Key defects and improvement strategies of LED lamp detection methods

Traditional methods for detecting the light, color, and electrical parameters of LEDs and their modules include electric pulse drive, CCD fast spectrum measurement, and measurement methods after thermal equilibrium under certain conditions. However, the measurement conditions and results of these methods are far from the actual working conditions of LEDs in lighting fixtures. This article introduces how to mark the Vf-TJ curve and control the measurement of the light, color, and electrical parameters of LEDs at a controlled junction temperature. This not only provides a target limit for how to ensure the working junction temperature of LEDs in lighting fixtures using LEDs, but also makes the measurement parameters of the light, color, and electrical parameters of LEDs and their modules closer to the actual application conditions. The article also introduces how to measure the junction temperature of LEDs and determine the functional relationship between the limit temperature of the LED reference point and the junction temperature in lighting fixtures using LEDs. This provides an effective way to quickly evaluate the working status and service life of lighting fixtures using LEDs.

I. Preface

For an emerging product, the development of the product itself always precedes the product standards and testing methods. Although the product standards and testing methods cannot precede the product research and development, they should keep up with the progress of product design and development as much as possible, because the process of formulating product standards and testing methods is a review, discussion and summary of the product research and development process. As long as the conditions are basically mature, the more timely the formulation of product standards and testing methods, the more it can reduce the blindness of the product research and development process. With the development of the LED lighting industry, it is basically time for us to review and summarize the standards and testing methods of LED lighting products.

Quote:
2. Current status and improvement methods of photoelectric parameters and detection methods of LED modules

1. Traditional LED module detection method

At present, there are two main methods for detecting traditional LED modules. The first method is to use pulse measurement, which is to fix the lighting LED module on the measuring device (such as the measuring position of the integrating sphere, etc.), and use the synchronous linkage of the pulse constant current power supply and the instantaneous measurement spectrometer, that is, when the LED emits a pulse current of tens of milliseconds to hundreds of milliseconds, the shutter of the instantaneous measurement spectrometer is opened synchronously to quickly detect the light parameters (luminous flux, light color parameters, etc.) emitted by the LED, and at the same time, the forward voltage drop and power parameters of the LED are also collected synchronously. Since the junction temperature of the LED is almost equal to the room temperature during the detection process in this way, the light efficiency of the measurement result is high, and the light color and electrical parameters are significantly different from the actual use. This is generally a rapid detection method adopted by LED chip (device) manufacturers, and it is not comparable to the actual application of the LED in the final lighting fixture.

The second detection method is to install the LED module on the detection device, and then attach a fixed heat sink (which may also have a base temperature control function) to apply the claimed working current to the LED. Influenced by the traditional lighting source detection method, the photoelectric parameters of the LED are measured only after the LED reaches thermal equilibrium. This method seems to be more rigorous, but in fact, its thermal equilibrium conditions and working conditions are still not well correlated with the state of such LEDs installed in the final lighting fixture. Therefore, the measured photoelectric parameters are still not comparable with the parameters of the actual application state in the future. The GB/T24824-2009/CIE 127-2007NEQ "Measurement Methods for Basic Performance of LED Modules for General Lighting" standard, which has been promulgated, stipulates in this regard: "When testing or measuring, the LED module should work in a thermal equilibrium state. While monitoring the ambient temperature, it is best to monitor the operating temperature of the LED module itself to ensure the reproducibility of the test. If it is possible to monitor the junction voltage of the LED module, the junction voltage should be monitored first. Otherwise, the temperature of the specified temperature measurement point of the LED module should be monitored." It can be seen that measuring the photoelectric parameters of LED modules under the condition of monitoring the junction voltage is the preferred solution to ensure the reproducibility of detection. However, the standard does not specify the detection of the light, color and electrical parameters of LED modules under the conditions of simulating the actual junction temperature.

2. Improvement of LED module measurement method

As we all know, the optical and electrical parameter characteristics of LED are closely related to its working junction temperature. For the same LED product, the junction temperature will not cause obvious differences in these parameters, which also causes obvious inconsistency in the measurement results of the optical, color and electrical parameters of the same LED. Therefore, the measurement of the optoelectronic parameters of LED should first be considered under the conditions of the set working junction temperature. In addition, due to differences in packaging processes and materials, the claimed maximum working junction temperature of LED is obviously different. In order to ensure that LED lighting products have the characteristics of high efficiency and longevity, the actual working junction temperature of LED should be significantly lower than the maximum working junction temperature. For example, the LED packaging methods and technologies that we currently use in large quantities have a covering layer of polymer silica gel plus phosphor in front of the LED's light emission. Practice has proved that in order to make such LED lighting fixtures, the time to maintain the luminous flux of 70% must be ≥60,000 hours, and its working junction temperature must be kept below 70℃～75℃. From the perspective of improving light efficiency and service life, it is better to keep the working junction temperature of LED below 60℃, but from the perspective of the shape, volume and cost performance of lighting equipment, it is most appropriate to control the maximum working junction temperature of LED to 70℃～75℃ on the basis of achieving the expected light efficiency and service life. In order to make the detection of light, color and electrical parameters of LED and its modules as close as possible to the actual application junction temperature state, it is necessary to solve the problem of how to measure the junction temperature of LED and detect the light, color and electrical parameters at this junction temperature.

(1) Currently, there are roughly two methods for measuring LED junction temperature:

1) The junction temperature is obtained by measuring the pin temperature, chip power dissipation and thermal resistance coefficient. However, due to the inaccuracy of power dissipation and thermal resistance coefficient, the measurement accuracy is relatively low.

2) Infrared thermal imaging method, using an infrared non-contact thermometer to directly measure the temperature of the LED chip, but it requires the device under test to be in an unpackaged state. In addition, there are special requirements for the refractive index of the LED packaging material. Otherwise, it cannot be accurately measured and the measurement accuracy is relatively low.

3) Determining junction temperature by using the peak shift of the luminescence spectrum is also a non-contact measurement method. It directly determines the junction temperature by using the bandgap width shift technology from the luminescence spectrum. This method requires high resolution accuracy of the spectral testing instrument, and the accuracy of the luminescence peak position is difficult to measure. A 1-nanometer error change in the spectral peak shift corresponds to a change of about 30 degrees in the measured junction temperature, so the measurement accuracy and repeatability are relatively low.

4) Nematic liquid crystal thermal imaging technology has high requirements for instrument resolution and can only measure unpackaged single bare chips, but cannot measure packaged LEDs.

5) Use the Vf-TJ relationship curve of the diode PN junction voltage and junction temperature to measure the junction temperature of the LED.

From the various LED junction temperature measurement methods introduced above, it can be seen that the method of using the change of the diode PN junction voltage to monitor the junction temperature is the most feasible and has the highest measurement accuracy. Therefore, in many integrated IC circuits, in order to detect the working junction temperature of the IC chip, one or several diodes are often engraved or inserted, and the change of their forward voltage drop is measured to achieve the purpose of measuring the chip junction temperature.

(2) The most advanced Vf-TJ measurement method in the world

At present, the internationally advanced Vf-TJ measurement method is to connect the LED to be measured with the lead wire and place it in a silicone oil cylinder, then heat the silicone oil cylinder to make the temperature of the silicone oil reach about 140℃, and then let the silicone oil in the cylinder cool naturally. As long as the temperature of the silicone oil drops slowly enough during cooling, it can be considered that the junction temperature of the LED is basically the same as the temperature of the LED's heat sink. In this process, according to the measured silicone oil temperature, a specified current pulse is instantaneously input to the LED every time it drops by 2℃ to 10℃, and its forward voltage drop at this temperature is measured, and the temperature and forward voltage drop of this measurement point are imported into the database of the computer software. Starting from about 140℃, as the temperature drops, the heat sink temperature and forward voltage drop are measured every time a set equal temperature drops, and the measurement is continued to about 25℃. When this set of measurement data is completed and imported into the database of the computer software, a Vf-TJ curve is generated by the software. This method belongs to the measurement method when the temperature drops, which is feasible for measurement. However, because the ambient temperature of the laboratory is constant (generally 25°C), and the oil temperature of the silicone oil cylinder drops from high to low, when the oil temperature of the silicone oil cylinder is high, the temperature difference with the ambient temperature of the laboratory is large, so the cooling speed is fast. In order to ensure the accuracy of the measurement, appropriate measures are taken to prevent the temperature of the silicone oil cylinder from dropping too fast when the temperature is high. However, when the temperature of the silicone oil cylinder is low, the cooling speed is too slow because the temperature difference with the room temperature is too small, which greatly prolongs the measurement time of this detection process. For the above reasons, this temperature drop measurement method cannot be short in the Vf-TJ calibration process (about 4 to 5 hours), otherwise it will produce obvious measurement errors. In addition, the oil cylinder of this detection device is fixed, and it takes a long time to measure the second group. In addition, the above-mentioned heating device is at the bottom outside the silicone oil cylinder, and there is a significant lag in heating, temperature control and measured temperature, which also causes the poor accuracy of this method in measuring junction temperature.

(3) New Vf-TJ detection method

The detection method invented by this mechanism adopts the measurement method when the temperature rises, and adopts the PID (integral, differential plus heating and non-heating time ratio control) method set by the computer to heat and control the temperature of the silicone oil cylinder, that is, in the initial stage of heating the silicone oil cylinder, the ratio of heating time to non-heating time is very small and adjustable, so that the temperature rise rate of the silicone oil cylinder can ensure the consistency of the LED junction temperature, heat sink and silicone oil temperature. As the silicone oil temperature gradually rises, the temperature difference with the room temperature also increases. At this time, the PID heating and temperature control system will automatically increase the ratio of heating time to non-heating time (actually increasing the heating power per unit time), so it can ensure that the temperature rise rate of the silicone oil in the silicone oil cylinder is always maintained at the set rate, and the oil temperature will not rise at different rates due to the difference between the silicone oil temperature and the ambient temperature. The silicone oil can be set to keep the temperature constant at any temperature value in the application temperature range, and the heating rate of 0.1℃/minute to 2℃/minute can also be achieved.

After each heating stage, there is a temperature control stage, that is, the heating stage and the temperature control stage form a step-type temperature control curve. As the temperature rises in a step-by-step manner, the forward voltage can be set to measure once every 0.5℃ rise, and can be gradually adjusted to measure once every 10℃ rise at an interval of 0.5℃. In order to ensure the timeliness of temperature control and temperature measurement, built-in heating is adopted. In addition, in order to ensure the consistency of oil temperature in the silicone oil cylinder, a magnetic induction stirring bar is added to the bottom of the cylinder. The stirring bar is driven by an external motor to rotate in the cylinder through magnetic induction. The rotation speed is adjustable, thereby ensuring that the temperature difference of the silicone oil in the cylinder is kept within 0.2℃. Because the temperature rise rate of the silicone oil is almost the same, and the step-by-step heating and temperature control are implemented, the measuring device can ensure accurate detection results under the condition of a reasonable temperature rise rate, and the detection time (about 2.5 hours from 25℃ to 140℃) can be significantly lower than the measurement time of the existing detection devices in the world. The existing testing devices in the world are single silicon oil cylinder structures. This measuring device adopts double silicon oil cylinder structure. After completing the measurement of one set of samples, a silicon oil cylinder can be replaced to immediately start the second set of LED testing. This measuring device measures the temperature and LED forward voltage drop at each measuring temperature point, imports them into the database and generates the Vf-TJ curve by the compiled software.

(4) Measurement of junction temperature of lighting LEDs and use of Vf-TJ relationship curve to guide the measurement of light, color and electrical parameters

After obtaining the Vf-TJ curve of the LED under test, the most important thing is to measure the light, color and electrical parameters under the condition of fixed junction temperature. The detection system is shown in Figure 1. Fix the LED under test in an integrating sphere with a temperature control/constant temperature base, pass the working current to the LED, and let the LED burn for 15 to 20 minutes to basically reach stability. Then quickly switch to the measurement current (that is, the measurement current of the previously calibrated Vf-TJ curve) to quickly measure the forward voltage Vf of the LED under test in a few milliseconds. By comparing it with the Vf corresponding to the set junction temperature value in the Vf-TJ curve, if there is a difference with the target value, the control program will automatically adjust the temperature of the constant temperature base to make the forward voltage Vf of the LED reach the junction voltage corresponding to the target junction temperature value. After quickly measuring Vf, the device will automatically return to the state of passing the working current to the LED. When the working current passes through the LED under test, and its junction temperature reaches the target value (that is, the Vf value corresponding to the target junction temperature value) and reaches thermal equilibrium, the system will automatically start the spectrometer to measure the light and color parameters and read its electrical parameters at the same time.

The most obvious advantage of the above measurement method is that in the actual application of LED, as long as the LED in the lighting fixture operates near the target junction temperature value, the parameters of this method have good simulation, which makes the measured parameters meaningful, and its light, color and electrical parameters also have good reproducibility of measurement results.

Figure 1 LED junction temperature measurement and light, color, and electrical parameter measurement system using the Vf-TJ curve under set junction temperature conditions.

3. Measurement of junction temperature after LED enters lighting fixture

1. Necessity of controlling and measuring junction temperature after LED enters lighting fixtures

When LEDs are used in lighting fixtures, people generally hope that they have a service life of tens of thousands of hours. However, to measure the light attenuation and life of lighting fixtures using LEDs, it often takes more than 300 days (6000 hours) according to the LM80 requirements of the US DOE. This is impossible to implement during the bidding and acceptance of many projects.

As an important parameter to measure the performance of an LED lighting fixture, junction temperature is the core element of reliability measurement of LED lighting fixtures in engineering applications. If the two quantitative indicators of the PN junction temperature of the LED in the lamp and the thermal resistance from the PN junction to a specified point of the heat sink can be accurately measured, it is not only possible to measure the heat dissipation characteristics of the lighting fixture using LEDs, but also to qualitatively know the approximate service life of various similar lighting fixtures using LEDs. In addition, it is also possible to know under what junction temperature conditions the measured values ​​of the light efficiency and other light parameters of the LED lighting fixture are measured, and to derive the functional relationship between a certain point (reference temperature point) on the heat sink of the power LED in the lighting fixture and the junction temperature, thereby guiding enterprises to correctly mark the temperature limit of the heat sink reference point.

2. Introduction to measurement methods

At present, the junction temperature of the PN junction of LEDs at home and abroad can only be measured for the junction temperature and thermal resistance of a single LED or a single LED module. There is no complete method for measuring the actual working junction temperature and thermal resistance of LEDs in lighting fixtures. The following introduces a complete method for measuring the actual working junction temperature and thermal resistance of LEDs in lighting fixtures.

1) Vf-TJ curve calibration

(1) Take an LED in the middle of a series LED group in the middle of the LED matrix in the lighting fixture as the LED to be tested, connect it according to the circuit in Figure 2, and stick a thermocouple on the heat sink (small heat sink of the LED itself) of this LED. Place the lamp in an environment of 25℃±2℃ for 6 to 12 hours (the placement time is determined by the size of the lamp to be tested), and then pass a measurement current If through the LED to be tested in Figure 2. If can be selected in the range of 2mA to 50mA depending on the power of the LED to be tested. The power-on measurement time is 0.005S to 2S. During this period, the forward voltage drop Vf of the LED to be tested is continuously measured to obtain the curve shown in Figure 3. From this curve, the value ΔVf of the Vf drop of the LED to be tested in the lighting fixture in a unit measurement time Δt when a constant measurement current is passed can be obtained. This value is reserved as the correction amount for the change of Vf caused by the measurement current in the following detection process. When the measurement time is less than 3ms and the measured current is relatively small, no correction is required.

Figure 2. Measurement circuit connection diagram of a LED string in an LED matrix of an LED lamp

Figure 3. Relationship between the drop in Vf of the LED under test ΔVf and the measurement time Δt when the current is measured within a unit measurement time.

(2) Adjust the three 2-pole 2-throw switching relays to the measuring position, and place the LED lamp in a programmable special heating box. The heating box adopts PID programming and sets the step-by-step heating method to heat the LED lamp in the box. The temperature control curve of the step-by-step heating is shown in Figure 4. Each step in Figure 4 is divided into a constant temperature time period and a heating time period. These two time periods can be set separately, and the setting range is any value from 1 minute to 30 minutes. According to the temperature value reflected by the thermocouple attached to the heat sink of the LED, and finally the forward voltage drop of the measured LED measured by the circuit of Figure 2 is stable, it means that the LED in the lamp has reached a thermal equilibrium at a certain set point temperature. When each constant temperature time period is about to end, start measuring the forward voltage drop Vf of the measured LED. According to the actual measured time △t, the correction is △Vf from Figure 3. Add the measured Vf value to △Vf to get Vf1' of D1 at this temperature which is not affected by the measured current, that is, Vf1'=Vf1+△Vf. Import this Vf1' and the temperature T1 measured by the thermoelectric corner into the set computer database. Repeat this step to get a set of corrected values. Automatically import this set of corrected values ​​into the database to generate the Vf-TJ curve of the LED in the lighting fixture.

Figure 3 Temperature control curve of the heating box with step-by-step heating

2) Measurement of thermal resistance of LEDs in lighting fixtures

Take out the lighting fixture that has completed the Vf-TJ relationship curve calibration in the heating box and cool it down, and then measure the LED thermal resistance according to the following steps.

(1) Place the lighting fixture in the windshield specified in Appendix D of GB 7000.1, and arrange the lighting fixture in the normal thermal test position. In addition to the thermoelectric corners that have been bonded to the LED D1 under test, thermoelectric corners (single or multiple thermoelectric corners) can also be bonded to certain designated points of the heat sink of the LED in the lighting fixture or even certain points on the lighting fixture housing according to the requirements of the test client. Connect each thermoelectric corner to a temperature meter and place the lighting fixture at 25℃±1℃ for 8 hours.

(2) According to the measured working current value output by the LED control device in the lighting fixture to D1, set the test constant current power supply, and pass a measured working current to D1 according to the circuit in Figure 2. Heat for 1 minute to 30 minutes. During this period, measure the Vf of D1 once every 1 minute with the original calibrated measurement current, and find out the corresponding junction temperature value according to the Vf-TJ curve. At the same time, monitor the measured temperature of the thermoelectric corner, and automatically import the measured junction temperature value and the measured temperature value of the monitored thermoelectric corner into the database. When the junction temperature obtained by the measured Vf reaches the maximum difference with the temperature measured by the thermoelectric corner, record the VfR value at this time and the temperature value TB of a certain point measured by the thermoelectric corner. Pass the VfR value through the Vf-TJ curve to obtain the instantaneous junction temperature value TfR of D1. Calculate the thermal resistance value from the PN junction of D1 to the heat sink or radiator or even the shell according to the thermal resistance RAB=(TfR-TB)/P formula.

Where:

TFR is the junction temperature of the LED at that moment obtained by looking up the forward voltage drop Vfa value of D1 based on the Vf-TJ curve when the difference between the PN junction temperature of D1 and the measured value of the thermoelectric corner reaches the maximum value.

TB——is the measurement value of the reference point measured by the thermoelectric corner at the moment when the junction temperature obtained by the measured Vf and the temperature measured by the thermoelectric corner reach the maximum difference (the reference point can be a heat sink, a point on the radiator, or a point on the radiator of the lamp housing).

P is the heating power when measuring the thermal resistance of the LED under test, which is the product of the measured working current and the average forward voltage drop of the LED under test during the junction temperature measurement process.

3) Measurement of LED junction temperature in lighting fixtures

Take the LED lighting fixture out of the dedicated heating box. This test can be carried out simultaneously with the thermal test of the lighting fixture. Put the LED lighting fixture in the windshield specified in Appendix D of GB 7000.1, and keep the lighting fixture in the normal working position. Adjust the three 2-pole 2-throw changeover relays to the working position, and perform a thermal test according to the requirements of 12.4 thermal test in GB7000.1 standard. Light up the LED matrix in the lighting fixture through the LED control device in the lighting fixture. At this time, the LED lighting fixture is in normal working condition. Observe the temperature value reflected by the thermocouple attached to the heat sink of the LED. When the temperature value reaches thermal equilibrium (temperature change is less than 1°C per hour), adjust the three 2-pole 2-throw changeover relays to the measuring position. Measure the forward voltage values ​​of the five tested LEDs five times in a row, each time with an interval of tens of milliseconds. Calculate the forward voltage drop of the tested LED at the moment of disconnecting the working current through a computer and special function calculation software, and find out the junction temperature value of the tested LED in the LED lighting fixture when it works continuously to thermal equilibrium based on the relationship curve between the forward voltage drop and the junction temperature. At the same time, the temperature value of the reference point on the heat sink when the lamp works continuously to thermal equilibrium can also be obtained.

IV. Review and Conclusion

The measurement and control of LED junction temperature is an important and indispensable step for LED to enter the lighting field. It organically combines LED devices with the front and back processes of LED lighting fixtures. By calibrating the Vf-TJ curve of a certain model of LED, and using this curve to guide and control the measurement of light, color, and electrical parameters of LED at a predetermined junction temperature, the measured values ​​of these parameters of LED are closer to the parameters of the actual application state. In addition, the determination of the predetermined junction temperature of LED also indicates the limit of heat dissipation control for LED lighting fixture designers. Similarly, by calibrating the Vf-TJ curve of the LED in the lighting fixture, the LED junction temperature of the lighting fixture under the rated ta conditions can be measured. This can not only objectively evaluate the rationality of the heat dissipation design of the LED fixture, but also reveal the functional relationship between the reference point temperature on the LED heat sink and the junction temperature, and further know the thermal resistance from the LED's PN junction to a certain point on the lighting fixture, thereby guiding the manufacturers of LED lighting fixtures to correctly mark the limit of the reference point temperature, and in mass production, can conveniently measure the reference point temperature to basically know the working junction temperature of the LED.

When LED lighting fixtures are working normally, the quality of their heat dissipation characteristics is directly related to the light efficiency, light decay and service life. The corresponding indicators are the PN junction temperature and heat resistance of the LED when it is working. If these two indicators are good, it means that the efficiency and service life of the lamp are guaranteed, just like the examination of the human body. If the blood test indicators, color CT examination and blood angiography results are all good, the person must be healthy. The significance of this inspection method is to establish the "blood test and color CT examination and blood angiography instrument" and its method for LED lighting fixtures. It can be foreseen that the establishment of this method will be a powerful driving force to guide the improvement of the design and manufacturing links of LED lighting fixtures and to promote the design and production technology of LED lighting fixtures to a higher level.