In the thesis, an LED driver circuit that is applied in low power lighting LED with constant output current and Power Factor Correction (PFC) is presented. The insulated Flyback converter is used for the LED driver. Power Factor Correction is realized with both the method of Voltage Follower Approach Control under Discontinuous Conduction Mode and the method of Boundary Conduction Control under Boundary Conduction Mode. Compared with Voltage Follower Approach Control, Boundary Conduction Control needs only output current feedback. Moreover, it possesses lesser magnetize inductance current, lesser electrical stress of elements, more flexible choice of elements specification, smaller output current ripples, and higher power factor under light load. The circuit design is expounded, and verified by IsSpice simulation and experiment result.

abstract: A novel integrated constant current LED driver design on a single chip is developed in this dissertation. The entire design consists of two sections. The first section is a DC-DC switching regulator (boost regulator) as the frontend power supply; the second section is the constant current LED driver system. In the first section, a pulse width modulated (PWM) peak current mode boost regulator is utilized. The overall boost regulator system and its related sub-cells are explained. Among them, an original error amplifier design, a current sensing circuit and slope compensation circuit are presented. In the second section – the focus of this dissertation – a highly accurate constant current LED driver system design is unveiled. The detailed description of this highly accurate LED driver system and its related sub-cells are presented. A hybrid PWM and linear current modulation scheme to adjust the LED driver output currents is explained. The novel design ideas to improve the LED current accuracy and channel-to-channel output current mismatch are also explained in detail. These ideas include a novel LED driver system architecture utilizing 1) a dynamic current mirror structure and 2) a closed loop structure to keep the feedback loop of the LED driver active all the time during both PWM on-duty and PWM off-duty periods. Inside the LED driver structure, the driving amplifier with a novel slew rate enhancement circuit to dramatically accelerate its response time is also presented. / Dissertation/Thesis / Doctoral Dissertation Electrical Engineering 2016

The main goal of this diploma thesis is to compare two circuits designed for LED powering with output current of units of Amperes. Both circuits are based on step-down converter topology, one with technology of synchronous rectifying and the other one without it. Calculations and selection of used components with real prototypes were made for both selected circuits. All selected components meet automotive qualification requirements for discrete products. Produced prototypes were compared in terms of functionality, efficiency, EMC and thermal radiation.

This diploma thesis summarizes basic findings about issues of making development samples of PCB. The emphasis is especially on required quality which complies with mentioned norms. The theoretical section includes methodology for evaluating quality dismounted boards, assembling and soldering, parameters of components under different environmental circumstances. The practical part is carried out in cooperation with the department EEG in R&D Automotive Lighting Jihlava. It is dedicated to the design and the preparation of development samples where the quality is assessed according to given methodology from the theoretical part. PCB are analyzed by an X-ray, metallographic grinding and other methods. Recommendations are given and based on results for improvements.

This dissertation presents a comparative study about the performance of silicon (Si) and gallium nitride (GaN) semiconductors in drivers for light emitting diodes. Hereby, it is expected to provide the theoretical background required for the development of future works using this new technology. Theoretical aspects related to the materials used in the manufacture of semiconductors and their implications in the final product, as well as the characteristics and peculiarities of GaN semiconductors are presented. The experimental development consisted of two case studies, each focused on a distinct topology with different types of GaN semiconductors. First, a comparison of Si and enhancement mode GaN transistors was carried out in a family of synchronous buck converters. Ten 48 V to 28.3 V and 22.6 W converters were designed with the same parameters, at five different switching frequencies, ranging from 100 kHz to 1 MHz. Efficiency and temperatures were measured in four different scenarios: with and without an external diode in parallel with the low-side switch and with two different dead-time values, 25 ns and 50 ns. Converters with GaN transistors showed higher efficiency and lower temperatures in all cases, with a maximum efficiency of 96.8% and a minimum of 94.5%. In addition, Si-based converters exhibited greater performance degradation as the switching frequency and dead time increased. In the second study, nine 75 W off-line integrated double buck-boost converters were developed and evaluated. Two different Si technologies were compared with a cascode GaN transistor at three switching frequencies, ranging from 50 to 150 kHz. Again, the efficiency and temperatures of the prototypes were measured. The converters with GaN demonstrated superior performance in all cases, yielding about 5% gain in efficiency over the worst tested Si semiconductor. In both cases, the converters’ loss distribution was presented based on simulation results. It was concluded that the gallium nitride transistors have the potential to replace silicon technology mainly due to its superior performance and requirement of small, or no change, in the original circuit. / A presente dissertação apresenta um estudo comparativo do desempenho de semicondutores de silício (Si) e nitreto de gálio (GaN) em circuitos utilizados na alimentação de diodos emissores de luz. Por meio deste, procura-se fornecer o embasamento teórico necessário para o desenvolvimento de trabalhos futuros utilizando esta nova tecnologia. São apresentados, inicialmente, aspectos teóricos relacionados aos materiais utilizados na fabricação de semicondutores e suas implicações no produto final, além das características e peculiaridades dos semicondutores GaN. O desenvolvimento experimental consistiu de dois estudos de caso, cada um focado em uma topologia distinta com tipos de semicondutores GaN diferentes. Primeiramente, realizou-se um comparativo de transistores Si e GaN do tipo intensificação em uma família de conversores buck síncronos. Dez conversores 48 V para 28,3 V e 22,6 W foram projetados, com os mesmos parâmetros, em cinco diferentes frequências de comutação, variando de 100 kHz a 1 MHz. Eficiência e temperaturas foram medidas em quatro diferentes cenários: com e sem um diodo externo em paralelo com o interruptor de roda-livre e com dois valores diferentes de tempo morto, 25 ns e 50 ns. Os conversores com transistores GaN apresentaram maior eficiência e menores temperaturas em todos os casos, com uma eficiência máxima de 96,8% e uma mínima de 94,5%. Além disso, os conversores com Si exibiram uma maior degradação de desempenho à medida que a frequência de comutação e o tempo morto aumentam. No segundo estudo, nove conversores duplo buck-boost integrados de 75 W com alimentação a partir da rede elétrica foram desenvolvidos e avaliados. Compararam-se duas tecnologias distintas de interruptores de Si com um transistor GaN do tipo cascode, em três frequências de comutação, variando de 50 a 150 kHz. Novamente, mediu-se a eficiência e temperaturas dos protótipos. Os conversores com GaN demonstraram desempenho superior em todos os casos, com um ganho de cerca de 5% no rendimento em relação ao pior semicondutor Si testado. Em ambos os estudos de caso, a distribuição de perdas dos conversores foi apresentada com base em resultados de simulação. Concluiu-se que os transistores de nitreto de gálio têm potencial para substituir a tecnologia de silício utilizada atualmente devido, principalmente, a seu desempenho superior e exigência de pouca, ou nenhuma, mudança no circuito original.

Recent trends in mobile broadband indicates that the available radio frequency (RF) spectrum will not be enough to support the data requirements of the immediate future. Visible light communication, which uses visible spectrum to transmit wirelessly could be a potential solution to the RF ’Spectrum Crunch’. Thus there is growing interest all over the world in this domain with support from both academia and industry. Visible light communication( VLC) systems make use of light emitting diodes (LEDs), which are semiconductor light sources to transmit information. A number of demonstrators at different data capacity and link distances has been reported in this area. One of the key problems holding this technology from taking off is the unavailability of power efficient, miniature LED drive schemes. Reported demonstrators, mostly using either off the shelf components or arbitrary waveform generators (AWGs) to drive the LEDs have only started to address this problem by adopting integrated drivers designed for driving lighting installations for communications. The voltage regulator based drive schemes provide high power efficiency (> 90 %) but it is difficult to realise the fast switching required to achieve the Mbps or Gbps data rates needed for modern wireless communication devices. In this work, we are exploiting CMOS technology to realise an integrated LED driver for VLC. Instead of using conventional drive schemes (digital to analogue converter (DAC) + power amplifier or voltage regulators), we realised a current steering DAC based LED driver operating at high currents and sampling rates whilst maintaining power efficiency. Compared to a commercial AWG or discrete LED driver, circuit realised utilisng complementary metal oxide semiconductor (CMOS) technology has resulted in area reduction (29mm2). We realised for the first time a multi-channel CMOS LED driver capable of operating up to a 500 MHz sample rate at an output current of 255 mA per channel and > 70% power efficiency. We were able to demonstrate the flexibility of the driver by employing it to realise VLC links using micro LEDs and commercial LEDs. Data rates up to 1 Gbps were achieved using this system employing a multiple input, multiple output (MIMO) scheme. We also demonstrated the wavelength division multiplexing ability of the driver using a red/green/blue commercial LED. The first integrated digital to light converter (DLC), where depending on the input code, a proportional number of LEDs are turned ON, realising a data converter in the optical domain, is also an output from this research. In addition, we propose a differential optical drive scheme where two output branches of a current DAC are used to drive two LEDs achieving higher link performance and power efficiency compared to single LED drive.

The aim of this diploma thesis is to design two different switching power supply circuits with integrated circuits suitable for controlling of LED. The theoretical part describes the principles of two main switching converters topologies – buck and boost. Furthermore, a universal H-bridge structure of switching supply is described. This is compared with the typical switching supply design. Based on specified requirements, two circuits are designed with selected IO which are suitable for using in the automotive industry. Their function and typical schema are described. In the practical part is realized design with both selected IC, choosing components and layout PCB. Set of tests are carried out with manufactured samples. Based on measured results are performed optimalization of samples, which are tested also. In the end of this thesis both samples are compared in terms of function, efficiency and EMC.

Lighting consumes approximately 20-25% of the energy produced worldwide. LED based lighting is rapidly becoming the preferred choice over incandescent and fluorescent based lighting. LEDs have advantages such as high efficacy, long operating lifetime and excellent lumen maintenance. Therefore, to gain benefits from LEDs for lighting purposes, they must be driven with efficient drivers which maintain high LED efficacy and long LED lifetime. A review of existing LED drivers is done, and their advantages and drawbacks are identified. Existing fault-tolerant drivers are also reviewed. Several dimming methods and their effects on the LED efficacy and lifetime are investigated. As a result, a converter with coupled inductors, suitable as an LED driver which has high efficiency and can maintain high LED efficacy, incorporated with a high efficiency dimming method, is chosen. For the proposed LED driver, a comprehensive analysis on the effects of coupling type and coupling coefficient on converter performance is done. This is carried out to establish the best coupled inductor structure and coupling coefficient, for the proposed LED driver. The coupled inductor obtained is used to achieve high LED efficacy and also used to eliminate the need for an output filtering capacitor. This results in a highly compact, high efficiency and low cost LED driver. A lossless method of LED string current sensing is proposed, so that driver efficiency is not negatively impacted. The LED driver and a digital control system are designed, with the fault-tolerant feature incorporated. The LED driver and the control system are simulated and practically implemented. The results obtained show excellent LED driver performance. The fault-tolerant feature can enable the driver to operate under fault conditions, saving repair costs and down time. Additionally, a novel digitally controlled LED driver, which can drive several independent multiple LED loads, is proposed. This novel driver is simulated and practically implemented; with the results showing excellent driver performance. The novel LED driver can simplify and reduce costs of existing LED lighting systems. / Dissertation (MEng)--University of Pretoria, 2013. / gm2014 / Electrical, Electronic and Computer Engineering / Unrestricted

Light-emitting diodes (LEDs) offer energy-efficient lighting and are widely adopted. However, LED drivers that regulate power can fail when subjected to voltage disturbances on the electrical grid. This research investigates how components within LED drivers durability when undergoing voltage impulses and swells using simulation-based methods. An LED driver circuit was modeled in LTspice circuit simulation software. Impulses from 35-65V and equivalent voltage swells were applied to the simulated driver. The electrical stresses on components were statistically analyzed using the design of experiments and general full factorial. This methodology identified the most vulnerable components and their common durability/failure mechanisms during impulse and swell events. The findings provide insights into design changes that harden drivers against grid disturbances. This study determines that higher capacitor voltage ratings improved voltage impulse and swell withstand. Additionally, adding a surge suppression diode across the LED minimized diode reverse breakdown during swells. This simulation-based approach enables the informed design of robust LED drivers that can withstand electrical grid perturbations through strategic hardening of the most vulnerable components. The methodology and findings provide a framework for the reliability optimization of LED drivers and other power electronic systems exposed to power quality disturbances.Keywords: LED driver, voltage disturbance, component failure, circuit simulation, design of experiments (DOE), general full factorial.

No description available.