As we are familiar with the laptop backlight, it is a combination of many LED beads, and the problem of color difference is becoming more and more obvious. If the use of LED beads is not the same batch, it is easy to lead to different wavelengths, color differences and other problems, this problem is inevitable even if the use of the same batch of LED beads. For the testing of LED chips, the integrating sphere test system is generally used.

    The use of integrating spheres to measure parameters such as luminous flux can make the measurement results more reliable. Integrating spheres reduce and remove errors caused by differences in the shape of the light, angle of divergence, etc. Such high precision and careful measurements are also driving LED integrating sphere manufacturers to increasingly choose higher performance power supplies configured in the system to power the LEDs to be tested.

LED Integrating Sphere Test System

  LED integrating sphere is mainly used to measure the luminous efficiency, color rendering and color difference of LED lights, and other parameters that characterize the performance of LED light sources. In the production line every day in a large number of tests, always facing two problems: First, how to improve the test efficiency Second, how to better protect the LED to be tested in the measurement.

It is understood that the current power supply industry, from the power supply on to the LED lamp power supply to the software control system to collect back stable voltage and current values, the total time is about 600ms, so on this basis is there a breakthrough possibility? The answer is yes. So what can be achieved by what method? In fact, the system is configured for the power supply to the LED to be tested * plays a fairly important role.

I. Configuration ** power supply can suppress overshoot and improve test efficiency

   General power supply in the moment of power on, due to the existence of the loop response time, the power supply output is not to reach the set output voltage value that stops the output. Therefore, during this period of time, the voltage will overshoot. Too high an overshoot voltage will have an impact on LED life, and in serious cases even breakdown LED.

For low-power LED lamps, the constant-current operating current is usually between a few tens to a few hundred milliamps. When working at low currents, the current and voltage creep rate decreases accordingly, sometimes the LED will light up the time, even longer than the time required to add up the voltage and current values from the normal power supply to the LED lamp to the software control system to collect back the stable voltage and current values.

The face of such as the above two problems, if in order to speed up the climbing speed, improve test efficiency, to improve the loop speed improvement measures, it usually brings more serious overshoot. However, if the overall test efficiency is reduced in order to take care of the overshoot problem, the overall test efficiency will be reduced again. Such a dilemma can be achieved by configuring **testing LED power supply, such as the IT6874A power supply* of Airtex can solve this problem head-on. The following two graphs are the waveforms of the voltage rise phase when the IT6874A DC power supply supplies 80V/20mA and 80V/400mA to the LEDs, respectively.

Figure 1: LED (80V/20mA) IT6874A climbing time: about 80ms

Figure 2: LED (80V/400mA) IT6874A creep time: approx. 50ms

From the above figure, we can see that the IT6874A DC power supply can basically meet the test requirements of high speed and no overshoot within 100ms time, while the general power supply needs about 600ms time, the IT6874A DC power supply of Airtex increases this figure by nearly 6 times, which is of great significance for the improvement of test efficiency.

I. Configure** power supply to make the integrating sphere test system to collect more stable values

When using integrating sphere testing, the stability of the power supply in the system is also very important. Stability is mainly reflected in.

(1) the error range of the power supply

    Integrating sphere test, if the use of power supply ** the higher the degree, the more the measured data tend to the true value of the batch test, the deviation between the LEDs to be tested also tend to **. For example, to carry out the test of LED lights, assuming that the power supply back to read the value of the allowable error range is 20mV, when measuring two LED lights (theoretical rating indicators are 350mA), #1-LED measurement value may be 330mA, #2-LED measurement value may be 370mA. assume that the true value of these two LED lights are 352mA, it can be seen that the 330mA and 370mA value is far from the real value, and also lead to two LED lights performance indicators deviation. If you use a ** higher degree of power supply, assuming that the voltage readback value of the allowable error range is 10mV power supply to test, the same two LED lights may be tested to get the value of 342mV and 362mV, the value is closer, can more truly reflect the performance indicators of individual LED lights and the range of deviations between the consistency of multiple LEDs.

(2) Stability of the collected values

   What does the stability of the acquisition value mean? May be to give an example. For example, the power supply can light up the LED lamp in 400mS time (the parameters of the LED lamp to be tested is 80V/20mA), at this time the software part of the integrating sphere test system to collect the voltage value of 80.099V. And when the LED lamp is turned off and on again, it is possible that the software reads the value of 82.152V. So simply use the oscilloscope, we can only see the waveform rise to the rated So simply using an oscilloscope, we can only see the waveform rise to the rated current and voltage value, but there is no guarantee that the value captured by the software also reaches a stable interval. This can affect the performance of LEDs in mass production testing due to the instability of the power supply values. In general, software engineering developers will delay to 500-600ms in order to ensure that the value is stable during continuous reading operation. Therefore, the time required for integrating sphere testing is about 600ms, instead of 400ms as measured by the oscilloscope.