The main topic of this thesis is the development of a chemical microsystem based on integration of a silicon-based resonant microsensor and a CMOS ASIC for portable sensing applications. Cantilever and disk-shape microresonators have been used as mass-sensitive sensors. Based on the characteristics of the microresonators, CMOS integrated interface and control electronics have been implemented. The CMOS ASIC utilizes the self-oscillation method, which incorporates the microresonator in an amplifying feedback loop as the frequency determining element. In this manner, the ASIC includes a main feedback loop to sustain oscillation at or close to the fundamental resonance frequency of the microresonator. For stable oscillation, an automatic gain control loop regulates the oscillation amplitude by controlling the gain of the main feedback loop. In addition, an automatic phase control loop has been included to adjust the phase of the main feedback loop to ensure an operating point as close as possible to the resonance frequency, resulting in improved frequency stability. The CMOS chip has been interfaced to cantilever and disk-shape microresonators and short-term frequency stabilities as low as 3.4×10-8 in air have been obtained with a 1 sec gate time. The performance of the implemented microsystem as a chemical sensor has been evaluated experimentally with microresonators coated with chemically sensitive polymer films. With a gas-phase chemical measurement setup constructed in this work, chemical measurements have been performed and different concentrations of VOCs, such as benzene, toluene and m-xylene have been detected with limits of detection of 5.3 ppm, 1.2 ppm and 0.35 ppm, respectively. To improve the long-term stability in monitoring applications with slowly changing analyte signatures, a method to compensate for frequency drift caused by environmental disturbances has been implemented on the CMOS chip. This method uses a controlled stiffness modulation generated by a frequency drift compensation circuit to track the changes in the resonator's Q-factor in response to variations in the environmental conditions. The measured Q-factor is then used to compensate for the frequency drift using an initial calibration step. The feasibility of the proposed method has been verified experimentally by compensating for temperature-induced frequency drift during gas-phase chemical measurements.

L'oscillateur optoélectronique (OEO) a été introduit pour la première fois en 1996 par S. Yao et L. Maleki, en tant qu'oscillateur microondes à très faible bruit de phase et obtenu par synthèse directe. Les développements de l'OEO concernent les applications en photonique microondes, télécommunications optiques, radar et traitement du signal. Mais l'OEO devrait aussi pouvoir être utilisé dans le domaine des capteurs. Dans cette thèse nous étudiants plusieurs aspects de l'OEO pour son application à la mesure d'indice de réfraction d'un liquide. Compte tenu de sa structure l'OEO dépend fortement des conditions ambiantes d'utilisation. S'il n'est pas bien optimisé ni contrôlé, il ne peut pas fonctionner correctement sur une longue durée. Nous avons étudié les influences de la température sur le modulateur électrooptique (EOM) et sur le comportement global de l'OEO. Un contrôle de température réduit de façon significative le phénomène de dérive de l'EOM. Afin de la supprimer complètement, nous avons mis au point une instrumentation construite autour d'une carte DSP, permettant de détecter et compenser la dérive du point de fonctionnement optique de l'EOM tout en contrôlant simultanément sa température. Une première technique est basée sur un signal de test, basse fréquence, appliqué à l'électrode DC du modulateur. Une deuxième solution consiste à travailler sur la puissance optique en sortie du modulateur. En combinant les deux on peut profiter des avantages de ces deux méthodes. Utilisant ainsi l'OEO nous avons testé plusieurs configurations pour mesurer l'indice de réfraction de quatre solutions chimiques bien connues, nous avons obtenu une variance de 3 pour mille. Les résultats sont en assez bon accord avec les publications correspondantes. Enfin nous avons aussi introduit une nouvelle méthode pour améliorer les mesures d'indice de réfraction faites à long terme en suivant, grâce à un analyseur vectoriel de réseau, les évolutions au cours du temps du temps de propagation dans la fibre optique. En introduisant à partir de cette mesure une correction aux mesures de la fréquence d'oscillation il est possible de réduire les fluctuations de cette fréquence à seulement 606 Hz, sur une durée de 62 h, ce que l'on peut comparer aux 8 GHz de l'oscillateur. Ainsi le rapport signal à bruit, peut être grandement amélioré lors de la mesure d'indice de réfraction et il doit être possible de diminuer la limite de détection des variations de l'indice de réfraction au cours du temps. / The optoelectronic oscillator (OEO) was first introduced in 1996 by S. Yao and L. Maleki as a very low phase noise microwave oscillator working in direct synthesis. The OEO developments concern applications in microwave photonics, optical telecommunication, radar and high speed signal processing systems but it should also be used in the sensing domain. In this thesis, we study several aspects to apply the OEO to liquid refractive index measurement. Because of its structure the OEO is very dependent on the ambient conditions. If the OEO is not optimized and controlled, it cannot operate well for long duration. We have analyzed the influences of temperature on the electrooptic modulator (EOM) and the global OEO behavior. Temperature control can significantly reduce the drift phenomena of the EOM. In order to totally remove this drift, we have developed a complete digital system, based on a DSP kit, to detect and compensate automatically the EOM optical bias point drift and to control simultaneously its temperature. The first technique is based on a dither signal at low frequency, injected to DC electrode of the EOM. The second one is based on the average optical output power of the EOM. A combination of these two techniques can take advantages from both of them. Using like that the OEO, we have tested several configurations to measure the refractive index of four classical chemical solutions leading to a standard deviation of 3 per thousand. The results are in rather good agreement with previous publications. Finally, we have introduced a new method to improve the long-term refractive index measurement by monitoring, with a vector network analyzer, the variations of the optical delay in the fiber loop of the OEO. Introducing by this way a correction to the long-term frequency measurement it is possible to reduce the oscillation frequency fluctuations to only 606 Hz, compared to the 8 GHz of the oscillator, for a duration of 62 hours. Therefore the signal-to-noise ratio in the refractive index measurement can be enhanced and so the detection resolution of the refractive index variations during time.

Speed, power and reliability of analogue integrated circuits (IC) exhibit temperature dependency through the modulation of one or several of the following variables: band gap energy of the semiconductor, mobility, carrier diffusion, current density, threshold voltage, interconnect resistance, and variability in passive components. Some of the adverse effects of temperature variations are observed in current and voltage reference circuits, and frequency drift in oscillators. Thermal instability of a voltage-controlled oscillator (VCO) is a critical design factor for radio frequency ICs, such as transceiver circuits in communication networks, data link protocols, medical wireless sensor networks and microelectromechanical resonators. For example, frequency drift in a transceiver system results in severe inter-symbol interference in a digital communications system. Minimum transconductance required to sustain oscillation is specified by Barkhausen’s stability criterion. However it is common practice to design oscillators with much more transconductance enabling self-startup. As temperature is increased, several of the variables mentioned induce additional transconductance to the oscillator. This in turn translates to a negative frequency drift. Conventional approaches in temperature compensation involve temperature-insensitive biasing proportional-to-absolute temperature, modifying the control voltage terminal of the VCO using an appropriately generated voltage. Improved frequency stability is reported when compensation voltage closely follows the frequency drift profile of the VCO. However, several published articles link the close association between oscillation amplitude and oscillation frequency. To the knowledge of this author, few published journal articles have focused on amplitude control techniques to reduce frequency drift. This dissertation focuses on reducing the frequency drift resulting from temperature variations based on amplitude control. A corresponding hypothesis is formulated, where the research outcome proposes improved frequency stability in response to temperature variations. In order to validate this principle, a temperature compensated VCO is designed in schematic and in layout, verified using a simulation program with integrated circuit emphasis tool using the corresponding process design kit provided by the foundry, and prototyped using standard complementary metal oxide semiconductor technology. Periodic steady state (PSS) analysis is performed using the open loop VCO with temperature as the parametric variable in five equal intervals from 0 – 125 °C. A consistent negative frequency shift is observed in every temperature interval (≈ 11 MHz), with an overall frequency drift of 57 MHz. However similar PSS analysis performed using a VCO in the temperature stabilised loop demonstrates a reduced negative frequency drift of 3.8 MHz in the first temperature interval. During the remaining temperature intervals the closed loop action of the amplitude control loop overcompensates for the negative frequency drift, resulting in an overall frequency spread of 4.8 MHz. The negative frequency drift in the first temperature interval of 0 to 25 °C is due to the fact that amplitude control is not fully effective, as the oscillation amplitude is still building up. Using the temperature stabilised loop, the overall frequency stability has improved to 16 parts per million (ppm)/°C from an uncompensated value of 189 ppm/°C. The results obtained are critically evaluated and conclusions are drawn. Temperature stabilised VCOs are applicable in applications or technologies such as high speed-universal serial bus, serial advanced technology attachment where frequency stability requirements are less stringent. The implications of this study for the existing body of knowledge are that better temperature compensation can be obtained if any of the conventional compensation schemes is preceded by amplitude control. / Dissertation (MEng)--University of Pretoria, 2013. / Electrical, Electronic and Computer Engineering / unrestricted