Test 4: RGB Triads
To ‘premix the colors at the source, another test setup was constructed with red, green, blue triads. Electrical consideration called for a distribution among the colors, using 10 red, 12 green and 12 blue LEDs. This will probably not result in white light with all LEDs driven to full power. However, by adjusting the duty cycle for each color, some useful information will still be obtained. the photo below shows the layout.
The front opening of the baffle shown was then closed off with a translucent sheet of Plexiglas not shown here) to diffuse the light and mix the colors. The resulting light was projected onto a white poster board some 20 inches away. By adjusting the duty cycle of each LED color (using the pulse-width controller), the overall color can be adjusted over a wide range. By driving only a single LED chain, pure monochromatic colors are obtained, of course.
However, one of the goals of these tests is to produce white light. Given the different efficiencies of the LEDs, the different response of the human eye to various colors and the subjective definition of “white” light, the results are as follows.
What I perceive as a pleasant white light, something between an incandescent and daylight, required duty cycles as follows:
Needless to say, using this ratio of LED quantities to obtain white light is not the thing to do. The test suggests new, different ratios. For example, one could use 20 red, 18 green and 6 blue LEDs, all driven to full capacity to achieve a similar result. This will be the subject of the next test.
TO BE CONTINUED ... Also see Remote Phosphor