While white LEDs have advantages over conventional light sources, they require an active control system to ensure that the desired lighting conditions remain predictable and stable over time. Unfortunately, today’s control systems often make use of characteristic curves for brightness and color temperature that have been developed under optimal thermal conditions. While these curves can allow a lighting system to react quickly to changing conditions, they may lead to inaccuracy because of the lack of actively monitored optical values and have no way of compensating for long term effects.
A control system should be able to react to long term effects such as loss of brightness of an LED and color drift due to the aging process of the converter mass. A system should also be able to react to short term effects such as the change of optical parameters over temperature, the failure of individual LEDs in the LED cluster, and changes to the operating point (brightness EV,
color temperature CCT) set by the end user. Such an active control system can now be realized with just one optical sensor.
To implement an active control system, an RGBW sensor can be used. RGBW sensors feature four photodiodes with bandpass filters for the red, green, blue, and white spectral components and contain an amplifier, analog and digital circuitry, and an I2C interface. The VEML6040 sensor from Vishay is an example of such a sensor.
An active feedback control system for mixed lighting using warm white and cool white LED clusters using an RGBW sensor is shown in Figure 2. The system is microcontroller-based and includes constant-current power supplies with an analog control input (0V to 10V) or a digital control input (Dali, DMX). The brightness, EV, and color temperature, CCT, are calculated from the measured RGBW sensor output using mapping matrices that are determined once for the lighting system and must be stored in an EEPROM.
The optional storage of a family of curves (CCT, EV as a function of the LED currents) is useful to set the initial operating point quickly when the user changes the brightness and color temperature. From this, the control loop moves to the final operating point according to the target values.
To verify the control circuit and the performance of the RGBW sensor, a lighting system was set up using two LED clusters (12W, Ra >90, max. 350mA) in warm white (CCT = 3050K) and cool white (CCT = 6700K). In addition to the RGBW sensor, the optical parameters were measured and checked using professional photometry and a spectral photometer.
First, the RGBW values of the sensor were determined for the light scenarios of 100% warm white, 100% cool white, and 50% warm white/50% cool white. The tristimulus XYZ values were then measured using a spectral photometer, and used to define the transfer matrix R,G,B -> X,Y,Z. With the help of this matrix, the actual optical quantities of brightness and color temperature were computed in the control system. The new currents for the warm white and cool white LED clusters were then calculated and passed to the power supplies thus closing the control loop.
As a measure of the quality of the control system, the error between the target value and the actual value of the color temperature and brightness was analyzed. To visualize the measured color temperatures, the relevant excerpt of the x/y color space from CIE1931 for the mixed area of the LED configuration is reproduced in Figure 3. In addition to the three calibrated values shown in blue, six mixed scenarios were entered shown in red. The color temperature of these scenarios was determined using the RGBW sensor. The percent error entered in each case is the difference between the RGBW sensor and a professional measurement reference.
The accuracy of the entire control system is decisively affected by the accuracy of the VEML6040 RGBW sensor and the control
algorithm. A measurement inaccuracy of CTT <2.5% over the reachable mixed color space of the white LED clusters used can be set with the RGBW sensor. For the complete control system with the VEML6040 RGBW sensor, a control accuracy of better than 4% for CCT and better than 5% for the brightness EV can be achieved.
It works. The measurements show that an active control system of the two most important optical parameters can be implemented simply with an RGBW sensor, maintaining performance that is stable over the long term and that reacts precisely to a change in the surroundings.