I am trying to determine resistor values to set the forward current per color of an RGB LED.

I would like to achieve approximate D65 (daylight white) when all LED channels are driven at maximum illuminance. This article details the process.

I am confused on how to calculate the x,y chromaticity coordinates from the LED datasheet if given the typical wavelength and luminous intensity.

Essentially what I am asking is: Are there any equations or methods to set the white balance of an RGB LED to D65 at maximum illuminance?

Here is a datasheet: enter image description here

  • \$\begingroup\$ @jonk I am aware of CIE 1931 and the color theory standards, I am just looking to ballpark the current values per color so when each channel is on the white point is around D65. Obviously we can then calibrate in software with the driver chip from there. \$\endgroup\$
    – usman510
    Jul 15 at 16:16
  • \$\begingroup\$ @jonk thanks for your insight. Yes I am attempting to achieve a very niche and somewhat complex goal - I am just starting out as an engineer and am working on designing LED lighting. I am building a MATLAB model to assist me in the long term, plugging in various values from each LED datasheet. We will most likely be driving the LED driver ICs using PWM through serial communication, so that is where we can set color. \$\endgroup\$
    – usman510
    Jul 15 at 17:24
  • \$\begingroup\$ Since you'll have to measure this stuff anyway to confirm, you may as well just set everything up for color calibration and not worry about deriving stuff ahead of time. That's how I did it the few times I needed a calibrated color out of RGB LEDs. To maintain the color over time, you'd need to characterize the aging. It takes a long time, since there's no way to accelerate it. \$\endgroup\$ Jul 16 at 6:07
  • \$\begingroup\$ @Kuba The OP isn't just calibrating color out of RGB LEDs. That can easily be a relative process. Getting to the D60 or D65 white points (and there are others) is about an international standard. Like calibrating a current mirror current up to NIST standards that would be usable in an experimental setup that could be replicated by independent researchers. It's not 'in my garage' kind of stuff. This is why the need for calibration standards (NIST or DIN traceable VIS-NIR Tungsten Halogen, for example, which ages fast) for the spectrophotometer and then the application of CIE models to that. \$\endgroup\$
    – jonk
    Jul 16 at 8:52


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.