Univ., Diss.--Hannover, 2001.

The thesis treats the topic of the characterization and optimization of CMOS compatible photo sensor elements (CPSE). To understand the functionality of a CPSE a pro founded basic knowledge is necessary in the fields of the ray-, wave and quantum optics. Therefore all in this context necessary and principal equations are presented at the beginning of this work. In addition belongs in particular the statistic investigation of the absorption and emission processes in the field of the quantum optics, since these processes are the basis to understand the elementary noise sources within a CPSE. After an introduction to the available technological realization each element of the equivalent circuit diagram is analyzed. The optical quantum efficiency is developed intended for arbitrary angles and optical multi-layer sequences. The internal quantum efficiency is determined with a new mathematical approach and offers thereby more possibilities of the interpretation of the results compared to the classical physical approach which usually proceed from approximation solutions. The physical operation limit of a sensor and thus its efficiency is specified by the power of the noise sources. In this work it is analytically proven that the statements of Einstein concerning the kTC noise of an individual RC combination also apply to coupled RC combinations. Thus a highly simplified viewpoint of the CPSE and the Switched Capacitor amplifier is possible. Several optimization strategies to improve the optical quantum efficiency, for example the oxide perforation, are introduced and realized. The electro optical parameters can be usually determined by the measurement of special test-structures. Furthermore in this work new test-structures are developed, manufactured and measured under the aspect of high insensitivity to measurement errors. These results lead to the development of an Isochor diagram which permits all important design parameters. The Isochor diagram contributes thereby considerably to the conceptual design of a high-dynamic camera including image sensor. In particular the demanded application in the automotive field can be considered in detail. The achievement of the required goal parameters is documented by the measurement of the image sensor in the camera.

A problem in todays Internet infrastructure may occur when a streaming multimedia application is to take place. The information content of video and audio signals that contain moving or changing scenes may simply be too great for Internet clients with low bandwidth capacity if no adaptation is performed. In order to satisfactorily reach clients with various bandwidth capacities some works such as receiver-driven multicast and resilient overlay networks (RON) have been developed. However these efforts mainly call for modification on router level management or place additional layer to the Internet structure, which is not recommended in the nearest future due to the highly acceptance level and widely utilization of the current Internet structure, and the lengthy and tiring standardization process for a new structure or modification to be accepted. We have developed an improved active network approach for distributed and dynamic streaming multimedia environment with heterogeneous bandwidth, such as the case of the Internet. Friendly active network system (FANS) is a sample of our approach. Adopting application level active network (ALAN) mechanism, FANS participants and available media are referred through its universal resource locator (url). The system intercepts traffic flowing from source to destination and performs media post-processing at an intermediate peer. The process is performed at the application level instead of at the router level, which was the original approach of active networks. FANS requires no changes in router level management and puts no additional requirement to the current Internet architecture and, hence, instantly applicable. In comparison with ALAN, FANS possesses two significant differences. From the system overview, ALAN requires three minimum elements: clients, servers, and dynamic proxy servers. FANS, on the other hand, unifies the functionalities of those three elements. Each of peers in FANS is a client, an intermediate peer, and a media server as well. Secondly, FANS members tracking system dynamically detects the existence of a newly joined computers or mobile device, given its url is available and announced. In ALAN, the servers and the middle nodes are priori known and, hence, static. The application level approach and better performance characteristics distinguished also our work with another similar work in this field, which uses router level approach. The approach offers, in general, the following improvements: FANS promotes QoS fairness, in which clients with lower bandwidth are accommodated and receive better quality of service FANS introduces a new algorithm to determine whether or not the involvement of intermediate peer(s) to perform media post-processing enhancement services is necessary. This mechanism is important and advantageous due to the fact that intermediate post-processing increases the delay and, therefore, should only be employed selectively. FANS considers the size of media data and the capacity of clients bandwidth as network parameters that determine the level of quality of service offered. By employing the above techniques, our experiments with the Internet emulator show that our approach improves the reliability of streaming media applications in such environment.

The increasing complexity of integrated circuits (IC) requires IC internal contactless test-techniques for their function and failure analysis. The magnetic force microscopy-based current contrast imaging test-system makes the measurement of IC internal currents and current contrast imaging of IC areas possible. In this book a magnetic force microscopy (MFM) based current contrast imaging (CCI) test-system is presented. This new technology offers a promising tool for advanced contactless circuit internal function and failure analysis of IC. Contactless current images of 100nm wide interconnection lines and currents down to 1.7µA are presented. In this work the first high frequency current measurements up to 4.7 GHZ are demonstrated. Furthermore, there are some new detrimental effects on the technique which are discovered, explained by a new model and some possibilities for elimination of these effects are developed and presented. At least a failure analysis at a real IC is demonstrated.

The capability of todays electronics bases on the use of integrated circuits, which contain millions of electronic structures. The trend towards decreasing structure width and increasing working frequencies of the integrated circuits produces a demand for special test techniques for function and failure analysis of new developed integrated circuits. The electric force microscopy (EFM) is a promising new test technique for contact-less, chip-internal voltage measurements with simultaneous high temporal and high spatial resolution. In this thesis the behaviour of the EFM measurement signal is investigated regarding dynamic voltage measurements on parallel sub-micrometer conducting lines. Therefore, the influence of the structure width, the probe geometry and of cross-talk on the measurement signal is analysed by means of measurements, calculations and simulations. Furthermore, a broad overview about the properties and possibilities of different EFM-measurement set-ups is given and the suitability of the so called "heterodyne mixing technique" for the measurement of periodic digital voltage is investigated.

The high degree of automation in industrial manufacturing and in many areas of the daily life requires as well as the continuous rising need for information of the society an technical measurement monitoring of various process and environmental parameters. Sensors convert the instantaneous value of a physical measured variable on basis of physical or chemical effects into an equivalent electrical signal. They build the interface between a technical system and its environment. Magnetic field sensors, like the sensor elements Hall-Element and magneto-resistor which are based on galvanomagnetic ef-fects, react to a magnetic interaction with the item under test or cause such an interac-tion. The starting point of the present study is built by the treatment of the components of galvanomagnetic sensor systems as shape elements out of defined material, which are used as information carrier by modification of the constructional parameters. The variation of number, arrangement and/or geometrical shape results in one- or multidi-mensionally structured system components, which allow the generation of defined sen-sor system topologies with custom-designed information content. Within the framework of this work different new sensor systems were developed, which open a broader appli-cation scope to the galvanomagnetic sensor technology by structured components and which permit the application of these sensors within purviews of the measuring tech-nique, which are dominated so far by sensors on basis of other, usually optical effects. If the allocation of a sensor signal to the structure of a system component is reproducible and biunique, then the principle of the structure variation also admits a purposeful cod-ing of defined discrete information. For this a beginning for the development of magnetic structure codes for the identification technology forms the conclusion of this thesis.

This work explains a method for optoelectronic integration of an heterojunction bipolar transistor (HBT) and a waveguide electroabsorption modulator (EAM). For this the epitactical layers of the individual devices based on III/V-semiconductors are merged into each other. The method has the advantage to reuse layers and results in a layer stack, which is easier to process technologically. Additionally to the manufacturing of the individual devices HBT and EAM, this integration enables a multifunctional device, which works in the optical and electronic regime simultaneously (HBT-EAM). This corresponds to an EAM with integrated amplifier. Presented are epitaxy, technological processing and measurement results.

In this thesis a circuit technology is described, which enables an integrated controling of an active matrix of micromechanic spatial light modulators (SLM). Possible applications areas for micromechanic spatial light modulators are projection of still or moving images, maskless lithography and adaptive optics. Especially in the areas of lithography and adaptive optics typical requirements of light modulators are high optical resolutions, high accuracy of modulation and fast addressing in order to enable a fast image repetition rate. The developed circuits are optimised for use as a maskless litography system and provide a minimum pixel pitch of 16 <font face="Symbol">m</font><i>m</i>. Demonstrators with image areas of 256 x 256 pixels and a modulation resolution of 16 steps on each pixel enabled a minimum geometric lithographic resolution of less than 20 nm, using today available optics.

In the field of radiological diagnostics, especially for the computed tomography a large number of photodetector elements (so called Pixels) is needed to assemble a large-area X-Ray-sensitive detector. These pixels should provide very high sensitivity, a very large dynamic range and a high signal-to-noise-ratio. Moreover, an in-pixel digitization of the signals is desired. To meet these specifications, a digital CMOS-Pixel comprising a special type of photodiode and a single-loop sigma-delta-modulator was developed, theoretically analyzed, fabricated and characterized. A matrix of 4x4 digital CMOS-pixels with each of them consuming an area of 1mm² has been fabricated. For further signal processing the concept of a multichannel decimation filter is presented.

This dissertation introduces an implantable LED microdisplay in the aim of returning a certain ability to see to blind people. The microdisplay is the basic device of the intraocular vision aid (IoVA) an implant developed for those whose cornea of the eye is considerably blurred due to accident or illness. This microdisplay consists of a LED-array which is connected to a CMOS driver circuit. By means of activating each pixel an image of the surroundings is generated and directly projected onto the retina. Within the scope of this dissertation the concepts and techniques of the intraocular vision aid are described. Physiological and technical parameters for application of the microdisplay are explained. That followed the material and applied technologies are defined. Finally the technological realization of the LED microdisplay is demonstrated as well as the first functional model on a worldwide scale.