Voltage-controlled oscillators are widely used circle blocks, particularly in phase- locked loops. As CMOS is the technology of choice for many applications, CMOS oscillators with low timing jitter are highly desired. In this thesis, two types of VCO based on differential ring oscillator and relaxation oscillator are proposed. We describes the effect of supply noise on the performance of differential ring and relaxation oscillators. Compared to the conventional VCO, the proposed VCOs have lower sensitivity to noise on the power supply and also provided a high linearity gain (Kvco) which decreases the VCO jitter in the PLL circuit and improve the system stability. Both of VCOs are designed in TSMC 0.35£gm 2P4M Mixed-Signal process technology.

La tendance dans la conception de cartes électroniques imprimées est de remplacer les traditionnels bus parallèles par des liens série rapides dont le débit peut atteindre plusieurs dizaines de Gigabit par seconde (Gbps). Cette thèse proposée par THALES Communications & Security en collaboration avec le laboratoire SATIE de l'ENS de Cachan a pour objectif de définir une approche adaptée au traitement des problèmes de liens multi-gigahertz, de manière à garantir le fonctionnement d’une carte numérique complexe (multicouches, haute densité d'intégration, ...) sans qu’une phase de prototypage ne soit nécessaire. Après un état de l’art, ce travail s’est organisé en trois parties : La première partie porte sur l'étude du canal de propagation. La décomposition spectrale des liens multi-gigabits couvrant plusieurs gigahertz voir plusieurs dizaines de gigahertz montre la nécessité d'employer des logiciels de simulations spécifiques au domaine des hyperfréquences. Une évaluation de certains solveurs électromagnétiques 3D parmi les plus récents a été réalisée afin d'extraire les paramètres S du canal de propagation de façon précise et rapide a partir des informations issues des logiciels de CAO utilisés à THALES. La seconde partie traite de la prise en compte des émetteurs, des récepteurs et des traitements numériques associés dans la simulation afin de réaliser des calculs de diagrammes de l'œil, de taux d’erreurs binaires (BER) et de jitter. L’utilisation de la norme IBIS-AMI, très récente, et la comparaison des performances aves d’autres outils, tel que HSPICE, a demandé l'évaluation de simulateurs circuit de dernière génération. Cette étape a été réalisée en étroite collaboration avec les éditeurs des logiciels car certains outils ne sont pas suffisamment matures pour s'inscrire dans un flot global de conception. Enfin, la chaîne de simulation complète ayant été validée par la mesure, nous avons effectué une analyse approfondie des différentes composantes du jitter en fonction des phénomènes physiques plus ou moins destructeurs pour la qualité du signal. Cela nous a ensuite permis d’établir les règles et la méthodologie de conception, en tenant compte des marges allouées à partir des résultats de l’analyse du jitter. / The designers of Printed Circuit Board (named “board” below) tend to use more and more multi-gigabit serial links rather than traditional parallel buses. It enables to push back the density limitations and to increase embedded functionalities of the board. This thesis is the result of collaboration with THALES Communications & Security and the SATIE laboratory of ENS Cachan. The goal of the thesis was to define an approach dedicated to the study of Multi-GigaHertz (MGH) signals in order to assure that digital complex boards work without costly multiple prototype designs. After an inventory of the state of the arts, this work was conducted in three parts: The firt part relates to the study of the propagation channel. The spectral power distribution of the multi-gigabit links ranges from DC to several dozens of gigahertz, it is the reason why specific simulation softwares usually used in the hyper-frequency field have to be used A benchmark of several most recent 3D ElectroMagnetic (EM) solvers has been achieved in order to quickly and accurately extract the S Parameter matrix of the propagation channel thanks to information from CAO softwares used in THALES The second part consisted to take into account the transmitters, the receivers and the digital treatments associated in the circuit simulation in order to calculate eye diagrams, Bit Error Rate (BER) and Jitter separation. The benchmark of the latest generation of channel simulators was needed for the use of the recent norm IBIS-AMI and the comparison of the performances with other tools, such as HSPICE. This step has been led in close collaboration with the simulation software suppliers because some tools are not mature enough to fit into a global design flow. Finally, thanks to the validation of the simulation flow with measurements, a deep sudy of the different components of the jitter has been conducted depending on the physical phenomenon being more or less destructive for the quality of the transmission. This study enabled to define design rules and design methodology taking into account the margins allocated from the results of the jitter analysis.

Synchronization means the aligning of the significant instants of one signal to the significant instants of another. In digital systems, where timing transfer between systems is required, synchronization is an important function. In this thesis new results on the performance and design of synchronization processes are presented.An inescapable consequence of the synchronization of external autonomous inputs in digital systems is the possibility of failure of digital devices used to capture the external signal. The anomalous behaviour of these devices is referred to as metastability. The most commonly used approach to controlling the problem of metastability is the use of synchronizers. A synchronizer can be designed to reduce the probability of metastable: failure but cannot eliminate it altogether. New high performance synchronizer designs are presented and analysed in this thesis.Another consequence of synchronization is the resulting disturbance of the significant epochs of timing signals. This disturbance is referred to as jitter. The characterization of jitter produced in synchronization processes is important in the design of digital systems. In this thesis, jitter characteristics are derived for two important applications that arise in digital communications systems. The characterization provides new insight into the dependence of the jitter on system parameters.

Jitter is a dominant factor contributing to a high bit error rate (BER) in high speed I/O circuitry, and it aggravates the quality of a clock signal from a phase-locked loop (PLL), subsequently impacting a given timing budget. The recent proliferation of systems-on-a-chip (SoCs) with help of technology scaling makes jitter measurement more challenging as the SoCs integrate more I/O circuitry and PLLs within a chip. Jitter has been, however, one of the most difficult parameters to measure accurately when validating the high speed serial I/O circuitry or PLLs, mostly due to its small value. External instruments with full-fledged high precision measurement hardware, along with comprehensive analysis tools, have been used for jitter measurement, but increased test cost from long test time, signal integrity, and human intervention prevent this approach from being used for high volume manufacturing testing. Built-in self-test (BIST) solutions have recently become attractive to overcome these drawbacks, but complicated analog circuit designs that are sensitive to ever increasing process variations, and associated complex analysis methods impede their adoption in the SoCs. This dissertation studies practical random jitter measurement methods that achieve measurement accuracy by exploiting a differential approach and make the proposed methods tester-friendly solutions for an automatic test equipment (ATE). We first propose a method of measuring the average value of the random jitter, rather than measuring the jitter at every clock cycle, that can be converted to the root-mean-square (RMS) value of the random jitter, which is the key indicator of the quantity of the random jitter. Then, we propose a simple but accurate delay measurement method which uses the proposed jitter measurement method for random jitter measurement when a reference signal, such as a golden PLL output in high speed I/O validation, is not available. The validity of the proposed random jitter measurement method is supported by measurement results from a test chip. The impact of substrate noise on the signal of interest is also shown with measurements using a test chip. To address the random jitter of a clock signal when the clock is operating in its functional mode, we demonstrate a novel method for random jitter measurement that explores the shmoo capability of a low-cost production tester without relying on any BIST circuitry. / text

The analysis of electromyographic (EMG) signals detected during muscle contraction provides important information to aid in the diagnosis and characterization of neuromuscular disorders. One important analysis measures neuromuscular jitter, which is the variability of the time intervals between two muscle fibre potentials (MFPs) belonging to the same motor unit over a set of discharges. Conventionally, neuromuscular jitter is measured using single fibre (SF) EMG techniques, which can identify individual MFPs by using a SF needle electrode. However, SF electrodes are expensive, very sensitive to needle movement and not easy to operate in practise. <br /><br /> A method is studied in this thesis for automatically measuring neuromuscular jitter in motor unit potentials (MUP), it measures jitter using routine EMG techniques, which detect MUPs using a concentric needle (CN) electrode. The method is based on the detection of near MFP contributions, which correspond to individual muscle fibre contributions to MUPs, and the identification of individual MFP pairs. The method was evaluated using simulated EMG data. After an EMG signal is decomposed into MUP trains, a second-order differentiator, McGill filter, is applied to detect near MFP contributions to MUPs. Then, using nearest neighbour clustering and minimum spanning tree algorithms, the sets of available filtered MUPs can be selected and individual MFPs can be identified according to the features of their shapes. Finally, individual MFP pairs are selected and neuromuscular jitter is measured. <br /><br /> Using the McGill filter, near MFP contributions to detected CN MUPs can be consistently detected across an ensemble of successive firings of a motor unit. The method is an extension of the work Sheng Ma, compared to previous works, more efficient algorithms are used which have demonstrated acceptable performance, and which can consistently measure neuromuscular jitter in a variety of EMG signals.

Die vorliegende Arbeit analysiert vorsprachliche Lautäußerungen des zweiten und dritten Lebensmonates von 19 Säuglingen mit oro-fazialen Spaltbildungen und 24 gesunden Säuglingen als Referenzgruppe erstmalig unter dem Aspekt des Einflusses einer erfolgten kieferorthopädischen Frühbehandlung. Die Untersuchungen konzentrierten sich dabei auf die Grundfrequenz und die Länge der Einzelvokalisationen sowie auf verschiedene Stimmstabilitätsparameter (Jitter, Shimmer, PPQ). Für die Grundfrequenz der Lautäußerungen konnten weder deutliche Unterschiede zwischen Säuglingen mit oro-fazialen Spaltbildungen und den gesunden Gleichaltrigen festgestellt werden, noch konnte ein maßgeblicher Einfluss der Oberkieferplatte auf die mittlere Grundfrequenz vorsprachlicher Vokalisationen gezeigt werden. Die Tatsache, dass bei nahezu allen Analysen der Stimmstabilität ein positiver Effekt der Oberkieferplatte zu verzeichnen war, spricht stark dafür, dass sich diese kieferorthopädische Maßnahme positiv auf die Stabilität der Phonation auswirkt. Die hier durchgeführten Analysen zeigen vor allem signifikante Unterschiede zwischen der Kontrollgruppe und der Gruppe mit rechtsseitiger LKGS-Spalte. Inwieweit dieser Befund tatsächlich Bestand hat, müssen weiterführende Studien zeigen, in denen insbesondere ein stärkeres Augenmerk auf adäquate begleitende pädaudiologische Untersuchungen gelegt werden sollte. / The present paper analyses spontaneous cries during the second and third month of life of 19 infants with cleft lip and palate compared to 24 non-cleft infants as a reference group. Research has been performed in the light of the influence of pre-orthodontic treatment for the first time. Tests have been focused on fundamental frequency (F0), length of the infant cries and on different parameters of the F0 stability (Jitter, Shimmer, PPQ). No significant varieties have been found between infants with cleft lip and palate and the non-cleft contemporaries. A decisive impact of the intra-oral plate on fundamental frequency of pre-linguistic vocalization has not been found either. Almost all tests have shown a positive effect of the pre-orthodontic treatment on voice stability. This strongly suggests that the orthodontic measure has a favourable impact on the stability of Phonation. The current tests show significant differences in particular between the non-cleft group and the group with unilateral cleft lip and palate. Whether these findings will be confirmed or not has to be shown by future research that is preferably focused on appropriate paediatric audiology investigations / examinations.

Effects of external vibrations (called g-jitter) on Marangoni convection in a liquid bridge were investigated on the International Space Station (ISS) and in ground-based experiments. In ISS, most dominant g-jitter frequency was noted to be ~110 Hz. ISS experiments suggested that the surface vibrations were mainly affected by the aspect ratio (length/diameter ratio), but not the imposed temperature gradient. Liquid bridge surface vibrations agreed well with Ichikawa et al.’s model. Ground-based experiments confirmed that increasing the volume ratio would cause the resonance frequency to increase. When a temperature difference was imposed between the upper and lower disks, for constant aspect and volume ratios, the resonance frequency tended to increase with the decreasing temperature difference. Furthermore, the shift in the resonance frequency due to a temperature difference, was found to be due to Marangoni convection and not due to reduced viscosity or surface tension of the fluid.

Effects of external vibrations (called g-jitter) on Marangoni convection in a liquid bridge were investigated on the International Space Station (ISS) and in ground-based experiments. In ISS, most dominant g-jitter frequency was noted to be ~110 Hz. ISS experiments suggested that the surface vibrations were mainly affected by the aspect ratio (length/diameter ratio), but not the imposed temperature gradient. Liquid bridge surface vibrations agreed well with Ichikawa et al.’s model. Ground-based experiments confirmed that increasing the volume ratio would cause the resonance frequency to increase. When a temperature difference was imposed between the upper and lower disks, for constant aspect and volume ratios, the resonance frequency tended to increase with the decreasing temperature difference. Furthermore, the shift in the resonance frequency due to a temperature difference, was found to be due to Marangoni convection and not due to reduced viscosity or surface tension of the fluid.

The analysis of electromyographic (EMG) signals detected during muscle contraction provides important information to aid in the diagnosis and characterization of neuromuscular disorders. One important analysis measures neuromuscular jitter, which is the variability of the time intervals between two muscle fibre potentials (MFPs) belonging to the same motor unit over a set of discharges. Conventionally, neuromuscular jitter is measured using single fibre (SF) EMG techniques, which can identify individual MFPs by using a SF needle electrode. However, SF electrodes are expensive, very sensitive to needle movement and not easy to operate in practise. <br /><br /> A method is studied in this thesis for automatically measuring neuromuscular jitter in motor unit potentials (MUP), it measures jitter using routine EMG techniques, which detect MUPs using a concentric needle (CN) electrode. The method is based on the detection of near MFP contributions, which correspond to individual muscle fibre contributions to MUPs, and the identification of individual MFP pairs. The method was evaluated using simulated EMG data. After an EMG signal is decomposed into MUP trains, a second-order differentiator, McGill filter, is applied to detect near MFP contributions to MUPs. Then, using nearest neighbour clustering and minimum spanning tree algorithms, the sets of available filtered MUPs can be selected and individual MFPs can be identified according to the features of their shapes. Finally, individual MFP pairs are selected and neuromuscular jitter is measured. <br /><br /> Using the McGill filter, near MFP contributions to detected CN MUPs can be consistently detected across an ensemble of successive firings of a motor unit. The method is an extension of the work Sheng Ma, compared to previous works, more efficient algorithms are used which have demonstrated acceptable performance, and which can consistently measure neuromuscular jitter in a variety of EMG signals.

The objective of the proposed research is to investigate new and more efficient techniques in numerical evaluation and experimental measurement of phase jitter impact on more general communication systems including dispersion management, filtering, and spectral inversion schemes. There has recently been a renewed effort to develop coherent optical communication systems. In particular, differential phase-shift keying (DPSK), which does not require a local oscillator to perform decoding, has focused the attention and is perceived to be the promising candidate for future optical communication systems updates. This motivates us to exploit DPSK in wavelength-division multiplexed systems. First, modulation formats based on phase show an increased robustness to nonlinear impairments such as cross-phase modulation (XPM) and nonlinear polarization rotation, primarily because the time-dependence of optical power is deterministic and periodic. Second, coherent formats allow a higher spectral efficiency since both in-phase and quadrature dimensions of the signal space are available to encode information. Optical phase is also used in intensity-modulated direct detection systems as an extra degree of freedom, for example to provide better resistance to intrachannel four-wave mixing (FWM), or to increase spectral efficiency in duobinary modulation. Finally, phase modulation outperforms its intensity counterpart in terms of sensitivity since a 3 dB improvement can be achieved when balanced detection is used. Nevertheless, DPSK-based formats show a different behavior to noise accumulated along the propagation. Noise-induced power fluctuations are converted into phase fluctuations by the Kerr effect and become a penalty source which limits the transmission system reach. In this context, there have been intense research activities for evaluating phase uncertainties but the previous studies assume an analytically determined pulse shape and a constant-dispersion optical link which is far from reflecting the actual and future structures of transmission lines.