| Popis: |
In recent years, high brightness light emitting diodes (HBLEDs) have increasingly attracted the interest of both industrial manufacturers and academic research community. Among the several aspects that make LED technology so attractive, the most appreciated characteristics are related to their robustness, high efficiency, small size, easy dimming capability, long lifetime, very short switch-on/switch-off times and mercury free manufacturing. Even if all such qualities would seem to give to solid state lighting a clear advantage over all the other kinds of competing technologies, the issues deriving from the need of LED technology improvement, on one hand, and of the development of suitable electronic ballasts to properly drive such solid state light sources, on the other, have so far hindered the expected practical applications. The latter problem, in particular, is nowadays considered the main bottleneck in view of a widespread diffusion of solid state technology in the general lighting market, as a suitable replacement of the still dominant solutions, namely halogen and fluorescent lamps. In fact, if it is true that some aspects of the devices’ technology (e.g. temperature dependent performance, light quality, efficiency droop, high price per lumen, etc…) still need further improvements, it is now generally recognized that one of the key requirements, for a large scale spread of solid state lighting, is the optimization of the driver. In particular, the most important specifications for a LED lamp ballast are: high reliability and efficiency, high power factor, output current regulation, dimming capability, low cost and volume minimization (especially in domestic general lighting applications). From this standpoint, the main goal is, therefore, to find out simple switched mode power converter topologies, characterized by reduced component count and low current/voltage stresses, that avoid the use of short lifetime devices like electrolytic capacitors. Moreover, if compactness is a major issue, also soft switching capability becomes mandatory, in order to enable volume minimization of the reactive components by increasing the switching frequency in the range of the hundreds of kHz without significantly affecting converter’s efficiency. It is worth mentioning that, in order to optimize HBLED operation, also other matters, like the lamp thermal management concern, should be properly addressed in order to minimize the stress suffered by the light emitting devices and, consequently, the deterioration of the light quality and of the expected lamp lifetime. However, being this work focused on the issues related to the research of innovative driving solutions, the aforementioned thermal management problems, as also all the topics related to the improvement of solid state devices’ technology, will be left aside. The main goal of the work presented in this thesis is, indeed, to find out, analyze and optimize new suitable topologies, capable of matching the previously described specifications and also of successfully facing the many challenges dictated by the future of general lighting. First of all, a general overview of solid state lighting features, of the state of the art of lighting market and of the main LED driving issues will be provided. After this first introduction, the offline driving concern will be extensively discussed and different ways of approaching the problem, depending on the specific application considered, will be described. The first kind of approach investigated is based on the use of a simple structure relying on a single power conversion stage, capable of concurrently ensuring: compliance with the standards limiting the input current harmonics, regulation of the load current and also galvanic isolation. The constraints deriving from the need to fulfil the EN 61000-3-2 harmonics standard requirements, when using such kind of solution for low power ( |
