNPs embedded in the polymer film. The sensors sensitivity increases by more than 400\% in an environment with relative humidity (RH) of 93\% and the sensors show selectivity towards VOCs of interest. The interface circuit was designed on Eagle PCB and was fabricated using a two-layer PCB. The fabricated interface circuit was simulated with variable resistance and was verified with experiments. The system is also tested at different power source voltages and it was found that the system performance is optimum at more than 5 volts. The sensor fabrication, testing methods, and results are presented and discussed along with interface circuit design, fabrication, and characterization. / 2022-05-8