Several questions about increasing LED power
The temperature rise effect of LED devices and its countermeasures elaborate on the influence of junction temperature rise on the LED light output intensity, the forward voltage of LED PN junction, and the color of light emission. It is pointed out that when the junction temperature increases, the output light intensity becomes weaker, the forward voltage decreases, and the emission wavelength shifts red. When the junction temperature rises high enough, these changes will change from reversible to irreversible permanent decay. The article further points out that LED input power is the only source of thermal effect of the device, and trying to improve the electro-optic conversion efficiency of the device and improve the heat dissipation capacity of the device are the main ways to reduce the temperature rise effect of the LED.
1. Introduction As we all know, LED is an electroluminescent device, its basic physical process is the conversion of electrical energy to light energy. The so-called increase in LED power is to increase the electrical input energy and at the same time obtain the largest possible optical power output. The light energy (luminous flux) generated by a unit of input power is usually referred to as the photoelectric conversion power, or light effect for short. Early LEDs had low light efficiency (-0.1 lm / w) and low brightness, and were usually only used to indicate bright and dark states as indicator lights. In the early 1990s, the emergence of ultra-bright quaternary LEDs led to an order of magnitude increase in device brightness, especially the advent of GaN-based blue, green, and white LEDs, which led to the application of LEDs. significant change. Solid-state lighting has become an important goal pursued by mankind in the 21st century. Obviously, continuously improving the input power and luminous efficiency of LED is the only way to achieve general lighting. Assuming that the luminous efficiency of the LED is 100 lm / w, then to reach the luminous flux of 600 lm emitted by an incandescent lamp of 40 lights (watts), the input power of the LED must reach 6w. However, the input power of a Φ5 standard LED is currently 0.04 to 0.07w, which is far from meeting the needs of practical lighting. A lot of practice shows that the basic reason why the LED cannot increase the input power is that the LED will emit a lot of heat during the operation, causing the die junction temperature to rise rapidly. The higher the input power, the greater the heating effect. The increase in temperature will cause changes and attenuation of device performance, and even failure. This article makes some concise discussions on the influence of the heating effect on the performance of power devices and how to reduce this heating effect.
2. The temperature rise of the LED device is estimated to set the wafer area to 1.2 × 1.2mm2, the thickness to 200um, and the GaAs substrate. Since the epitaxial layer is very thin, the difference between the material of the epitaxial layer and the substrate is ignored, and the influence of the electrode is not considered, then the volume of the wafer is about 2.88 × 10 × 4 cm3. The specific gravity of the GaAs crystal is 5.318 (g / cm3), so the wafer weight is about 15.3 × 10-4 g. Let the working current of the device be 100mA, if about 90% of the electric power is converted into heat, then the device can calculate the temperature of the wafer to reach 5 × after the device is turned on for 20 minutes without considering the heat dissipation from the wafer to the surrounding environment 105 ° C, the specific heat data of the GaAs crystal used in the calculation is 0.33 Joules / gram o degree. The severity of its temperature rise effect can be seen. Here only the wafer is considered as a uniform heating element. If the concentration effect of temperature rise at the junction is considered, the situation will be more serious. Fortunately, during the heating process of the wafer, the wafer cannot be in an adiabatic state. Instead, it always exchanges heat with the surrounding medium or environment in some way, and finally reaches thermal equilibrium to maintain the temperature of the wafer at a low temperature. On the level.
3. The influence of junction temperature on LED performance
1. The effect of junction temperature on LED light output. Experiments indicate that the light output intensity of InGaAlP LEDs emitting red and yellow light and InGaN LEDs emitting blue and green light depend significantly on the junction temperature of the device. In other words, when the junction temperature of the LED increases, the output light intensity of the device will gradually decrease; and when the junction temperature decreases, the light output intensity will increase. Generally, this change is reversible and recoverable When the temperature returns to the original value, the light intensity will return to the original state. Figure 1 (a) indicates the change in the relative amount of light output of InGaAlP LEDs with temperature. Here, 25 ° C is used as the reference point for device performance. It can be seen from the figure that InGaAlP orange LEDs have higher temperature sensitivity than red LEDs. When the junction temperature rises to 100 ° C, the output flux of the amber device drops by 75%. Figure 1 (b) is another set of InGaAlP LED light output temperature data, set the value of the LED to 100 when 25? C, then when the junction temperature rises to 100? C, 640nm, 620nm and 590nm InGaAlP LED light The output is 42%, 30% and 20% of the original value.
The mathematical expression of the effect of junction temperature on light output is shown in equation (1):
ΦV (T2) = ΦV (T1) e-kΔT (1)
Among them, ΦV (T2) represents the luminous flux output of the junction temperature T2; ΦV (T1) represents the luminous flux output of the junction temperature T1; K is the temperature coefficient; ΔT = T2-T1.
In general, the K value can be determined by experiment. The K value related to InGaAlP LED is shown in Table 1: From the above table, for InGaAlP LED, the temperature coefficient is only related to the emission wavelength of the device, and whether it is transparent to the substrate Irrelevant, further experiments indicate that the shorter the emission wavelength of InGaAlP, the greater the K value. The faster the luminous flux of the device decays with increasing temperature. For InGaN series LEDs, the change of the luminous flux with temperature is much smaller than that of InGaAlP LED. Typical results are shown in Figure 2. It can be seen from the figure that as the emission wavelength becomes shorter, the change in light output flux with temperature becomes less obvious. Table 2 lists the relative values ​​of light output flux at a junction temperature of 100 ° C relative to 25 ° C.
Equation (2) points out another representation of the change in light output flux with junction temperature ΦT2 = ΦT1e- (T2-T1 / T0) (2)
Here T0 represents a characteristic temperature. The T0 value is related to the material. The experiment indicates that for red InGaAlP LEDs, T0 = 85? C; for amber InGaAlP LEDs, T0≈85? C; and for InGaN LEDs, the T0 value is about 840? C, It shows that the temperature coefficient of InGaN devices is much smaller than that of InGaAlP devices that emit red and yellow light, that is, the rate at which the luminous flux decreases with increasing temperature is much smaller than that of InGaAlP LEDs.
In general, the effect of decreasing the light output flux as the junction temperature increases is reversible, that is, when the temperature returns to the initial temperature, the light output flux will have a restorative increase. The mechanism of this effect is obviously due to the fact that some relevant parameters of the material will change with temperature, resulting in changes in device parameters. As the temperature increases, the concentration of electrons and holes will increase, the band gap will decrease, and the electron mobility will also decrease. Changes in these parameters must cause changes in the output luminous flux of the device. However, when the temperature returns to the initial state, the expression of the device parameters will also disappear, and the output luminous flux will also return to the initial state value.
Table 3 shows the test results of the high-power device AP-HLR-01. Between each measurement, a -40? C-140? C cold and hot loop aging test was performed. The measurement data points out that each measurement data can be repeated well, and the cold and hot loop aging test does not change the performance of the device, indicating that under certain conditions, the performance of the LED device changes with the current is reversible. Obviously, the decrease in light effect at high currents is caused by the temperature rise. When the test current is reduced to a small current, the light effect data returns to the original value.
2. The decay of device performance at high temperature In addition to the recoverable change of LED light output characteristics at high temperature, an irreversible permanent decay will also occur over time. Figure 3 shows the decay of the light output flux of Lumileds' model Luxeon high-power device with time. It can be seen from the figure that for the same type of LED device, at the same operating current, the higher the junction temperature, the faster the output light intensity of the device will decay. For a definite device, in general, the size of the junction temperature depends on the operating current and the ambient temperature. After the operating current is fixed, the higher the ambient temperature, the higher the junction temperature and the faster the decay rate of device performance. Conversely, when the ambient temperature is determined, the greater the operating current of the device, the higher the junction temperature and the faster the rate of device performance decay.
Figure 4 shows the decay curve of the relative value of the output luminous flux with time for a typical InGaAlP device at different operating currents. Obviously, when the operating current of the device increases, the light output characteristics of the device will decay faster.
In order to ensure the normal working conditions of an LED device, it is necessary to make the junction temperature of the device lower than a certain value Tj. For this reason, when the ambient temperature rises, the operating current should be appropriately reduced until the ambient temperature rises to the critical temperature Tj, and the operating current is reduced to zero. At this time, the junction temperature will be equal to the ambient temperature, as shown in Figure 5.
There are usually two reasons for the permanent attenuation of the output performance of LED devices under high temperature conditions. One reason is the proliferation of defects in the material. It is well known that modern high-brightness LED devices usually use MOCVD technology on heterogeneous substrates such as GaAs and sapphire. The epitaxial growth is made of InGaAlP or InGaN and other materials. In order to improve the luminous efficiency, the epitaxial materials all contain a multi-layer structure. Due to the more or less lattice mismatch between the epitaxial layers, a large number of Structural defects such as element faults, at higher temperatures, these defects will rapidly multiply and multiply until they invade the light-emitting region, forming a large number of non-radiative recombination centers, which seriously reduces the injection efficiency and luminous efficiency of the device. In addition, under high temperature conditions, micro-defects in the material and fast-spreading impurities from the interface and the electric board will also be introduced into the light-emitting region, forming a large number of deep energy levels, which will also accelerate the performance decay of the LED device.
At high temperatures, the denaturation of the LED package epoxy is another major cause of LED performance decay and even failure. Generally, there is an important characteristic of encapsulating epoxy for LED, that is, when the epoxy temperature exceeds a specific temperature Tg = 125? C, the characteristics of the encapsulating epoxy will change from a rigid glass-like state to a soft Rubbery substance. At this time, the expansion coefficient of the material increases sharply, forming an obvious inflection point, as shown in Figure 6. The temperature corresponding to this inflection point is the glass transition temperature of epoxy resin, and its value is usually 125 ° C. When the device changes near or above this temperature, significant expansion or contraction will occur, resulting in additional stress on the wafer board and leads, and excessive fatigue or even detachment damage. In addition, when the epoxy is at a higher temperature (even if it does not exceed the transition temperature Tg), especially the encapsulated epoxy adjacent to the chip will gradually denature and yellow, which affects the light transmission performance of the encapsulated epoxy. This is a subtle process. As the working time increases, the LED will gradually lose its luster. Obviously, the higher the working temperature, the faster this process will be. To solve this difficulty, especially in the manufacturing process of high-power devices, some advanced packaging structures have abandoned epoxy resin materials and used some more stable materials such as glass and PC to make lenses; another important method The epoxy is not directly contacted with the surface of the wafer, and is filled with a gel-like transparent silicone with stable performance. Practice has proved that through such improvements, the performance and stability of the device have been significantly improved.
3. The influence of junction temperature on the emission wavelength
The luminous wavelength of LEDs can be generally divided into two types: peak wavelength and dominant wavelength. The former indicates the wavelength with the largest light intensity, and the dominant wavelength can be determined by the X and Y chromaticity coordinates, reflecting the colors perceived by human eyes. Obviously, the change in the LED emission wavelength caused by the junction temperature will directly cause the human eye to feel differently about the LED emission color. For an LED device, the value of the forbidden band width of the material in the light-emitting area directly determines the wavelength or color of light emitted by the device. InGaAlP and InGaN materials belong to group III-V compound semiconductors. Their properties are similar to GaAs. When the temperature is increased, the band gap of the material will be reduced, resulting in a longer emission wavelength of the device and a red-shifted color.
The change in wavelength with junction temperature can usually be expressed as follows:
λ (T2) = λ (T1) + ΔToK (nm /? C) (3)
Among them: λ (T2) represents the wavelength at the junction temperature T2; λ (T1) represents the wavelength at the junction temperature T1; K represents the coefficient of wavelength change with temperature.
Table 4 indicates the dominant wavelength and peak wavelength K of InGaAlP and InGaN devices. From the table, it can be seen that for InGaN there are InGaAlP LEDs, the change in peak wavelength with temperature is greater than the change in dominant wavelength with temperature, among which InGaAlP LEDs are even worse.
The human eye's sensitivity to color perception at different wavelengths is very different, as shown in Figure 7: In the blue, green, and yellow regions, a small wavelength change will cause a change in the human eye's perception. Therefore, higher requirements are imposed on the temperature rise effect of blue, green and yellow devices. Generally speaking, the human eye can perceive the wavelength change of 2 ～ 5nm; while the red light wavelength change, the human eye feels relatively dull, but it can also feel the 15nm wavelength difference. To quantify the human eye's perception of colors at different wavelengths, some company's products list the relationship between color and wavelength with the color bin of the main wavelength, as shown in Table 5.
Obviously, for the amber (yellow) color, because the human eye is the most sensitive, the wavelength interval of the color bin is very fine, only 2-3nm, but for the red region, the interval is expanded to 15nm. That is to say, why the requirements for the color calibration and uniformity of the yellow traffic lights are higher, and the color requirements of the red traffic lights are relatively lower.
4. The influence of junction temperature on LED forward voltage. Forward voltage is an important parameter to judge the performance of LED. Its value depends on the characteristics of semiconductor materials, wafer size, device formation and electrode manufacturing process. Relative to the forward current of 20mA, the forward voltage of InGaAlP LEDs is generally between 1.8V and 2.2V, while the forward voltage of blue and green InGaN LEDs is between 3.0V and 3.5V. Under low current approximation, the forward voltage drop of the LED device can be expressed by equation (4):
Vf = (nKT / q) ln (If / Io) + RsIf (4)
Where Vf is the forward voltage, If is the forward current, Io is the reverse saturation current, q is the electron charge, K is the Boltzmann constant, Rs is the series resistance, and n is one that characterizes the perfection of the P / N junction The parameters are between 1-2. On the right side of the analysis formula (4), it is found that the reverse saturation current Io is closely related to the temperature, and the value of Io increases with the increase of junction temperature, resulting in a decrease in the value of the forward voltage Vf. The experiment points out that under the condition that the input current is constant, for a certain LED device, the relationship between the forward voltage drop of the two terminals and the temperature can be expressed by equation (5):
VfT = VfTo + K (T-To) (5)
In the formula, VfT and VFTo represent the forward voltage drop when the junction temperature is T and To, respectively. K is the coefficient of voltage drop with temperature. For InGaAlP and InGaN LED, the K value can be roughly shown in Table 6. Some people gave detailed experimental data, as shown in Table 7 and Table 8.
The change in voltage with temperature is recoverable, but as in the case of high temperature, due to the large proliferation and accumulation of junction defects and impurities, it will also cause an increase in the extra recombination current, and the forward voltage will drop. In general, constant current is a better mode of LED work. For example, under constant voltage conditions, the forward voltage drops and the forward current increases due to the temperature rise effect, and a vicious cycle is formed, which eventually leads to device damage.
4. Countermeasures to reduce LED temperature rise effect
The input power of the LED is the only source of the thermal effect of the device. Part of the energy becomes radiant light energy, and the rest eventually becomes heat, which raises the temperature of the device. Obviously, the main method to reduce the temperature rise effect of LED is to try to improve the electro-optical conversion efficiency (also called external quantum efficiency) of the device to convert as much input power as possible into light energy. Another important way is to improve the device The heat dissipation capacity of the heat dissipates the heat generated by the junction temperature into the surrounding environment through various channels.
1. Quantum efficiency of LED devices The so-called quantum efficiency of LED devices is the ability of the electrical energy of the device to be converted into light energy. This electro-optical conversion ability can usually be defined as the external quantum efficiency ηex, which is the injection efficiency of the device ηJ, internal The sum of quantum efficiency ηi, electron transport efficiency ηf and light output efficiency ηo.
ηex = ηJoηioηfoηo (6)
For InGaAlP and InGaN LED devices, due to a large difference between the forbidden band width Eg and the doping concentration on both sides of the PN junction, usually ηJ -1; due to the structure of the device light-emitting region, all epitaxial growth is formed, and the light-emitting region The PN junction is a sudden junction, and the electron transport efficiency is also close to 1. In addition, in view of the fact that the current InGaAlP and InGaN device structures and growth processes are very mature, practice has proved that modern technology is sufficient to increase the internal quantum efficiency to a level close to 100%. Therefore, the external quantum efficiency of the LED device mainly depends on the light extraction efficiency ηo. If the die is regarded as an optical cavity with an absorption coefficient of α and a volume of v, surrounded by N faces with an area of ​​Ai, then the die The light extraction efficiency can be expressed as:
ηN = ΣAiTI / [Σ (1-Ri) Ai + 4αv] (7)
Here, TI and Ri are the transmittance and reflectance of Ai, respectively. For an actual LED die, calculations show that the small transmittance of the wafer surface is the main reason why the light efficiency of the LED device becomes very small. The reason is due to the large refractive index difference between the two sides of the wafer surface. As shown in Figure 8, when the light in the wafer hits the surface along aspect 1 and into the air along direction 2, according to the law of refraction:
n1Sinθ1 = n2Sinθ2 (8)
Generally, the refractive index of the wafer material is n1≈3.6, and the refractive index of air is n2 = 1. It can be calculated that the critical angle at which the total reflection occurs at the interface (θ2 = 90?) Θ1 = θ0 = 16.2 ?, that is to say, only 4% of the beam from the inside of the wafer to the surface can exit the surface, and most of the remaining light energy is reflected. Back inside the wafer material and absorbed by (substrate).
2. Several ways to improve the efficiency of LED light output (1) Transparent substrate technology
InGaAlP LEDs are usually prepared by epitaxially growing InGaAlP light-emitting regions and GaP window regions on GaAs substrates. Compared with InGaAlP, GaAs material has a much smaller band gap. Therefore, when short-wavelength light enters the GaAs substrate from the light-emitting region and the window surface, it will be fully absorbed, which is the main reason for the low light extraction efficiency of the device. . A Bragg reflector region is grown between the substrate and the confinement layer, which can reflect the light perpendicularly directed to the substrate back to the light-emitting region or window, which partially improves the light-emitting characteristics of the device. A more effective method is to first remove the GaAs substrate and replace it with a fully transparent GaP crystal. The removal of the substrate absorption region in the wafer has increased the quantum efficiency from 4% to 25-30%. Three years ago, in order to further reduce the absorption of the electrode region, some people made this transparent substrate type InGaAlP device into the shape of a truncated inverted cone, so that the quantum efficiency has been greatly improved, as shown in Figure 9. Obviously, this truncated inverted cone shape device increases the light transmission area even more. In the red region, the external quantum efficiency of this type of device can exceed 50%.
Figure 10 shows the relationship between the luminous flux of various devices and the forward current, which clearly shows the difference in luminous flux of the three types of devices. For devices that absorb substrates, due to the low quantum efficiency, a large part of the input energy becomes heat. Under a small forward current, the junction temperature of the device rises very high, causing the luminous flux to drop rapidly. For LED devices with transparent substrates, because a considerable part of the input electrical energy becomes light energy, the heating effect is relatively reduced, so that the device can work in a much larger current state.
(2) Metal film reflection technology If the transparent substrate process first originated from HP, Lumileds and other companies in the United States, then the metal film reflection method has been largely researched and developed by some companies in Japan, Taiwan and other places. This process not only avoids the transparent substrate patent, but also facilitates large-scale production. The effect can be said to be similar to the transparent substrate method. This process is usually referred to as the MB process, and its basic points are shown in FIG. 11. First, the GaAs substrate is removed, then an Al metal film is vapor-deposited on the surface and the Si substrate surface at the same time, and then welded together under a certain temperature and pressure. In this way, the light irradiated from the light-emitting layer to the substrate is reflected to the wafer surface by the Al-based metal film layer, thereby improving the luminous efficiency of the device by more than 2.5 times. The experiment proves that when the current is 400mA and 800mA, the luminous flux of MB type red LED can reach 37lm and 74lm respectively. Such devices have entered small batch production in Japan's Sanken Electric, Taiwan Guolian, and brand new companies. Compared with traditional devices, the light efficiency has been greatly improved. In addition to the device with MB structure, Taiwan Guolian also developed a new generation device called GB type high brightness InGaAlP LED. The so-called GB is the abbreviation of English Giga Bright. This process uses a new type of transparent adhesive to bond the LED epitaxial wafer with GaAs absorption substrate and a sapphire substrate, and then remove the GaAs absorption substrate and make electrodes on the epitaxial layer to obtain High luminous efficiency.
(3) Surface microstructure technology Surface microstructure technology is another effective technology to improve the light output efficiency of the device. The basic point of this technology is to etch a large number of small structures with the size of light wavelength on the surface of the wafer, each of which has a truncated angle The tetrahedron shape not only expands the light output area, but also changes the direction of light refraction at the wafer surface, thereby significantly improving the light transmission efficiency. Fig. 12 indicates the N light-emitting modes of LED wafers with a textured structure. Due to the presence of textured edges, many lights that are originally larger than the critical angle can be transmitted out of the device surface through reflection or refraction at the edge. Obviously, the existence of the texture structure on the surface is equivalent to a substantial increase in the thickness of the window layer in the light emitting mechanism. The thinner the window layer and the deeper the texture erosion, the more obvious the increase in light output. Measurements indicate that for devices with a window layer thickness of 20 μm, the light extraction efficiency can increase by 30%. When the window layer thickness is reduced to 10 μm, the light extraction efficiency will be improved by 60%. For LED devices with a wavelength of 585-625nm, the luminous efficiency can reach 30lm / w after the texture structure is made, which is close to the level of transparent substrate devices.
(4) Flip chip technology usually shows the structure of blue-green and white LED as shown in Figure 13. The GaN-based LED structure layer is grown on the sapphire substrate by MOCVD technology, and the light emitted from the P / N junction light emitting region is emitted through the above P-type region. Due to the poor conductivity of P-type GaN, in order to obtain good current spreading, a metal electrode layer composed of Ni-Au needs to be formed on the surface of the P region by evaporation technology. The lead in the P region is led out through this layer of metal film. In order to obtain good current spreading, the Ni-Au metal electrode layer should not be too thin. For this reason, the luminous efficiency of the device will be greatly affected, and it is common to take into account both current expansion and light extraction efficiency. But no matter what the situation, the presence of the metal film will always make the light transmission performance worse. In addition, the presence of wire solder joints also affects the light extraction efficiency of the device.
The structure using GaN LED flip chip can fundamentally eliminate the above problems, as shown in Figure 14. Since the wafer is flip-chip mounted on the Si-based pad, the light emitted by the LED is directly emitted through the sapphire, and the above-mentioned Ni-Au metal film and lead electrode are not present, so there is no loss of the emitted light, plus vapor deposition on the underlying P-GaN layer With Ag reflective film, the intensity of the outgoing light is further enhanced. Figure 15 shows how the quantum efficiency of blue-green LEDs varies with peak wavelength. The experiment indicates that in the peak wavelength region of 450 ~ 530nm, the quantum efficiency of flip-chip power LED devices is 1.6 times higher than that of ordinary devices.
3. Analysis of LED heat dissipation mechanism As mentioned above, for a conventional LED device, more than 90% of the input power will be converted into heat. In order for the device to maintain a suitable temperature and work properly, this heat must be dissipated to the surrounding environment through a medium such as a tube substrate.
The temperature difference generated by the unit thermal power transmission between two nodes is usually defined as the thermal resistance between the two nodes, and the mathematical expression is: Rθ = ΔT / PD (9)
Where Rθ is the thermal resistance between nodes 1 and 2, ΔT is the temperature difference between nodes 1 and 2, and PD is the thermal power flow between two points. The thermal resistance Rθ represents the heat dissipation capacity between two points. The larger the Rθ, the worse the heat dissipation capacity; conversely, the smaller the Rθ, the stronger the heat dissipation capacity. When the electric power V = VFoIF is applied to the LED, a large amount of heat will be generated at the PN junction of the device, causing the wafer temperature to rise rapidly. Due to the good heat dissipation characteristics of the device, most of the heat will pass through the silver paste, the case, and the heat dissipation substrate , PCB is emitted to the surrounding environment, thereby suppressing the temperature rise of the device wafer.
Similar to the electrical resistance characteristics in electricity, the thermal resistance also has the same algorithm. When n LEDs are mounted on the same substrate, the heat flow diagram is shown in Figure 17.
In the figure, Tj, Tc and TB respectively represent the PN junction area of ​​a certain LED tube. The temperature, RθJ-C, RθC-B and RθB-A at the case and substrate represent the PN junction and case, the case and the substrate respectively 〠The thermal resistance between the substrate and the environment, then the total thermal resistance of the LED array can be expressed as: RθJ-A = RθJ-C + RθC-B + RθB-A (10)
Among them, 1 / ΣRθJ-C = Σ (1 / RθJ-C), 1 / ΣRθC-B = Σ (1 / RθC-B), the basic condition that the above formula satisfies is that all LEDs in the array have exactly the same parameters.
Thermal resistance Rθ is an important parameter of LED. When we know the Rθ value of the thermal resistance of a certain device, then according to equation (9), the junction temperature of LED can be obtained: Tj = TA + PDRθJ-A (11)
Where Tj is the PN junction temperature of the device, TA is the ambient temperature, PD = IoV is the dissipated power of the device, and RθJ-A is the thermal resistance between the PN junction of the device and the environment.
Obviously, the thermal resistance of the LED will seriously affect the use conditions and performance of the device. Figure 18 points out the relationship between the maximum forward current of the device with different thermal resistance values ​​and the ambient temperature. The forward current increases proportionally. When the thermal resistance is large, due to the rise of the PN junction temperature, when the forward current increases to a certain value, the luminous flux will tend to saturate and gradually decrease. For an LED tube, trying to reduce the thermal resistance between the PN junction and the adopted environment is the fundamental way to improve the heat dissipation capability of the device. Since epoxy glue is a low thermal conductivity material, the heat generated at the PN junction is difficult to be dissipated upwards to the environment through the transparent epoxy. Most of the heat passes through the substrate, silver paste, case, epoxy adhesive layer, PCB The heat sink spreads downward. Obviously, the thermal conductivity of related materials will directly affect the heat dissipation efficiency of the device. Table 9 and Table 10 indicate the thermal conductivity values ​​of several commonly used substrates and heat sink materials. The data of silver paste and epoxy are not listed in the table. Their thermal conductivity values ​​are 2.7 and 0.2 to 0.8 (w / mk). The experiment indicates that for a common type (Φ5) LED, the total thermal resistance from the ambient temperature of the PN junction region is between 300 and 600? C / w; for a power LED device with a good structure, the total heat resistance Resistance is about 15 ~ 30? C / w. The huge difference in thermal resistance shows that ordinary devices can only work normally under very small input power conditions, and the power dissipation of power devices can be as large as watts or even higher.
A whteboard magnet is usd for teaching,put paper on the blackboard
Custom Magnetic whiteboard , flat whiteboard magnet, neodymium whiteboard magnet
HU NAN YUBANG MAGNETIC MATERIAL CO.,LTD , https://www.ybmagnet.com