LED street light failure

LED street light failure calculation

N-hour light decay = 1 - (N-hour flux / 0 hours of luminous flux) Edit this paragraph causes LED street light failure

LED street light decay products is light in the transmission of the signal weakened, and at this stage of the global LED manufacturers have made light LED products with different degree of decline, there are also high-power LED light failure, which is directly related to temperature and, mainly by the chip, phosphor and packaging technologies decision. At present, the white LED market, its light failure may be the first to enter the civilian lighting problems. Decline for the LED light that there are two major factors:

A, LED product quality problems in itself:

1, using the LED chip body is not good, light fading rapidly. 2, the production process defects, LED chip cooling is not a good foot from the PIN to export, leading to the chip LED chip temperature is too high attenuation intensified.

Second, the use conditions of:

1, LED constant current driver, some of the reasons for the LED voltage-driven LED attenuation over. 2, the drive current is greater than the rated drive conditions. In fact, lead to bad light LED products, many reasons, the most important thing is the hot issue, although many manufacturers are not special focus in the secondary cooling products, but long-term use of these sub-LED product, the optical decay heat level of attention than has The LED products to high. The thermal resistance of LED chip itself, the impact of silver plastic, the substrate cooling effect, and the gold colloid and also has a relationship with the bad light. LED street light failure analysis

Most of the white LED is a blue LED light $ phosphor and get. LED light failure caused by two main reasons, one is light blue LED itself decline, decline than the blue LED light red, yellow, green LED faster. There is also a bad light is a phosphor, the phosphor decay at high temperatures is very serious. Its various brands of LED light decline is different. LED manufacturers are usually able to give a standard curve of light decline. Cree's light like the United States to decline curve shown in Figure 1.

Figure 1 Cree's LED light decay curve can be seen from the figure, LED light failure and its junction temperature is related to the so-called PN junction semiconductor junction temperature is the temperature, the higher the junction temperature light failure occurred earlier, but also is the shorter life expectancy. It can be seen from the chart, if the junction temperature of 105 degrees, brightness to 70% of the life expectancy of over ten thousand hours, 95 degrees there is 20,000 hours, while the junction temperature cools to 75 degrees, the life there are 5 million hours of 65 degrees can also be extended to 90,000 hours. So the key to longevity is to reduce the junction temperature. However, these data only for Cree's LED. Not suitable for other companies LED. For example, the company's LuxeonK2 Lumiled light decay curve is shown in Figure 2.

Figure 2 Lumiled company LuxeonK2 light decay curve when the junction temperature increase from 115 ℃ to 135 ℃, it will cause reduced life expectancy from 50,000 hours to 20,000 hours. Other bad light curve of each company should be able to obtain a copy of the original. How can extend the life of LED

Can be concluded from the figure, to extend its life, the key is to reduce the junction temperature. The key to reducing the junction temperature is to have a good heat sink. LED produced a timely manner to distribute the heat out. Here we are not prepared to discuss how to design a heat sink, but rather to discuss a radiator cooling effect which is relatively good question. In fact, this is a junction temperature of the measurement problem, if we are able to measure any kind of heat sink the junction temperature can be achieved, then not only compare the radiator cooling effect, but also know that after using this radiator LED life can achieve.

How to measure the junction temperature

Junction temperature measurement appears to be a temperature problem, but to measure the junction temperature of the LED's internal, we can not get into a PN junction thermometer or thermocouple to measure its temperature. Of course it can still shell temperature measurement using thermocouples, and then gives the thermal resistance Rjc (junction to case), it can calculate the junction temperature. However, after the radiator is installed, the problem has become complicated again. Because usually LED is soldered to the aluminum plate, and aluminum plate and installed the radiator, if the only measure the temperature of the radiator shell, then to calculate the junction temperature must know a lot of resistance values. Including Rjc (junction to case), Rcm (shell to the aluminum plate, in fact, which should also include the printed version of the thermal resistance of thin films), Rms (aluminum plate to the radiator), Rsa (radiator to the air), which as long as a data inaccuracy will affect the accuracy of the test. Figure 3 shows the thermal resistance of each LED to the radiator diagram. Which combined a lot of resistance, making it a more limited accuracy. That is, the heat sink from the surface temperature measured to infer more accurate junction temperature difference.

Figure 3 LED to the heat sink thermal resistance diagram Fortunately, each an indirect method of measuring temperature, it is to measure voltage. Then the junction temperature and voltage on which it? How this relationship is like? We should first talk in LED voltage characteristic.

LED IV characteristics of the temperature coefficient

We know that LED is a semiconductor diode, diode, like it and all have a volt-ampere characteristic, but also, and, like all of the semiconductor diode, the IV characteristics have a temperature. Its characteristics is that when the temperature is rising, and voltammetric characteristics of the left. Figure 4 LED's voltage characteristic plotted the temperature characteristics.

Figure 4 LED voltage characteristics of the temperature characteristics of the LED is assumed to supply constant current Io, the junction temperature of T1, the voltage V1, and when the junction temperature rise for the T2, the voltage characteristic left, the current Io is not change, voltage becomes V2. Is the temperature difference between the two voltage removed, the temperature coefficient can be obtained in order to mV / oC said. For ordinary silicon diode, the temperature coefficient is about -2mV/oC. However, most LED is not made of silicon materials, it should also go to the temperature coefficient of determination. Fortunately, various LED manufacturers data sheet gives most of its temperature coefficient. For example, Cree's XLamp7090XR-E high-power LED, the temperature coefficient of -4mV/oC. Than ordinary silicon diode 2 times larger. In the United States Puri array LED (BXRA) to give a more detailed data.

However, they are given the data, its scope is too large, so that the value of lost use. Anyway, just know that the temperature coefficient of LED can easily measure the forward voltage LED calculate the junction temperature of the LED.

How the LED junction temperature measurement

Now on to Cree's XLamp7090XR-E as an example. To illustrate how the LED junction temperature measurement. LED installation requirements have been put to the radiators, and is a constant current drive for power. At the same time connected to the LED should go out two lines cited. Previously put in the power connected to the output voltage meter (LED's positive and negative), then connect the power, not heat up while the LED before and immediately read the voltmeter reading, which is equivalent to the value of V1, and then wait at least 1 hour, so it has reached thermal equilibrium, and then measured again, LED voltage across the equivalent of V2. Subtract these two values, the difference obtained. Longer be removed 4mV look to a junction temperature can be drawn. In fact, LED mostly in series and then parallel to many, that does not matter, when the voltage difference between the LED in series by many of the common contribution, so should the voltage difference divided by the LED in series divided by the number of go 4mV, you can get the junction temperature. For example, LED strings 2 and 10, the first measured voltage is 33V, the second time after the heat balance of the measured voltage is 30V, the voltage difference of 3V. This number divided by the first LED series number (10), are 0.3V, divided by 4mV, can be 75 degrees. Assume that the environment before starting the temperature is 20 degrees, this time the junction temperature should be 95 degrees. Results obtained using this method temperature, certainly better than the heat sink temperature measurement using thermocouples again derive the junction temperature must be accurate lot. To infer from the life of the junction temperature should be as simple as check the curve in Figure 1, you can know the junction temperature of 95 degrees corresponds to life when you can get LED's life is 20,000 hours of. However, this method is used indoor LED lighting still has some credibility, if applied to the outdoor LED lighting, especially in high-power LED street light, where there are many uncertain factors. The biggest problem is the LED lights in the radiator cooling efficiency decreases with time. This is due to dust, bird feces and makes the accumulation of heat efficiency. And because there are very strong UV outdoors, but also will reduce the life of LED. UV mainly to the aging of epoxy encapsulated play a big role, if the use of silica gel, could be improved. UV aging of phosphors also have some bad effect, but not very serious. However, this method is used to compare the relative effects of the two radiator cooling is more effective. Obviously, the smaller the voltage characteristic left of the radiator, the heat, the better. .