LED bulb handy for custom setups

[ray_parkhurst]

hmm, I'm not so sure about that. Usually one can drive even white (ie blue led with yellow phosphor) with PMW without significant color shift.

From what I've read the main cause of color shifts are binning to start with (ie manufacturing tolerances) and temperature (obviously PMW and current can have a significant effect on temperature).

But even if they shift colors, the effect on CRI will be very small since the main problems are usually the gaps in cyan (around 480nm) and deep red (over 630nm), which won't be affected a lot by temperature/current (they will shift a few nm, but the gaps will still be there)

PWM does not vary the current through the LED, only the percentage of time it is driven.

It's easy to verify on a given LED if the color is constant over drive level. Just drive at 10%, 50%, 100%, and 140% and check the color temp. I've done that on a few LEDs and found that I needed noticeable WB shifts to compensate for the different currents.

I'm not saying the wavelengths will shift, only that the relative intensities across different wavelengths will change.

[chris_ma]

PWM does not vary the current through the LED, only the percentage of time it is driven.

yes I know that, what I meant that it will effect temperature. sorry for being unclear about that.

It's easy to verify on a given LED if the color is constant over drive level. Just drive at 10%, 50%, 100%, and 140% and check the color temp. I've done that on a few LEDs and found that I needed noticeable WB shifts to compensate for the different currents.

Interesting, this could also partly due to temperature though since higher current leads to higher temperature.
But the main point I was trying to make was not that color doesn't shift, but the CRI probably depends very little on how we drive an LED.

I'm not saying the wavelengths will shift, only that the relative intensities across different wavelengths will change.

which wouldn't affect CRI significantly.
Color temperature is a rather loose term on LEDs anyway, what is typically used is correlated color temperature (CCT) which is not saying too much because two LEDs with the same CCT can have a widely different white points (basically only specified a line in the chromatic diagram rather then a fixed point).

CRI is something else again and does not depend on the CCT.

chris

[ray_parkhurst]

PWM does not vary the current through the LED, only the percentage of time it is driven.

yes I know that, what I meant that it will effect temperature. sorry for being unclear about that.

It's easy to verify on a given LED if the color is constant over drive level. Just drive at 10%, 50%, 100%, and 140% and check the color temp. I've done that on a few LEDs and found that I needed noticeable WB shifts to compensate for the different currents.

Interesting, this could also partly due to temperature though since higher current leads to higher temperature.
But the main point I was trying to make was not that color doesn't shift, but the CRI probably depends very little on how we drive an LED.

I'm not saying the wavelengths will shift, only that the relative intensities across different wavelengths will change.

which wouldn't affect CRI significantly.
Color temperature is a rather loose term on LEDs anyway, what is typically used is correlated color temperature (CCT) which is not saying too much because two LEDs with the same CCT can have a widely different white points (basically only specified a line in the chromatic diagram rather then a fixed point).

CRI is something else again and does not depend on the CCT.

chris

Yes, it's for sure a pretty complicated set of parameters. I have not done much analysis on different LEDs which all white balance correctly but due to different CRIs have a different captured color balance. I did buy that Pico target a while back and intend to do such studies, but have not found the time or motivation yet to do the study.