Light-emitting diode (LED) based visible light communication (VLC) has recently been promoted by lighting industry for providing services and information for consumers in shopping malls. Communication system designs require well defined signal-to-noise ratio (SNR). Being semiconductor devices, both of LEDs and photo-detectors are highly temperature-sensitive. This research explores the electrical and thermal sensitivities of LED and photo-detector together for VLC applications, and highlights the discoveries that the SNR of LED based VLC could vary substantially over the full operating power range of the LED lighting systems. A new analysis of the SNR based on the photo-electro-thermal theory has been developed. It can be used as the theoretical and design tool to predict such SNR variations. The validity of the new analysis has been confirmed with practical measurements that are consistent with theoretical predictions. The analytic approach can in principle be used as a design tool for LED based VLC.
21 December 2018 in IEEE Journal of Emerging and Selected Topics in Power Electronics
Based on the dynamic Photo-Electro-Thermal (PET) theory for LED systems, a comprehensive theory for analyzing and modeling phosphor coating in white light-emitting diodes (LEDs) as a dynamic light converter has been presented in Part-I of this paper. This Part-2 of the work provides the details of the model parameter determination, model setup and experimental verification. The procedures of obtaining the parameters in the experimental setup are explained. Using the comprehensive theory, a dynamic PET computer model of a phosphor-coated LED device has been developed. The simulation results compare well with the practical measurements, confirming the validity of the comprehensive theory in Part-1 of this paper. The new model provides novel equations for new information such as the power loss in phosphor coating and transients of the light emitted from the phosphor coating. Such information can be used for the designs of LED drivers and systems.
7 November 2017 in IEEE Transactions on Electron Devices
With the emergence of a multichip Light-emitting diodes (LEDs) device based on different chip arrays, a new modeling for a multichip LED device with different array becomes urgent. This paper presents a compact thermal model (CTM) for the multichip LED device with different chip array for nonuniform thermal properties. Based on a general 3-D CTM, LED chips can be arranged in various array forms according to the multichip LED construction design. By linking the thermal matrix to junction temperature prediction, the proposed CTM provides accurate predictions of temperature distribution of the multichip LED device, compared to the practical physical model and infrared radiation measurement. The modeling technique has been successfully demonstrated and verified. Based on the proposed method, this paper presents the multichip LED device with a staggered array to optimize the LED positions so that heat concentration of the LED array can be reduced.
13 September 2024 in IEEE Transactions on Electron Devices
Recent developments in artificial intelligence (AI) have triggered growing studies in artificial synaptic transistors. However, achieving tunable synaptic behaviors and improving synaptic performance by simple processes is still challenging. Herein, a facile regulation strategy by introducing a polymer insulator to a semiconductor was proposed, for the first time, to implement an optoelectronic synaptic transistor (OST) using a solution-based method. The organic OST comprises a biodegradable polyvinyl alcohol (PVA) electret with hydroxyl group dipoles, enabling the effectively coordinated regulation of synaptic weight and the enhancement of synaptic properties. Essential synaptic characteristics were realized, such as paired-pulse facilitation (PPF), excitatory postsynaptic currents (EPSCs), long-term potentiation, and depression (LTP/D). The results indicate that enhanced synaptic performance should be associated with the improvement of thin film quality and surface morphologies due to blending insulators, as well as the carrier migration between semiconductors and PVA. More importantly, the controllable light logical operation was perfectly simulated. Pattern recognition test of MNIST handwritten digits was also researched based on artificial neural networks (ANNs), a high recognition accuracy of 91.38% was achieved using appropriate insulator blending ratios. Therefore, our work should open a new pathway for designing optoelectronic synapse devices and expand their further application in brain-inspired computing.
3 March 2023 in IEEE Transactions on Instrumentation and Measurement
Given the growing interest in micro-scale thermal management of mini- and micro-light-emitting diodes (LEDs), it’s crucial to develop an effective two-dimensional (2D) junction temperature measurement method. In this paper, the microscopic hyperspectral imaging (μ-HSI) technique has been proposed to probe the 2D photo-thermal distribution of the flip-chip mini-LEDs. Running on a two-axis rotating platform, μ-HSI is able to directly focus on the epitaxial layer surface (episurface) and lateral side of the LED chips to acquire their 2D spectral distributions. Subsequently, the full width at half maximum matrix was extracted to deduce the 2D junction temperature distributions. As a result, the location and temperature distribution of the active layer, i.e., PN junction, can be determined accurately. The mean temperature of the episurface approximately matches that of the active layer. The Raman shift method was applied to verify the accuracy of the μ-HSI technique. In addition, the extra fine structures as well as temperature distributions of the interface between the episurface and sapphire or electrodes are distinguished and analyzed by this technique. This work offers an effective tool for the quality improvement and thermal management of mini-LEDs.
Based on the energy flow inside phosphor materials, a prediction model and its modeling method for the thermal effect and optical properties of phosphor converted white light emitting diode (LED) are proposed in this paper. With the phosphor characteristics (excitation wavelength, phosphor temperature, thickness, light absorption and conversion coefficients), it is possible to accurately quantify the heat dissipation coefficient and optical power of the white LED device, and then to estimate the emission power and luminous flux. The modeling consists of a series of optical and thermal evaluation, which are simple for LED designers to comply. Using samples of LED sources and phosphor material, the proposed prediction model has been validated in experiments, where good agreements between calculations and practical measurements have been achieved.