We are using RGB LEDs in our product and we require to emit white light to indicate a certain state. But because of the tolerance of the resistors and the LED itself it is hard to achieve the same white color in all the devices(some are reddish white, some are yellowish). We are driving the RGB leds using PWM with 255 duty cycle.

In order to solve this we decided to use a color sensor to calibrate each RGB led on the product using a color sensor(We are using http://wiki.seeed.cc/Grove-I2C_Color_Sensor/) . What we would like to do is to store a calibration value that corrects R, G and B (e.g. by some multiplication factor) so that mixing of colors can be achieved accurately for not just white but any color. But unfortunately color sensor channels(RGB) data and its correlation to PWM value is not straight forward from what we understand. Could anybody help us out with a solution.

Thanks in Advance, Prasad


closed as unclear what you're asking by Voltage Spike, ThreePhaseEel, Dmitry Grigoryev, Peter Smith, uint128_t Dec 18 '16 at 21:01

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Is the color sensor attached to your product, or will it belong to an external piece of calibration equipment? \$\endgroup\$ – calcium3000 Dec 15 '16 at 18:53
  • \$\begingroup\$ Is there a question here? \$\endgroup\$ – Daniel Dec 15 '16 at 19:09
  • \$\begingroup\$ We had a project about emitting white light with a CCT range of 3000-6000K using RGB LEDs. Tried a lot of RGB LEDs: cheap Chinese strips, Cree and even Osram.. Tried even 65536 steps PWM. But we got nothing. One example: We tried to drive an RGB strip with PWM and measured CCT of the mix after every single duty cycle step. When we measured it as 4200K with a calibrated chromameter, it looked "pink" compared to real 4200K white strip. After about 1.5 years of R&D, we concluded that emitting white light with RGB LEDs is nearly impossible. If you find a good solution, I'd like to hear. Good luck! \$\endgroup\$ – Rohat Kılıç Dec 15 '16 at 19:24
  • 1
    \$\begingroup\$ @Passerby This question follows the format: We want to do a thing. We tried a thing. Thing didn't work. ????? Sorry not sorry, but the more I study color the more I realize that the stuff in the ?????? could get someone a PhD. Jonk knows what's up. \$\endgroup\$ – Daniel Dec 16 '16 at 4:21
  • 1
    \$\begingroup\$ You can't get 'good' white out of RGB. It's just not possible. the CCT doesn't mean much when the CRI is 40! \$\endgroup\$ – Daniel Dec 16 '16 at 14:57