Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules

Current-voltage characteristics of real high-power LED chips and their defects were studied. Process of modeling LED matrix by using Monte Carlo method based on the measured real LED chips was considered and performed. It was shown that the risk of variating LED chip parameters for stability of a fi...

Повний опис

Збережено в:
Бібліографічні деталі
Дата:2015
Автор: Shynkarenko, V.V.
Формат: Стаття
Мова:English
Опубліковано: Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України 2015
Назва видання:Semiconductor Physics Quantum Electronics & Optoelectronics
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/121145
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules / V.V. Shynkarenko // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2015. — Т. 18, № 1. — С. 200-204. — Бібліогр.: 5 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id nasplib_isofts_kiev_ua-123456789-121145
record_format dspace
spelling nasplib_isofts_kiev_ua-123456789-1211452025-02-23T17:31:34Z Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules Shynkarenko, V.V. Current-voltage characteristics of real high-power LED chips and their defects were studied. Process of modeling LED matrix by using Monte Carlo method based on the measured real LED chips was considered and performed. It was shown that the risk of variating LED chip parameters for stability of a final LED matrix as a result of the non-uniform heating up of individual elements in a series-parallel circuit. The solution allowing to significantly reducing the influence of LED chips input parameters variationon the final parameters of LED matrix was proposed. The author expresses his sincere gratitude to the corresponding member of NAS of Ukraine V.M. Sorokin for useful discussions and interest to the work, professor R.V. Konakova for discussions and her help in writing this paper, and senior researcher Ya.Ya. Kudryk for substantial assistance in performing the experiments. 2015 Article Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules / V.V. Shynkarenko // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2015. — Т. 18, № 1. — С. 200-204. — Бібліогр.: 5 назв. — англ. 1560-8034 DOI: 10.15407/spqeo18.02.1200 PACS 85.60.Jb https://nasplib.isofts.kiev.ua/handle/123456789/121145 en Semiconductor Physics Quantum Electronics & Optoelectronics application/pdf Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
description Current-voltage characteristics of real high-power LED chips and their defects were studied. Process of modeling LED matrix by using Monte Carlo method based on the measured real LED chips was considered and performed. It was shown that the risk of variating LED chip parameters for stability of a final LED matrix as a result of the non-uniform heating up of individual elements in a series-parallel circuit. The solution allowing to significantly reducing the influence of LED chips input parameters variationon the final parameters of LED matrix was proposed.
format Article
author Shynkarenko, V.V.
spellingShingle Shynkarenko, V.V.
Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules
Semiconductor Physics Quantum Electronics & Optoelectronics
author_facet Shynkarenko, V.V.
author_sort Shynkarenko, V.V.
title Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules
title_short Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules
title_full Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules
title_fullStr Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules
title_full_unstemmed Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules
title_sort development of modeling methods and algorithm for calculating the current-voltage characteristics of led modules
publisher Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
publishDate 2015
url https://nasplib.isofts.kiev.ua/handle/123456789/121145
citation_txt Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules / V.V. Shynkarenko // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2015. — Т. 18, № 1. — С. 200-204. — Бібліогр.: 5 назв. — англ.
series Semiconductor Physics Quantum Electronics & Optoelectronics
work_keys_str_mv AT shynkarenkovv developmentofmodelingmethodsandalgorithmforcalculatingthecurrentvoltagecharacteristicsofledmodules
first_indexed 2025-11-24T02:25:03Z
last_indexed 2025-11-24T02:25:03Z
_version_ 1849636792769183744
fulltext Semiconductor Physics, Quantum Electronics & Optoelectronics, 2015. V. 18, N 2. P. 200-204. doi: 10.15407/spqeo18.02.1200 © 2015, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 200 PACS 85.60.Jb Development of modeling methods and algorithm for calculating the current-voltage characteristics of LED modules V.V. Shynkarenko V. Lashkaryov Institute of Semiconductor Physics, NAS of Ukraine, 41, prospect Nauky, 03028 Kyiv, Ukraine; e-mail: shynkarenko@isp.kiev.ua Abstract. Current-voltage characteristics of real high-power LED chips and their defects were studied. Process of modeling LED matrix by using Monte Carlo method based on the measured real LED chips was considered and performed. It was shown that the risk of variating LED chip parameters for stability of a final LED matrix as a result of the non-uniform heating up of individual elements in a series-parallel circuit. The solution allowing to significantly reducing the influence of LED chips input parameters variation on the final parameters of LED matrix was proposed. Keywords: LED, modeling, current-voltage characteristics. Manuscript received 09.12.14; revised version received 23.03.15; accepted for publication 27.05.15; published online 08.06.15. Transition from the LED chips to LED modules has led to the need to solve a number of additional tasks – analysis of heat distribution in a similar module [1-3], but it ignores the fact that the LED chips that are the basis for LED modules have variation of parameters, which leads to the non-uniform current flow. To estimate this factor, it is necessary to have software to predict the effect of the connecting the real LED chips on the parameters of the future LED module. It is important to take into account the possible defects of LED chips (their errors on quality control, the influence of defects location, variation of parameters) and to calculate their effect on the general characteristics of the final LED module. This paper describes the stages of creating and first results of operating such a program that allows to simulate the behavior of the final LED module, in dependence on the number and charac- teristics of the elements in each circuit of series-parallel connection. To obtain information about the actual LED chips, the LED modules prepared in the conditions of domestic production based on the chips of company “EDISON Optocorporation” before filling them with phosphor described in [4] were investigated. An appearance of the LED module is shown in Fig. 1. The module is made on the basis of the printed circuit board with an aluminum basis, on which LED chips were mounted and connected to each other by a series-parallel circuit shown in Fig. 2. Two identical modules were studied. To study the current-voltage characteristics (CVC) of real chips on the given module, the connect tracks of one LED module were cutted and wire terminals were connected to each individual chip. The direct measurements of each chip were performed with taking into account the chip location on the printed circuit board and its connection. The CVC measurements of LED module chip are shown in Fig. 3. A number of anomalies in the current- voltage characteristics were detected. These anomalies were caused by soldering defects (Fig. 3a, chips with abnormally low currents at the applied voltage 3 V, their photo is shown in Fig. 4), as well as defects in these chips (Fig. 3, three curves for relatively high direct currents with the applied forward (a, b) and reverse (c) voltages 1 V), which can accelerate their degradation. Semiconductor Physics, Quantum Electronics & Optoelectronics, 2015. V. 18, N 2. P. 200-204. doi: 10.15407/spqeo18.02.1200 © 2015, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 201 a) b) Fig. 1. Appearance of the LED module (a), the numbering chips (b). Fig. 2. The series-parallel circuit connection of chips in the LED module. Two chips with higher values of series resistance attracted attention to them (see the dotted curves in Fig. 3a). The current through these chips was significantly lower than that through the middle chips. It leads to overloading the neighboring, parallel to them, LED chips and, as a result, to acceleration of their degradation during operation. For further analyzing, each CVC of LED chip seemed to be approached in the approximation of two parallel ideal diodes and serial nonlinear resistance (Fig. 5). In this case, CVC of high-power LED is described by the equation (1) 1 2 21010 11 10               URURReIeI I UCUC , (1) where I1 and І0 are the cut-off currents for diodes D1 and D0, C0 and C1 are the corresponding exponential factors, R1 is the coefficient associated with resistance as well as R0 and R2 take into account the non-linearity of the resistance on heating. In the program, to simplify the calculations, this formula evolves to the form . 1 10101010 1 1 2 210 1100               URURR I UCiUCi (2) 0 1 2 3 4 10 -11 10 -9 10 -7 10 -5 10 -3 10 -1 I, A U, V Normal CVC CVC with high current CVC of chip with soldering defect a) 1 2 3 4 0.00 0.05 0.10 0.15 Normal CVC CVC with high current CVC of chip with soldering defect I, A U, V b) 0 1 2 3 4 10 -11 10 -9 10 -7 10 -5 10 -3 10 -1 Normal CVC CVC with high current CVC of chip with soldering defect I, A U, V c) Fig. 3. Forward (a, b) and reverse (c) branches of CVC of LED module chips: (a) in linear coordinates, (b) in semi-logarithmic ones. The numbering is the same as that in Fig. 1b. Semiconductor Physics, Quantum Electronics & Optoelectronics, 2015. V. 18, N 2. P. 200-204. doi: 10.15407/spqeo18.02.1200 © 2015, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 202 Study of coefficients of real chips indicated the clear (more than 0.95) linear correlation between the parameters of each element model (i0 and C0, i1 and C1, R0, R2 and R2, respectively). Basic values of coefficients are listed in Table 1. Table 1. Parameters approximation coefficients of the 21 LED chips in (2). P ar am et er T h e ar it h m et ic m ea n M ed ia n (m o st p ro b ab le ) S ta n d ar d d ev ia ti o n A sy m m et ry co ef fi ci en t i0 –27.6983 –27.7038 0.600160 0.635762 i1 –19.4955 –19.4671 0.646147 1.341593 R0 1.5737 1.6687 0.374458 –0.914235 a) b) Fig. 4. Photos of chips with drawn contacts (a) and defect of soldering the contact (b). D1 D0 Rheat Fig. 5. Model of high-power LED: D0, D1 – ideal diodes, Rheat – non-linear resistance responsible for the heating up of a high- power LED during operation. By comparing the values of the arithmetic mean and the median (Table 1), one can see that the distribution of most coefficients is not significantly different from normal one, although the parameter of nonlinear resistance have small (about 10%) peak asymmetry. It indicates a greater number of small (close to the minimum values) R0 values. The calculated values of the asymmetry coefficient indicate distributions asymmetry of i1, i0 values and median bias towards the smaller values. In general, the data deviation from the normal distribution can be considered as non-significant, which allows to use a normal symmetrical distribution with the calculated median as the expected value in the model. Because of the complexity to account for the small values of the terms, each of denominators of the fractions were separately calculated, then compared with others, the general current was calculated (Fig. 6) basing on the comparison results. Fig. 6. Algorithm for calculating the CVC for high-power LED with known parameters. Semiconductor Physics, Quantum Electronics & Optoelectronics, 2015. V. 18, N 2. P. 200-204. doi: 10.15407/spqeo18.02.1200 © 2015, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 203 2 3 10 -10 10 -8 10 -6 10 -4 10 -2 10 0 I, A U, V Fig. 7. Real current-voltage characteristics of LEDs (black) and modeled current-voltage characteristics, based on them, by the Monte-Carlo method (white). 10 12 14 16 18 20 22 24 1E-10 1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 0.01 0.1 I, A U, V Experiment Modeling Fig. 8. Comparison of current-voltage characteristics of the LED module in the operation voltage range and the modeled current-voltage characteristics of the LED module with a variation of chip properties. As in the most programming languages there is only a random number generator with a uniform distribution. For generating variation of LED parameters coinciding with the normal distribution, we used the formula [5]:   22 22 00 ln2 ),( ba ba axx    , where х0 is the expected value, σ 2 – dispersion of a random value, a and bare random variables with uniform distribution, in which relations –1 < a < 1 and –1 < b < 1 are valid. To test the results of modeling the LED CVC by using the Monte Carlo method, we selected group of initial parameters obeying the specified normal distribution (Fig. 7, white curves). For comparison, in the same Fig. 7 CVC of actual LED chips (excepting the anomalous CVC of chips) are represented with black colour. A good agreement of the model with the experiment not only in the operation limits of voltages (2.8…3 V), but also in the wider range is seen. Table 2. Mean values and deviation of currents flowing through LEDs in the LED matrix. N u m b er o f m o d el ed L E D M ea n c u rr en t, А S ta n d ar d d ev ia ti o n M in im u m v al u e, А M ax im u m v al u e, А 400 0.248206 0.045829 0.130364 0.390092 400 0.105404 0.017155 0.058014 0.153495 400 0.084945 0.013367 0.049298 0.126556 400 0.075775 0.012356 0.043098 0.115647 400 0.063335 0.009298 0.037065 0.089466 400 0.050814 0.007464 0.029898 0.07512 400 0.025733 0.002934 0.01736 0.033899 400 0.012698 0.000994 0.009763 0.015713 400 0.005002 0.000199 0.004354 0.00559 Calculating CVC of the group of high-power LEDs, the size of the LED matrix and the type of their connection were determined, and then the iterative methods were used. Voltage was calculated on a group of LEDs connected in parallel, and then, on the basis of the obtained results, currents flowing through each LED in the curcuit with accounting the general resistance were measured. From the obtained data, the flowing current through each LED in the matrix was modeled (Table 2), and the distribution of currents was examined. The results indicate that variation of the initial parameters of the LED chips leads to a greater variation between them immediately during operation, without effecting significantly on the total current-voltage characteristics (Fig. 8). At the same time, the integral luminosity of the matrix is close to the nominal one. Even when an error welding of conduction contacts to the chip occurs, the chips in parallel will shine brighter due to the increased current flow, which will result to 5…7% reduction in the total integrated luminosity of general LED matrix. At the same time, this variation leads to overheating and rapid degradation of the individual LED chips, which can significantly effect on the service life of the LED matrix. Conclusions Analysis of the current-voltage characteristics of high- power LED chips of the company “EDISON Optocorporation” in the composition of LED modules had shown the presence of defects in connecting the wires to the chip and abnormally high current flow upon applied low voltages (the first of which is a defect of manufacturer of LED modules, the second – defect of chip supplier). The developed software allows to simulate numerically the behavior of LED modules connected practically in any combination, to analyze the currents through any part of a modeled switching circuits. Modeling the LED chip by two parallel ideal diodes and Semiconductor Physics, Quantum Electronics & Optoelectronics, 2015. V. 18, N 2. P. 200-204. doi: 10.15407/spqeo18.02.1200 © 2015, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 204 consistent nonlinear resistance, manifested in the heating up stage of CVC, had shown high correlation coefficients between the internal parameters of individual elements and close to normal distribution of the basic parameters of these elements. This result confirms the correctness of the chosen model. Analysis of existing switching-on of chips for a series-parallel circuit had shown that verification of the integrated luminosity or CVC of final product does not allow to identify the defects of supply wires or overvalued current transport due to the redistribution of the current in the LED module. However, this redistribution leads to an increase of current through the neighboring LEDs and their local overheating on the plate, which should lead to acceleration of degradation of the LED module. Addition of internal sorting lines grouping the LED chips by the current value at operation voltages will allow to reduce variation between the chips and to increase the service life of finished products. Acknowledgements The author expresses his sincere gratitude to the corresponding member of NAS of Ukraine V.M. Sorokin for useful discussions and interest to the work, professor R.V. Konakova for discussions and her help in writing this paper, and senior researcher Ya.Ya. Kudryk for substantial assistance in performing the experiments. References 1. Luqiao Yin, Lianqiao Yang, Weiqiao Yang et al., Design and analysis of multi-chip LED module with ceramic substrate // Solid-State Electronics, 54(12), p. 1520-1524 (2010). 2. Ye Huaiyu, SauKoh, Henk van Zeijl, A.W.J. Gielen, Zhang Guoqi, A review of passive thermal management of LED module // J. Semiconductors, 32(1), 014008 (2010). 3. Dun-Yen Kang, Enboa Wu, Da-Ming Wang, Modeling white light-emitting diodes with phosphor layers // Appl. Phys. Lett. 89(23), 231102 (2006); doi: 10.1063/1.2400111. 4. V.V. Shynkarenko, Ya.Ya. Kudryk, I.V. Mukha, R.Ya. Kudryk, Modeling the characteristics of a high-power LED module with real chips // Abstracts of the 6-th Intern. Sci.-Techn. Conf. “Sensor Electronics and Microsystem Technologies” (SEMST-6). – Ukraine, Odessa, 2014, Sept. 29 – Oct. 3, p. 58 (in Ukrainian). 5. J. Bucknall, The Tomes of Delphi Algorithms and Data Structures. Wordware Publ., Inc. Co., 2002.